diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -2,2753 +2,1973 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.7777777777777778,
+  "epoch": 0.7142857142857143,
   "eval_steps": 1000,
-  "global_step": 35000,
+  "global_step": 25000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 2.2222222222222223e-05,
-      "grad_norm": 1.0980026721954346,
+      "epoch": 2.857142857142857e-05,
+      "grad_norm": 1.1264785528182983,
       "learning_rate": 0.0,
-      "loss": 1.6613,
+      "loss": 1.4622,
       "step": 1
     },
     {
-      "epoch": 0.0022222222222222222,
-      "grad_norm": 1.037724256515503,
-      "learning_rate": 1.1e-06,
-      "loss": 1.5146,
+      "epoch": 0.002857142857142857,
+      "grad_norm": 1.0415701866149902,
+      "learning_rate": 1.4142857142857144e-06,
+      "loss": 1.4319,
       "step": 100
     },
     {
-      "epoch": 0.0044444444444444444,
-      "grad_norm": 1.0775114297866821,
-      "learning_rate": 2.2111111111111113e-06,
-      "loss": 1.5285,
+      "epoch": 0.005714285714285714,
+      "grad_norm": 1.1746091842651367,
+      "learning_rate": 2.8428571428571432e-06,
+      "loss": 1.4189,
       "step": 200
     },
     {
-      "epoch": 0.006666666666666667,
-      "grad_norm": 1.065529227256775,
-      "learning_rate": 3.322222222222222e-06,
-      "loss": 1.5211,
+      "epoch": 0.008571428571428572,
+      "grad_norm": 1.1301525831222534,
+      "learning_rate": 4.271428571428572e-06,
+      "loss": 1.4293,
       "step": 300
     },
     {
-      "epoch": 0.008888888888888889,
-      "grad_norm": 1.0316686630249023,
-      "learning_rate": 4.433333333333334e-06,
-      "loss": 1.502,
+      "epoch": 0.011428571428571429,
+      "grad_norm": 1.0607796907424927,
+      "learning_rate": 4.9999753285470756e-06,
+      "loss": 1.4205,
       "step": 400
     },
     {
-      "epoch": 0.011111111111111112,
-      "grad_norm": 1.034487247467041,
-      "learning_rate": 4.999985075284212e-06,
-      "loss": 1.5011,
+      "epoch": 0.014285714285714285,
+      "grad_norm": 1.1491715908050537,
+      "learning_rate": 4.999771876927458e-06,
+      "loss": 1.4197,
       "step": 500
     },
     {
-      "epoch": 0.013333333333333334,
-      "grad_norm": 1.0352669954299927,
-      "learning_rate": 4.999861998793635e-06,
-      "loss": 1.5045,
+      "epoch": 0.017142857142857144,
+      "grad_norm": 1.0873078107833862,
+      "learning_rate": 4.999362935318198e-06,
+      "loss": 1.4364,
       "step": 600
     },
     {
-      "epoch": 0.015555555555555555,
-      "grad_norm": 1.096940517425537,
-      "learning_rate": 4.999614608725878e-06,
-      "loss": 1.5074,
+      "epoch": 0.02,
+      "grad_norm": 1.0659881830215454,
+      "learning_rate": 4.998748537335728e-06,
+      "loss": 1.4507,
       "step": 700
     },
     {
-      "epoch": 0.017777777777777778,
-      "grad_norm": 1.0137542486190796,
-      "learning_rate": 4.999242917383197e-06,
-      "loss": 1.525,
+      "epoch": 0.022857142857142857,
+      "grad_norm": 1.1764490604400635,
+      "learning_rate": 4.99792873348571e-06,
+      "loss": 1.4398,
       "step": 800
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.089389443397522,
-      "learning_rate": 4.998746943249126e-06,
-      "loss": 1.5126,
+      "epoch": 0.025714285714285714,
+      "grad_norm": 1.0576765537261963,
+      "learning_rate": 4.996903591158886e-06,
+      "loss": 1.4203,
       "step": 900
     },
     {
-      "epoch": 0.022222222222222223,
-      "grad_norm": 1.0824161767959595,
-      "learning_rate": 4.998126710987552e-06,
-      "loss": 1.5018,
+      "epoch": 0.02857142857142857,
+      "grad_norm": 1.111843228340149,
+      "learning_rate": 4.995673194625541e-06,
+      "loss": 1.4203,
       "step": 1000
     },
     {
-      "epoch": 0.022222222222222223,
-      "eval_loss": 1.5183053016662598,
-      "eval_runtime": 103.5736,
-      "eval_samples_per_second": 132.225,
-      "eval_steps_per_second": 2.066,
+      "epoch": 0.02857142857142857,
+      "eval_loss": 1.4505008459091187,
+      "eval_runtime": 103.009,
+      "eval_samples_per_second": 132.95,
+      "eval_steps_per_second": 2.077,
       "step": 1000
     },
     {
-      "epoch": 0.024444444444444446,
-      "grad_norm": 0.9782128930091858,
-      "learning_rate": 4.997382251441495e-06,
-      "loss": 1.4755,
+      "epoch": 0.03142857142857143,
+      "grad_norm": 1.037828803062439,
+      "learning_rate": 4.994237645028573e-06,
+      "loss": 1.443,
       "step": 1100
     },
     {
-      "epoch": 0.02666666666666667,
-      "grad_norm": 1.07709538936615,
-      "learning_rate": 4.996513601631571e-06,
-      "loss": 1.4818,
+      "epoch": 0.03428571428571429,
+      "grad_norm": 1.1225452423095703,
+      "learning_rate": 4.992597060375177e-06,
+      "loss": 1.4519,
       "step": 1200
     },
     {
-      "epoch": 0.028888888888888888,
-      "grad_norm": 1.0353506803512573,
-      "learning_rate": 4.99552080475415e-06,
-      "loss": 1.4891,
+      "epoch": 0.037142857142857144,
+      "grad_norm": 1.032313346862793,
+      "learning_rate": 4.990751575527151e-06,
+      "loss": 1.4358,
       "step": 1300
     },
     {
-      "epoch": 0.03111111111111111,
-      "grad_norm": 1.0822919607162476,
-      "learning_rate": 4.994403910179209e-06,
-      "loss": 1.515,
+      "epoch": 0.04,
+      "grad_norm": 1.1252490282058716,
+      "learning_rate": 4.988701342189802e-06,
+      "loss": 1.4102,
       "step": 1400
     },
     {
-      "epoch": 0.03333333333333333,
-      "grad_norm": 1.0209099054336548,
-      "learning_rate": 4.993162973447879e-06,
-      "loss": 1.4816,
+      "epoch": 0.04285714285714286,
+      "grad_norm": 1.0545426607131958,
+      "learning_rate": 4.986446528899478e-06,
+      "loss": 1.4142,
       "step": 1500
     },
     {
-      "epoch": 0.035555555555555556,
-      "grad_norm": 1.089007019996643,
-      "learning_rate": 4.99179805626968e-06,
-      "loss": 1.5098,
+      "epoch": 0.045714285714285714,
+      "grad_norm": 1.08208429813385,
+      "learning_rate": 4.983987321009718e-06,
+      "loss": 1.4247,
       "step": 1600
     },
     {
-      "epoch": 0.03777777777777778,
-      "grad_norm": 1.144646167755127,
-      "learning_rate": 4.990309226519452e-06,
-      "loss": 1.5034,
+      "epoch": 0.04857142857142857,
+      "grad_norm": 1.042827844619751,
+      "learning_rate": 4.98132392067601e-06,
+      "loss": 1.4078,
       "step": 1700
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 1.1199110746383667,
-      "learning_rate": 4.988696558233985e-06,
-      "loss": 1.4831,
+      "epoch": 0.05142857142857143,
+      "grad_norm": 1.029168725013733,
+      "learning_rate": 4.978456546839175e-06,
+      "loss": 1.4255,
       "step": 1800
     },
     {
-      "epoch": 0.042222222222222223,
-      "grad_norm": 1.0624563694000244,
-      "learning_rate": 4.986960131608329e-06,
-      "loss": 1.4673,
+      "epoch": 0.054285714285714284,
+      "grad_norm": 1.1674017906188965,
+      "learning_rate": 4.975385435207367e-06,
+      "loss": 1.4428,
       "step": 1900
     },
     {
-      "epoch": 0.044444444444444446,
-      "grad_norm": 1.0910651683807373,
-      "learning_rate": 4.985100032991814e-06,
-      "loss": 1.4661,
+      "epoch": 0.05714285714285714,
+      "grad_norm": 1.2838454246520996,
+      "learning_rate": 4.972110838236704e-06,
+      "loss": 1.4327,
       "step": 2000
     },
     {
-      "epoch": 0.044444444444444446,
-      "eval_loss": 1.5076568126678467,
-      "eval_runtime": 98.7691,
-      "eval_samples_per_second": 138.657,
-      "eval_steps_per_second": 2.167,
+      "epoch": 0.05714285714285714,
+      "eval_loss": 1.4425562620162964,
+      "eval_runtime": 97.037,
+      "eval_samples_per_second": 141.132,
+      "eval_steps_per_second": 2.205,
       "step": 2000
     },
     {
-      "epoch": 0.04666666666666667,
-      "grad_norm": 1.0498489141464233,
-      "learning_rate": 4.9831163548837506e-06,
-      "loss": 1.4798,
+      "epoch": 0.06,
+      "grad_norm": 1.0720206499099731,
+      "learning_rate": 4.968633025110507e-06,
+      "loss": 1.4312,
       "step": 2100
     },
     {
-      "epoch": 0.04888888888888889,
-      "grad_norm": 1.042786717414856,
-      "learning_rate": 4.9810091959288324e-06,
-      "loss": 1.5013,
+      "epoch": 0.06285714285714286,
+      "grad_norm": 1.0312304496765137,
+      "learning_rate": 4.964952281717177e-06,
+      "loss": 1.4405,
       "step": 2200
     },
     {
-      "epoch": 0.051111111111111114,
-      "grad_norm": 1.0386837720870972,
-      "learning_rate": 4.978778660912231e-06,
-      "loss": 1.4999,
+      "epoch": 0.06571428571428571,
+      "grad_norm": 1.0791317224502563,
+      "learning_rate": 4.961068910626692e-06,
+      "loss": 1.4407,
       "step": 2300
     },
     {
-      "epoch": 0.05333333333333334,
-      "grad_norm": 1.0705187320709229,
-      "learning_rate": 4.976424860754386e-06,
-      "loss": 1.4647,
+      "epoch": 0.06857142857142857,
+      "grad_norm": 1.0809016227722168,
+      "learning_rate": 4.956983231065733e-06,
+      "loss": 1.434,
       "step": 2400
     },
     {
-      "epoch": 0.05555555555555555,
-      "grad_norm": 1.0342007875442505,
-      "learning_rate": 4.973947912505481e-06,
-      "loss": 1.4925,
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.059635043144226,
+      "learning_rate": 4.952695578891449e-06,
+      "loss": 1.4114,
       "step": 2500
     },
     {
-      "epoch": 0.057777777777777775,
-      "grad_norm": 1.073187232017517,
-      "learning_rate": 4.971347939339638e-06,
-      "loss": 1.4865,
+      "epoch": 0.07428571428571429,
+      "grad_norm": 1.0659129619598389,
+      "learning_rate": 4.948206306563842e-06,
+      "loss": 1.4374,
       "step": 2600
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 1.103935718536377,
-      "learning_rate": 4.968625070548778e-06,
-      "loss": 1.4802,
+      "epoch": 0.07714285714285714,
+      "grad_norm": 1.0818511247634888,
+      "learning_rate": 4.943515783116794e-06,
+      "loss": 1.4196,
       "step": 2700
     },
     {
-      "epoch": 0.06222222222222222,
-      "grad_norm": 1.07283616065979,
-      "learning_rate": 4.965779441536202e-06,
-      "loss": 1.4717,
+      "epoch": 0.08,
+      "grad_norm": 1.1003646850585938,
+      "learning_rate": 4.9386243941277374e-06,
+      "loss": 1.4508,
       "step": 2800
     },
     {
-      "epoch": 0.06444444444444444,
-      "grad_norm": 1.0456539392471313,
-      "learning_rate": 4.962811193809845e-06,
-      "loss": 1.5038,
+      "epoch": 0.08285714285714285,
+      "grad_norm": 1.086207628250122,
+      "learning_rate": 4.933532541685949e-06,
+      "loss": 1.4354,
       "step": 2900
     },
     {
-      "epoch": 0.06666666666666667,
-      "grad_norm": 1.0673285722732544,
-      "learning_rate": 4.959720474975257e-06,
-      "loss": 1.4866,
+      "epoch": 0.08571428571428572,
+      "grad_norm": 1.0702838897705078,
+      "learning_rate": 4.928240644359507e-06,
+      "loss": 1.4262,
       "step": 3000
     },
     {
-      "epoch": 0.06666666666666667,
-      "eval_loss": 1.501826524734497,
-      "eval_runtime": 99.5576,
-      "eval_samples_per_second": 137.559,
-      "eval_steps_per_second": 2.15,
+      "epoch": 0.08571428571428572,
+      "eval_loss": 1.438844919204712,
+      "eval_runtime": 96.9421,
+      "eval_samples_per_second": 141.27,
+      "eval_steps_per_second": 2.208,
       "step": 3000
     },
     {
-      "epoch": 0.06888888888888889,
-      "grad_norm": 1.043306827545166,
-      "learning_rate": 4.956507438728246e-06,
-      "loss": 1.5114,
+      "epoch": 0.08857142857142856,
+      "grad_norm": 1.1206424236297607,
+      "learning_rate": 4.922749137160875e-06,
+      "loss": 1.4445,
       "step": 3100
     },
     {
-      "epoch": 0.07111111111111111,
-      "grad_norm": 1.004623293876648,
-      "learning_rate": 4.953172244847243e-06,
-      "loss": 1.483,
+      "epoch": 0.09142857142857143,
+      "grad_norm": 1.0971518754959106,
+      "learning_rate": 4.917058471511149e-06,
+      "loss": 1.4117,
       "step": 3200
     },
     {
-      "epoch": 0.07333333333333333,
-      "grad_norm": 1.0519309043884277,
-      "learning_rate": 4.949715059185356e-06,
-      "loss": 1.4949,
+      "epoch": 0.09428571428571429,
+      "grad_norm": 1.1263982057571411,
+      "learning_rate": 4.9111691152029436e-06,
+      "loss": 1.4294,
       "step": 3300
     },
     {
-      "epoch": 0.07555555555555556,
-      "grad_norm": 1.048803687095642,
-      "learning_rate": 4.9461360536621225e-06,
-      "loss": 1.4688,
+      "epoch": 0.09714285714285714,
+      "grad_norm": 1.0150455236434937,
+      "learning_rate": 4.905081552361943e-06,
+      "loss": 1.4357,
       "step": 3400
     },
     {
-      "epoch": 0.07777777777777778,
-      "grad_norm": 1.002150058746338,
-      "learning_rate": 4.942435406254959e-06,
-      "loss": 1.4721,
+      "epoch": 0.1,
+      "grad_norm": 1.0511361360549927,
+      "learning_rate": 4.898796283407099e-06,
+      "loss": 1.438,
       "step": 3500
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 1.043863296508789,
-      "learning_rate": 4.9386133009903086e-06,
-      "loss": 1.4819,
+      "epoch": 0.10285714285714286,
+      "grad_norm": 1.1033008098602295,
+      "learning_rate": 4.892313825009499e-06,
+      "loss": 1.4162,
       "step": 3600
     },
     {
-      "epoch": 0.08222222222222222,
-      "grad_norm": 1.0765719413757324,
-      "learning_rate": 4.934669927934496e-06,
-      "loss": 1.4905,
+      "epoch": 0.10571428571428572,
+      "grad_norm": 1.1107470989227295,
+      "learning_rate": 4.885634710049891e-06,
+      "loss": 1.4267,
       "step": 3700
     },
     {
-      "epoch": 0.08444444444444445,
-      "grad_norm": 1.0662930011749268,
-      "learning_rate": 4.93060548318427e-06,
-      "loss": 1.491,
+      "epoch": 0.10857142857142857,
+      "grad_norm": 1.0580041408538818,
+      "learning_rate": 4.878759487574882e-06,
+      "loss": 1.4287,
       "step": 3800
     },
     {
-      "epoch": 0.08666666666666667,
-      "grad_norm": 1.0453810691833496,
-      "learning_rate": 4.926420168857054e-06,
-      "loss": 1.496,
+      "epoch": 0.11142857142857143,
+      "grad_norm": 1.0198274850845337,
+      "learning_rate": 4.871688722751799e-06,
+      "loss": 1.438,
       "step": 3900
     },
     {
-      "epoch": 0.08888888888888889,
-      "grad_norm": 0.9957125782966614,
-      "learning_rate": 4.922114193080893e-06,
-      "loss": 1.5116,
+      "epoch": 0.11428571428571428,
+      "grad_norm": 1.1063220500946045,
+      "learning_rate": 4.864422996822239e-06,
+      "loss": 1.4078,
       "step": 4000
     },
     {
-      "epoch": 0.08888888888888889,
-      "eval_loss": 1.5021920204162598,
-      "eval_runtime": 99.9798,
-      "eval_samples_per_second": 136.978,
-      "eval_steps_per_second": 2.14,
+      "epoch": 0.11428571428571428,
+      "eval_loss": 1.4405826330184937,
+      "eval_runtime": 97.1575,
+      "eval_samples_per_second": 140.957,
+      "eval_steps_per_second": 2.203,
       "step": 4000
     },
     {
-      "epoch": 0.09111111111111111,
-      "grad_norm": 1.0647602081298828,
-      "learning_rate": 4.917687769984112e-06,
-      "loss": 1.4787,
+      "epoch": 0.11714285714285715,
+      "grad_norm": 1.0360065698623657,
+      "learning_rate": 4.8569629070542775e-06,
+      "loss": 1.414,
       "step": 4100
     },
     {
-      "epoch": 0.09333333333333334,
-      "grad_norm": 1.0057638883590698,
-      "learning_rate": 4.9131411196846536e-06,
-      "loss": 1.479,
+      "epoch": 0.12,
+      "grad_norm": 1.0539647340774536,
+      "learning_rate": 4.849309066693382e-06,
+      "loss": 1.3992,
       "step": 4200
     },
     {
-      "epoch": 0.09555555555555556,
-      "grad_norm": 1.0767698287963867,
-      "learning_rate": 4.908474468279145e-06,
-      "loss": 1.4646,
+      "epoch": 0.12285714285714286,
+      "grad_norm": 1.0678602457046509,
+      "learning_rate": 4.8414621049119935e-06,
+      "loss": 1.4226,
       "step": 4300
     },
     {
-      "epoch": 0.09777777777777778,
-      "grad_norm": 1.0766099691390991,
-      "learning_rate": 4.903688047831647e-06,
-      "loss": 1.4902,
+      "epoch": 0.12571428571428572,
+      "grad_norm": 1.1174051761627197,
+      "learning_rate": 4.833422666757811e-06,
+      "loss": 1.4149,
       "step": 4400
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 1.0425220727920532,
-      "learning_rate": 4.89878209636212e-06,
-      "loss": 1.4865,
+      "epoch": 0.12857142857142856,
+      "grad_norm": 1.1076269149780273,
+      "learning_rate": 4.825191413100764e-06,
+      "loss": 1.4219,
       "step": 4500
     },
     {
-      "epoch": 0.10222222222222223,
-      "grad_norm": 1.1266907453536987,
-      "learning_rate": 4.893756857834579e-06,
-      "loss": 1.4701,
+      "epoch": 0.13142857142857142,
+      "grad_norm": 1.0237882137298584,
+      "learning_rate": 4.816769020578685e-06,
+      "loss": 1.4063,
       "step": 4600
     },
     {
-      "epoch": 0.10444444444444445,
-      "grad_norm": 1.0631765127182007,
-      "learning_rate": 4.888612582144971e-06,
-      "loss": 1.4875,
+      "epoch": 0.13428571428571429,
+      "grad_norm": 1.0634537935256958,
+      "learning_rate": 4.808156181541694e-06,
+      "loss": 1.4077,
       "step": 4700
     },
     {
-      "epoch": 0.10666666666666667,
-      "grad_norm": 1.0484580993652344,
-      "learning_rate": 4.8833495251087415e-06,
-      "loss": 1.4834,
+      "epoch": 0.13714285714285715,
+      "grad_norm": 1.1134625673294067,
+      "learning_rate": 4.799353603995275e-06,
+      "loss": 1.4589,
       "step": 4800
     },
     {
-      "epoch": 0.10888888888888888,
-      "grad_norm": 1.0509796142578125,
-      "learning_rate": 4.877967948448117e-06,
-      "loss": 1.4828,
+      "epoch": 0.14,
+      "grad_norm": 1.069698691368103,
+      "learning_rate": 4.790362011542085e-06,
+      "loss": 1.4063,
       "step": 4900
     },
     {
-      "epoch": 0.1111111111111111,
-      "grad_norm": 1.0983880758285522,
-      "learning_rate": 4.872468119779088e-06,
-      "loss": 1.458,
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.1093010902404785,
+      "learning_rate": 4.7811821433224665e-06,
+      "loss": 1.4225,
       "step": 5000
     },
     {
-      "epoch": 0.1111111111111111,
-      "eval_loss": 1.5034407377243042,
-      "eval_runtime": 100.2041,
-      "eval_samples_per_second": 136.671,
-      "eval_steps_per_second": 2.136,
+      "epoch": 0.14285714285714285,
+      "eval_loss": 1.4438061714172363,
+      "eval_runtime": 98.2976,
+      "eval_samples_per_second": 139.322,
+      "eval_steps_per_second": 2.177,
       "step": 5000
     },
     {
-      "epoch": 0.11333333333333333,
-      "grad_norm": 1.065687894821167,
-      "learning_rate": 4.866850312598101e-06,
-      "loss": 1.4696,
+      "epoch": 0.1457142857142857,
+      "grad_norm": 1.0884599685668945,
+      "learning_rate": 4.7718147539536865e-06,
+      "loss": 1.4347,
       "step": 5100
     },
     {
-      "epoch": 0.11555555555555555,
-      "grad_norm": 1.0348268747329712,
-      "learning_rate": 4.8611148062684625e-06,
-      "loss": 1.4943,
+      "epoch": 0.14857142857142858,
+      "grad_norm": 1.0088622570037842,
+      "learning_rate": 4.762260613467909e-06,
+      "loss": 1.4254,
       "step": 5200
     },
     {
-      "epoch": 0.11777777777777777,
-      "grad_norm": 1.1039857864379883,
-      "learning_rate": 4.855261886006437e-06,
-      "loss": 1.4881,
+      "epoch": 0.15142857142857144,
+      "grad_norm": 1.1340473890304565,
+      "learning_rate": 4.75252050724889e-06,
+      "loss": 1.4101,
       "step": 5300
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 1.048181414604187,
-      "learning_rate": 4.849291842867075e-06,
-      "loss": 1.4849,
+      "epoch": 0.15428571428571428,
+      "grad_norm": 1.093491554260254,
+      "learning_rate": 4.7425952359674225e-06,
+      "loss": 1.4256,
       "step": 5400
     },
     {
-      "epoch": 0.12222222222222222,
-      "grad_norm": 1.1050995588302612,
-      "learning_rate": 4.84320497372973e-06,
-      "loss": 1.4778,
+      "epoch": 0.15714285714285714,
+      "grad_norm": 1.0808088779449463,
+      "learning_rate": 4.732485615515511e-06,
+      "loss": 1.4093,
       "step": 5500
     },
     {
-      "epoch": 0.12444444444444444,
-      "grad_norm": 1.084206223487854,
-      "learning_rate": 4.837001581283301e-06,
-      "loss": 1.4881,
+      "epoch": 0.16,
+      "grad_norm": 1.100080132484436,
+      "learning_rate": 4.722192476939309e-06,
+      "loss": 1.4263,
       "step": 5600
     },
     {
-      "epoch": 0.12666666666666668,
-      "grad_norm": 1.0708404779434204,
-      "learning_rate": 4.83068197401118e-06,
-      "loss": 1.4873,
+      "epoch": 0.16285714285714287,
+      "grad_norm": 1.096901774406433,
+      "learning_rate": 4.7117166663708025e-06,
+      "loss": 1.4084,
       "step": 5700
     },
     {
-      "epoch": 0.1288888888888889,
-      "grad_norm": 1.043641448020935,
-      "learning_rate": 4.824246466175909e-06,
-      "loss": 1.4772,
+      "epoch": 0.1657142857142857,
+      "grad_norm": 1.1885929107666016,
+      "learning_rate": 4.7010590449582525e-06,
+      "loss": 1.4146,
       "step": 5800
     },
     {
-      "epoch": 0.13111111111111112,
-      "grad_norm": 1.0868722200393677,
-      "learning_rate": 4.817695377803553e-06,
-      "loss": 1.476,
+      "epoch": 0.16857142857142857,
+      "grad_norm": 1.082043170928955,
+      "learning_rate": 4.690220488795406e-06,
+      "loss": 1.4201,
       "step": 5900
     },
     {
-      "epoch": 0.13333333333333333,
-      "grad_norm": 1.1160252094268799,
-      "learning_rate": 4.8110290346677875e-06,
-      "loss": 1.4829,
+      "epoch": 0.17142857142857143,
+      "grad_norm": 1.0647767782211304,
+      "learning_rate": 4.679201888849481e-06,
+      "loss": 1.436,
       "step": 6000
     },
     {
-      "epoch": 0.13333333333333333,
-      "eval_loss": 1.492410659790039,
-      "eval_runtime": 100.2411,
-      "eval_samples_per_second": 136.621,
-      "eval_steps_per_second": 2.135,
+      "epoch": 0.17142857142857143,
+      "eval_loss": 1.4336808919906616,
+      "eval_runtime": 97.8921,
+      "eval_samples_per_second": 139.899,
+      "eval_steps_per_second": 2.186,
       "step": 6000
     },
     {
-      "epoch": 0.13555555555555557,
-      "grad_norm": 1.0023784637451172,
-      "learning_rate": 4.804247768273695e-06,
-      "loss": 1.4704,
+      "epoch": 0.1742857142857143,
+      "grad_norm": 1.1146217584609985,
+      "learning_rate": 4.668004150887924e-06,
+      "loss": 1.4132,
       "step": 6100
     },
     {
-      "epoch": 0.13777777777777778,
-      "grad_norm": 1.0485386848449707,
-      "learning_rate": 4.797351915841285e-06,
-      "loss": 1.484,
+      "epoch": 0.17714285714285713,
+      "grad_norm": 1.0890520811080933,
+      "learning_rate": 4.656628195403952e-06,
+      "loss": 1.4047,
       "step": 6200
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 1.0743563175201416,
-      "learning_rate": 4.79034182028872e-06,
-      "loss": 1.4831,
+      "epoch": 0.18,
+      "grad_norm": 1.033389687538147,
+      "learning_rate": 4.645074957540887e-06,
+      "loss": 1.4272,
       "step": 6300
     },
     {
-      "epoch": 0.14222222222222222,
-      "grad_norm": 0.961910605430603,
-      "learning_rate": 4.783217830215264e-06,
-      "loss": 1.4857,
+      "epoch": 0.18285714285714286,
+      "grad_norm": 1.1013028621673584,
+      "learning_rate": 4.63334538701528e-06,
+      "loss": 1.4402,
       "step": 6400
     },
     {
-      "epoch": 0.14444444444444443,
-      "grad_norm": 1.056557536125183,
-      "learning_rate": 4.775980299883949e-06,
-      "loss": 1.4888,
+      "epoch": 0.18571428571428572,
+      "grad_norm": 1.0814400911331177,
+      "learning_rate": 4.6214404480388455e-06,
+      "loss": 1.4031,
       "step": 6500
     },
     {
-      "epoch": 0.14666666666666667,
-      "grad_norm": 1.0664172172546387,
-      "learning_rate": 4.768629589203955e-06,
-      "loss": 1.478,
+      "epoch": 0.18857142857142858,
+      "grad_norm": 1.0447463989257812,
+      "learning_rate": 4.609361119239197e-06,
+      "loss": 1.4453,
       "step": 6600
     },
     {
-      "epoch": 0.14888888888888888,
-      "grad_norm": 1.0604673624038696,
-      "learning_rate": 4.761166063712719e-06,
-      "loss": 1.476,
+      "epoch": 0.19142857142857142,
+      "grad_norm": 1.1220800876617432,
+      "learning_rate": 4.5971083935794026e-06,
+      "loss": 1.4148,
       "step": 6700
     },
     {
-      "epoch": 0.1511111111111111,
-      "grad_norm": 1.0558980703353882,
-      "learning_rate": 4.753590094557745e-06,
-      "loss": 1.4832,
+      "epoch": 0.19428571428571428,
+      "grad_norm": 1.107762098312378,
+      "learning_rate": 4.584683278276356e-06,
+      "loss": 1.4285,
       "step": 6800
     },
     {
-      "epoch": 0.15333333333333332,
-      "grad_norm": 1.0912115573883057,
-      "learning_rate": 4.745902058478163e-06,
-      "loss": 1.4816,
+      "epoch": 0.19714285714285715,
+      "grad_norm": 1.1005544662475586,
+      "learning_rate": 4.572086794717985e-06,
+      "loss": 1.4328,
       "step": 6900
     },
     {
-      "epoch": 0.15555555555555556,
-      "grad_norm": 1.0631625652313232,
-      "learning_rate": 4.738102337785981e-06,
-      "loss": 1.4655,
+      "epoch": 0.2,
+      "grad_norm": 1.033148169517517,
+      "learning_rate": 4.559319978379287e-06,
+      "loss": 1.4111,
       "step": 7000
     },
     {
-      "epoch": 0.15555555555555556,
-      "eval_loss": 1.4969485998153687,
-      "eval_runtime": 101.1288,
-      "eval_samples_per_second": 135.421,
-      "eval_steps_per_second": 2.116,
+      "epoch": 0.2,
+      "eval_loss": 1.4392390251159668,
+      "eval_runtime": 97.5914,
+      "eval_samples_per_second": 140.33,
+      "eval_steps_per_second": 2.193,
       "step": 7000
     },
     {
-      "epoch": 0.15777777777777777,
-      "grad_norm": 1.080399990081787,
-      "learning_rate": 4.730191320347084e-06,
-      "loss": 1.4823,
+      "epoch": 0.20285714285714285,
+      "grad_norm": 1.052509069442749,
+      "learning_rate": 4.546383878737207e-06,
+      "loss": 1.4113,
       "step": 7100
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 1.0477869510650635,
-      "learning_rate": 4.722169399561937e-06,
-      "loss": 1.4784,
+      "epoch": 0.2057142857142857,
+      "grad_norm": 1.0561904907226562,
+      "learning_rate": 4.533279559184373e-06,
+      "loss": 1.4275,
       "step": 7200
     },
     {
-      "epoch": 0.1622222222222222,
-      "grad_norm": 1.0932061672210693,
-      "learning_rate": 4.714036974346028e-06,
-      "loss": 1.4761,
+      "epoch": 0.20857142857142857,
+      "grad_norm": 1.0787992477416992,
+      "learning_rate": 4.520008096941676e-06,
+      "loss": 1.4084,
       "step": 7300
     },
     {
-      "epoch": 0.16444444444444445,
-      "grad_norm": 1.0640238523483276,
-      "learning_rate": 4.705794449110029e-06,
-      "loss": 1.4732,
+      "epoch": 0.21142857142857144,
+      "grad_norm": 1.0198429822921753,
+      "learning_rate": 4.506570582969719e-06,
+      "loss": 1.4029,
       "step": 7400
     },
     {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 1.0664732456207275,
-      "learning_rate": 4.697442233739684e-06,
-      "loss": 1.4606,
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.0664575099945068,
+      "learning_rate": 4.492968121879142e-06,
+      "loss": 1.4049,
       "step": 7500
     },
     {
-      "epoch": 0.1688888888888889,
-      "grad_norm": 1.0107780694961548,
-      "learning_rate": 4.688980743575429e-06,
-      "loss": 1.4571,
+      "epoch": 0.21714285714285714,
+      "grad_norm": 1.0929675102233887,
+      "learning_rate": 4.479201831839812e-06,
+      "loss": 1.4169,
       "step": 7600
     },
     {
-      "epoch": 0.1711111111111111,
-      "grad_norm": 1.1194672584533691,
-      "learning_rate": 4.680410399391734e-06,
-      "loss": 1.4884,
+      "epoch": 0.22,
+      "grad_norm": 1.1445673704147339,
+      "learning_rate": 4.465272844488908e-06,
+      "loss": 1.4033,
       "step": 7700
     },
     {
-      "epoch": 0.17333333333333334,
-      "grad_norm": 1.0674372911453247,
-      "learning_rate": 4.671731627376184e-06,
-      "loss": 1.478,
+      "epoch": 0.22285714285714286,
+      "grad_norm": 1.064433217048645,
+      "learning_rate": 4.4511823048378986e-06,
+      "loss": 1.43,
       "step": 7800
     },
     {
-      "epoch": 0.17555555555555555,
-      "grad_norm": 1.0933210849761963,
-      "learning_rate": 4.662944859108278e-06,
-      "loss": 1.4611,
+      "epoch": 0.2257142857142857,
+      "grad_norm": 1.0845831632614136,
+      "learning_rate": 4.436931371178416e-06,
+      "loss": 1.4441,
       "step": 7900
     },
     {
-      "epoch": 0.17777777777777778,
-      "grad_norm": 1.0179678201675415,
-      "learning_rate": 4.654050531537975e-06,
-      "loss": 1.4983,
+      "epoch": 0.22857142857142856,
+      "grad_norm": 1.0980095863342285,
+      "learning_rate": 4.42252121498704e-06,
+      "loss": 1.4015,
       "step": 8000
     },
     {
-      "epoch": 0.17777777777777778,
-      "eval_loss": 1.4869590997695923,
-      "eval_runtime": 100.5,
-      "eval_samples_per_second": 136.269,
-      "eval_steps_per_second": 2.129,
+      "epoch": 0.22857142857142856,
+      "eval_loss": 1.4309405088424683,
+      "eval_runtime": 97.7381,
+      "eval_samples_per_second": 140.119,
+      "eval_steps_per_second": 2.19,
       "step": 8000
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 1.1465884447097778,
-      "learning_rate": 4.645049086963961e-06,
-      "loss": 1.4749,
+      "epoch": 0.23142857142857143,
+      "grad_norm": 1.1431641578674316,
+      "learning_rate": 4.407953020829001e-06,
+      "loss": 1.4249,
       "step": 8100
     },
     {
-      "epoch": 0.18222222222222223,
-      "grad_norm": 1.0336287021636963,
-      "learning_rate": 4.6359409730116546e-06,
-      "loss": 1.4854,
+      "epoch": 0.2342857142857143,
+      "grad_norm": 1.0139048099517822,
+      "learning_rate": 4.393227986260801e-06,
+      "loss": 1.3958,
       "step": 8200
     },
     {
-      "epoch": 0.18444444444444444,
-      "grad_norm": 1.102246642112732,
-      "learning_rate": 4.62672664261095e-06,
-      "loss": 1.4625,
+      "epoch": 0.23714285714285716,
+      "grad_norm": 1.0676871538162231,
+      "learning_rate": 4.378347321731773e-06,
+      "loss": 1.4204,
       "step": 8300
     },
     {
-      "epoch": 0.18666666666666668,
-      "grad_norm": 1.0804390907287598,
-      "learning_rate": 4.617406553973687e-06,
-      "loss": 1.4831,
+      "epoch": 0.24,
+      "grad_norm": 1.1097986698150635,
+      "learning_rate": 4.363312250484577e-06,
+      "loss": 1.4335,
       "step": 8400
     },
     {
-      "epoch": 0.18888888888888888,
-      "grad_norm": 0.9855327606201172,
-      "learning_rate": 4.607981170570875e-06,
-      "loss": 1.4745,
+      "epoch": 0.24285714285714285,
+      "grad_norm": 1.083742380142212,
+      "learning_rate": 4.348124008454644e-06,
+      "loss": 1.436,
       "step": 8500
     },
     {
-      "epoch": 0.19111111111111112,
-      "grad_norm": 1.0226573944091797,
-      "learning_rate": 4.598450961109637e-06,
-      "loss": 1.493,
+      "epoch": 0.24571428571428572,
+      "grad_norm": 1.072716474533081,
+      "learning_rate": 4.332783844168581e-06,
+      "loss": 1.424,
       "step": 8600
     },
     {
-      "epoch": 0.19333333333333333,
-      "grad_norm": 1.175445556640625,
-      "learning_rate": 4.588816399509905e-06,
-      "loss": 1.4798,
+      "epoch": 0.24857142857142858,
+      "grad_norm": 1.1168031692504883,
+      "learning_rate": 4.317293018641536e-06,
+      "loss": 1.4262,
       "step": 8700
     },
     {
-      "epoch": 0.19555555555555557,
-      "grad_norm": 1.1075881719589233,
-      "learning_rate": 4.579077964880855e-06,
-      "loss": 1.4634,
+      "epoch": 0.25142857142857145,
+      "grad_norm": 1.1102938652038574,
+      "learning_rate": 4.301652805273535e-06,
+      "loss": 1.4141,
       "step": 8800
     },
     {
-      "epoch": 0.19777777777777777,
-      "grad_norm": 1.0150587558746338,
-      "learning_rate": 4.569236141497075e-06,
-      "loss": 1.4736,
+      "epoch": 0.2542857142857143,
+      "grad_norm": 1.1052049398422241,
+      "learning_rate": 4.285864489744809e-06,
+      "loss": 1.4221,
       "step": 8900
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 1.0608808994293213,
-      "learning_rate": 4.559291418774489e-06,
-      "loss": 1.4796,
+      "epoch": 0.2571428571428571,
+      "grad_norm": 1.0475815534591675,
+      "learning_rate": 4.269929369910103e-06,
+      "loss": 1.4145,
       "step": 9000
     },
     {
-      "epoch": 0.2,
-      "eval_loss": 1.4837573766708374,
-      "eval_runtime": 102.4509,
-      "eval_samples_per_second": 133.674,
-      "eval_steps_per_second": 2.089,
+      "epoch": 0.2571428571428571,
+      "eval_loss": 1.428357481956482,
+      "eval_runtime": 98.1292,
+      "eval_samples_per_second": 139.561,
+      "eval_steps_per_second": 2.181,
       "step": 9000
     },
     {
-      "epoch": 0.20222222222222222,
-      "grad_norm": 0.9918996095657349,
-      "learning_rate": 4.54924429124602e-06,
-      "loss": 1.4812,
+      "epoch": 0.26,
+      "grad_norm": 1.0066262483596802,
+      "learning_rate": 4.253848755691992e-06,
+      "loss": 1.4049,
       "step": 9100
     },
     {
-      "epoch": 0.20444444444444446,
-      "grad_norm": 1.1614868640899658,
-      "learning_rate": 4.539095258536991e-06,
-      "loss": 1.4708,
+      "epoch": 0.26285714285714284,
+      "grad_norm": 1.131996512413025,
+      "learning_rate": 4.2376239689731955e-06,
+      "loss": 1.3991,
       "step": 9200
     },
     {
-      "epoch": 0.20666666666666667,
-      "grad_norm": 1.1184130907058716,
-      "learning_rate": 4.528844825340286e-06,
-      "loss": 1.4598,
+      "epoch": 0.26571428571428574,
+      "grad_norm": 1.1413109302520752,
+      "learning_rate": 4.2212563434879175e-06,
+      "loss": 1.3744,
       "step": 9300
     },
     {
-      "epoch": 0.2088888888888889,
-      "grad_norm": 1.1380512714385986,
-      "learning_rate": 4.5184935013912505e-06,
-      "loss": 1.4683,
+      "epoch": 0.26857142857142857,
+      "grad_norm": 1.073792576789856,
+      "learning_rate": 4.204747224712209e-06,
+      "loss": 1.422,
       "step": 9400
     },
     {
-      "epoch": 0.2111111111111111,
-      "grad_norm": 1.0119566917419434,
-      "learning_rate": 4.508041801442342e-06,
-      "loss": 1.4877,
+      "epoch": 0.2714285714285714,
+      "grad_norm": 1.0397651195526123,
+      "learning_rate": 4.188097969753363e-06,
+      "loss": 1.4064,
       "step": 9500
     },
     {
-      "epoch": 0.21333333333333335,
-      "grad_norm": 1.0013171434402466,
-      "learning_rate": 4.497490245237534e-06,
-      "loss": 1.4825,
+      "epoch": 0.2742857142857143,
+      "grad_norm": 1.1306557655334473,
+      "learning_rate": 4.171309947238357e-06,
+      "loss": 1.4408,
       "step": 9600
     },
     {
-      "epoch": 0.21555555555555556,
-      "grad_norm": 1.0520776510238647,
-      "learning_rate": 4.48683935748647e-06,
-      "loss": 1.4602,
+      "epoch": 0.27714285714285714,
+      "grad_norm": 1.1982935667037964,
+      "learning_rate": 4.154384537201347e-06,
+      "loss": 1.4151,
       "step": 9700
     },
     {
-      "epoch": 0.21777777777777776,
-      "grad_norm": 1.013963222503662,
-      "learning_rate": 4.476089667838369e-06,
-      "loss": 1.4701,
+      "epoch": 0.28,
+      "grad_norm": 1.1465263366699219,
+      "learning_rate": 4.137323130970225e-06,
+      "loss": 1.4211,
       "step": 9800
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 1.0059009790420532,
-      "learning_rate": 4.465241710855688e-06,
-      "loss": 1.4631,
+      "epoch": 0.28285714285714286,
+      "grad_norm": 0.9817516803741455,
+      "learning_rate": 4.120127131052244e-06,
+      "loss": 1.4089,
       "step": 9900
     },
     {
-      "epoch": 0.2222222222222222,
-      "grad_norm": 1.0225909948349,
-      "learning_rate": 4.4542960259875415e-06,
-      "loss": 1.4647,
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.150546908378601,
+      "learning_rate": 4.1027979510187285e-06,
+      "loss": 1.4191,
       "step": 10000
     },
     {
-      "epoch": 0.2222222222222222,
-      "eval_loss": 1.4831452369689941,
-      "eval_runtime": 102.5084,
-      "eval_samples_per_second": 133.599,
-      "eval_steps_per_second": 2.088,
+      "epoch": 0.2857142857142857,
+      "eval_loss": 1.429203748703003,
+      "eval_runtime": 98.6571,
+      "eval_samples_per_second": 138.814,
+      "eval_steps_per_second": 2.169,
       "step": 10000
     },
     {
-      "epoch": 0.22444444444444445,
-      "grad_norm": 1.0490069389343262,
-      "learning_rate": 4.44325315754287e-06,
-      "loss": 1.4887,
+      "epoch": 0.2885714285714286,
+      "grad_norm": 1.0928316116333008,
+      "learning_rate": 4.085337015388876e-06,
+      "loss": 1.4155,
       "step": 10100
     },
     {
-      "epoch": 0.22666666666666666,
-      "grad_norm": 1.0999810695648193,
-      "learning_rate": 4.432113654663379e-06,
-      "loss": 1.4609,
+      "epoch": 0.2914285714285714,
+      "grad_norm": 1.1372452974319458,
+      "learning_rate": 4.067745759512654e-06,
+      "loss": 1.4229,
       "step": 10200
     },
     {
-      "epoch": 0.2288888888888889,
-      "grad_norm": 1.093153715133667,
-      "learning_rate": 4.420878071296227e-06,
-      "loss": 1.4731,
+      "epoch": 0.29428571428571426,
+      "grad_norm": 1.1249101161956787,
+      "learning_rate": 4.0500256294528084e-06,
+      "loss": 1.4178,
       "step": 10300
     },
     {
-      "epoch": 0.2311111111111111,
-      "grad_norm": 1.0634323358535767,
-      "learning_rate": 4.4095469661664794e-06,
-      "loss": 1.4376,
+      "epoch": 0.29714285714285715,
+      "grad_norm": 1.1340339183807373,
+      "learning_rate": 4.032178081865995e-06,
+      "loss": 1.4125,
       "step": 10400
     },
     {
-      "epoch": 0.23333333333333334,
-      "grad_norm": 1.0290603637695312,
-      "learning_rate": 4.398120902749326e-06,
-      "loss": 1.4456,
+      "epoch": 0.3,
+      "grad_norm": 1.0652027130126953,
+      "learning_rate": 4.014204583883038e-06,
+      "loss": 1.4283,
       "step": 10500
     },
     {
-      "epoch": 0.23555555555555555,
-      "grad_norm": 1.0774028301239014,
-      "learning_rate": 4.386600449242057e-06,
-      "loss": 1.4789,
+      "epoch": 0.3028571428571429,
+      "grad_norm": 1.1057724952697754,
+      "learning_rate": 3.996106612988321e-06,
+      "loss": 1.4046,
       "step": 10600
     },
     {
-      "epoch": 0.23777777777777778,
-      "grad_norm": 1.0706713199615479,
-      "learning_rate": 4.374986178535812e-06,
-      "loss": 1.4872,
+      "epoch": 0.3057142857142857,
+      "grad_norm": 1.089181661605835,
+      "learning_rate": 3.977885656898337e-06,
+      "loss": 1.4199,
       "step": 10700
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 1.0337339639663696,
-      "learning_rate": 4.363278668187086e-06,
-      "loss": 1.4742,
+      "epoch": 0.30857142857142855,
+      "grad_norm": 1.0804879665374756,
+      "learning_rate": 3.959543213439393e-06,
+      "loss": 1.4259,
       "step": 10800
     },
     {
-      "epoch": 0.24222222222222223,
-      "grad_norm": 1.1255840063095093,
-      "learning_rate": 4.351478500389014e-06,
-      "loss": 1.4554,
+      "epoch": 0.31142857142857144,
+      "grad_norm": 1.0948872566223145,
+      "learning_rate": 3.941080790424483e-06,
+      "loss": 1.4143,
       "step": 10900
     },
     {
-      "epoch": 0.24444444444444444,
-      "grad_norm": 1.110970377922058,
-      "learning_rate": 4.3395862619424164e-06,
-      "loss": 1.4661,
+      "epoch": 0.3142857142857143,
+      "grad_norm": 1.1653496026992798,
+      "learning_rate": 3.92249990552934e-06,
+      "loss": 1.4343,
       "step": 11000
     },
     {
-      "epoch": 0.24444444444444444,
-      "eval_loss": 1.4758148193359375,
-      "eval_runtime": 102.2906,
-      "eval_samples_per_second": 133.883,
-      "eval_steps_per_second": 2.092,
+      "epoch": 0.3142857142857143,
+      "eval_loss": 1.4226573705673218,
+      "eval_runtime": 98.9066,
+      "eval_samples_per_second": 138.464,
+      "eval_steps_per_second": 2.164,
       "step": 11000
     },
     {
-      "epoch": 0.24666666666666667,
-      "grad_norm": 1.094433307647705,
-      "learning_rate": 4.3276025442266175e-06,
-      "loss": 1.4567,
+      "epoch": 0.3171428571428571,
+      "grad_norm": 1.0654685497283936,
+      "learning_rate": 3.903802086167676e-06,
+      "loss": 1.4102,
       "step": 11100
     },
     {
-      "epoch": 0.24888888888888888,
-      "grad_norm": 1.097430944442749,
-      "learning_rate": 4.31552794317004e-06,
-      "loss": 1.4803,
+      "epoch": 0.32,
+      "grad_norm": 1.01749849319458,
+      "learning_rate": 3.884988869365626e-06,
+      "loss": 1.408,
       "step": 11200
     },
     {
-      "epoch": 0.2511111111111111,
-      "grad_norm": 1.0710690021514893,
-      "learning_rate": 4.3033630592205675e-06,
-      "loss": 1.437,
+      "epoch": 0.32285714285714284,
+      "grad_norm": 1.1105825901031494,
+      "learning_rate": 3.866061801635399e-06,
+      "loss": 1.4276,
       "step": 11300
     },
     {
-      "epoch": 0.25333333333333335,
-      "grad_norm": 1.1534795761108398,
-      "learning_rate": 4.291108497315691e-06,
-      "loss": 1.4698,
+      "epoch": 0.32571428571428573,
+      "grad_norm": 1.0666981935501099,
+      "learning_rate": 3.8470224388481485e-06,
+      "loss": 1.3964,
       "step": 11400
     },
     {
-      "epoch": 0.25555555555555554,
-      "grad_norm": 1.019482970237732,
-      "learning_rate": 4.27876486685242e-06,
-      "loss": 1.4622,
+      "epoch": 0.32857142857142857,
+      "grad_norm": 1.090728759765625,
+      "learning_rate": 3.827872346106073e-06,
+      "loss": 1.3981,
       "step": 11500
     },
     {
-      "epoch": 0.2577777777777778,
-      "grad_norm": 1.1455426216125488,
-      "learning_rate": 4.26633278165698e-06,
-      "loss": 1.4656,
+      "epoch": 0.3314285714285714,
+      "grad_norm": 1.069846272468567,
+      "learning_rate": 3.808613097613759e-06,
+      "loss": 1.4053,
       "step": 11600
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 1.0760246515274048,
-      "learning_rate": 4.253812859954292e-06,
-      "loss": 1.4768,
+      "epoch": 0.3342857142857143,
+      "grad_norm": 1.1465699672698975,
+      "learning_rate": 3.7892462765487836e-06,
+      "loss": 1.3944,
       "step": 11700
     },
     {
-      "epoch": 0.26222222222222225,
-      "grad_norm": 1.0122239589691162,
-      "learning_rate": 4.241205724337223e-06,
-      "loss": 1.4773,
+      "epoch": 0.33714285714285713,
+      "grad_norm": 1.068352222442627,
+      "learning_rate": 3.769773474931558e-06,
+      "loss": 1.4284,
       "step": 11800
     },
     {
-      "epoch": 0.2644444444444444,
-      "grad_norm": 1.0859477519989014,
-      "learning_rate": 4.22851200173563e-06,
-      "loss": 1.4737,
+      "epoch": 0.34,
+      "grad_norm": 1.1487313508987427,
+      "learning_rate": 3.7501962934944704e-06,
+      "loss": 1.3894,
       "step": 11900
     },
     {
-      "epoch": 0.26666666666666666,
-      "grad_norm": 1.0537041425704956,
-      "learning_rate": 4.2157323233851855e-06,
-      "loss": 1.4655,
+      "epoch": 0.34285714285714286,
+      "grad_norm": 1.1034648418426514,
+      "learning_rate": 3.7305163415502936e-06,
+      "loss": 1.4184,
       "step": 12000
     },
     {
-      "epoch": 0.26666666666666666,
-      "eval_loss": 1.4727766513824463,
-      "eval_runtime": 102.6842,
-      "eval_samples_per_second": 133.37,
-      "eval_steps_per_second": 2.084,
+      "epoch": 0.34285714285714286,
+      "eval_loss": 1.4204550981521606,
+      "eval_runtime": 99.0133,
+      "eval_samples_per_second": 138.315,
+      "eval_steps_per_second": 2.161,
       "step": 12000
     },
     {
-      "epoch": 0.2688888888888889,
-      "grad_norm": 1.1021567583084106,
-      "learning_rate": 4.202867324795982e-06,
-      "loss": 1.4693,
+      "epoch": 0.3457142857142857,
+      "grad_norm": 1.08807373046875,
+      "learning_rate": 3.710735236859895e-06,
+      "loss": 1.4208,
       "step": 12100
     },
     {
-      "epoch": 0.27111111111111114,
-      "grad_norm": 0.9724922180175781,
-      "learning_rate": 4.18991764572093e-06,
-      "loss": 1.4585,
+      "epoch": 0.3485714285714286,
+      "grad_norm": 1.142823338508606,
+      "learning_rate": 3.6908546054992523e-06,
+      "loss": 1.4292,
       "step": 12200
     },
     {
-      "epoch": 0.2733333333333333,
-      "grad_norm": 1.140974760055542,
-      "learning_rate": 4.17688393012395e-06,
-      "loss": 1.4437,
+      "epoch": 0.3514285714285714,
+      "grad_norm": 1.0997464656829834,
+      "learning_rate": 3.670876081725784e-06,
+      "loss": 1.4058,
       "step": 12300
     },
     {
-      "epoch": 0.27555555555555555,
-      "grad_norm": 1.05283784866333,
-      "learning_rate": 4.163766826147943e-06,
-      "loss": 1.477,
+      "epoch": 0.35428571428571426,
+      "grad_norm": 1.1083920001983643,
+      "learning_rate": 3.650801307844004e-06,
+      "loss": 1.4152,
       "step": 12400
     },
     {
-      "epoch": 0.2777777777777778,
-      "grad_norm": 1.0787185430526733,
-      "learning_rate": 4.1505669860825645e-06,
-      "loss": 1.4659,
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.1371042728424072,
+      "learning_rate": 3.630631934070527e-06,
+      "loss": 1.4259,
       "step": 12500
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 1.0746320486068726,
-      "learning_rate": 4.137285066331781e-06,
-      "loss": 1.459,
+      "epoch": 0.36,
+      "grad_norm": 1.0470432043075562,
+      "learning_rate": 3.610369618398404e-06,
+      "loss": 1.3952,
       "step": 12600
     },
     {
-      "epoch": 0.2822222222222222,
-      "grad_norm": 1.0252412557601929,
-      "learning_rate": 4.123921727381234e-06,
-      "loss": 1.4802,
+      "epoch": 0.3628571428571429,
+      "grad_norm": 1.0853626728057861,
+      "learning_rate": 3.5900160264608395e-06,
+      "loss": 1.4005,
       "step": 12700
     },
     {
-      "epoch": 0.28444444444444444,
-      "grad_norm": 1.0793858766555786,
-      "learning_rate": 4.110477633765395e-06,
-      "loss": 1.4574,
+      "epoch": 0.3657142857142857,
+      "grad_norm": 1.0409729480743408,
+      "learning_rate": 3.569572831394265e-06,
+      "loss": 1.431,
       "step": 12800
     },
     {
-      "epoch": 0.2866666666666667,
-      "grad_norm": 1.1628038883209229,
-      "learning_rate": 4.096953454034514e-06,
-      "loss": 1.4711,
+      "epoch": 0.36857142857142855,
+      "grad_norm": 1.1226378679275513,
+      "learning_rate": 3.5490417137007997e-06,
+      "loss": 1.4112,
       "step": 12900
     },
     {
-      "epoch": 0.28888888888888886,
-      "grad_norm": 1.0130637884140015,
-      "learning_rate": 4.08334986072138e-06,
-      "loss": 1.4693,
+      "epoch": 0.37142857142857144,
+      "grad_norm": 1.0430322885513306,
+      "learning_rate": 3.528424361110115e-06,
+      "loss": 1.3999,
       "step": 13000
     },
     {
-      "epoch": 0.28888888888888886,
-      "eval_loss": 1.4742803573608398,
-      "eval_runtime": 102.7825,
-      "eval_samples_per_second": 133.242,
-      "eval_steps_per_second": 2.082,
+      "epoch": 0.37142857142857144,
+      "eval_loss": 1.423007845878601,
+      "eval_runtime": 99.2113,
+      "eval_samples_per_second": 138.039,
+      "eval_steps_per_second": 2.157,
       "step": 13000
     },
     {
-      "epoch": 0.2911111111111111,
-      "grad_norm": 1.0166757106781006,
-      "learning_rate": 4.069667530307876e-06,
-      "loss": 1.4463,
+      "epoch": 0.3742857142857143,
+      "grad_norm": 1.1154820919036865,
+      "learning_rate": 3.507722468440688e-06,
+      "loss": 1.4097,
       "step": 13100
     },
     {
-      "epoch": 0.29333333333333333,
-      "grad_norm": 1.0768380165100098,
-      "learning_rate": 4.0559071431913335e-06,
-      "loss": 1.4553,
+      "epoch": 0.37714285714285717,
+      "grad_norm": 1.1299182176589966,
+      "learning_rate": 3.4869377374604886e-06,
+      "loss": 1.4064,
       "step": 13200
     },
     {
-      "epoch": 0.29555555555555557,
-      "grad_norm": 1.0084826946258545,
-      "learning_rate": 4.042069383650704e-06,
-      "loss": 1.4671,
+      "epoch": 0.38,
+      "grad_norm": 1.1046215295791626,
+      "learning_rate": 3.4660718767470854e-06,
+      "loss": 1.4234,
       "step": 13300
     },
     {
-      "epoch": 0.29777777777777775,
-      "grad_norm": 1.0666933059692383,
-      "learning_rate": 4.028154939812527e-06,
-      "loss": 1.4785,
+      "epoch": 0.38285714285714284,
+      "grad_norm": 1.0251668691635132,
+      "learning_rate": 3.445126601547193e-06,
+      "loss": 1.4097,
       "step": 13400
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 1.0286009311676025,
-      "learning_rate": 4.014164503616713e-06,
-      "loss": 1.4562,
+      "epoch": 0.38571428571428573,
+      "grad_norm": 1.0839489698410034,
+      "learning_rate": 3.4241036336356757e-06,
+      "loss": 1.401,
       "step": 13500
     },
     {
-      "epoch": 0.3022222222222222,
-      "grad_norm": 1.1066243648529053,
-      "learning_rate": 4.000098770782136e-06,
-      "loss": 1.4636,
+      "epoch": 0.38857142857142857,
+      "grad_norm": 1.0709606409072876,
+      "learning_rate": 3.40300470117401e-06,
+      "loss": 1.4164,
       "step": 13600
     },
     {
-      "epoch": 0.30444444444444446,
-      "grad_norm": 1.0686429738998413,
-      "learning_rate": 3.985958440772031e-06,
-      "loss": 1.4483,
+      "epoch": 0.3914285714285714,
+      "grad_norm": 1.0628767013549805,
+      "learning_rate": 3.3818315385682255e-06,
+      "loss": 1.409,
       "step": 13700
     },
     {
-      "epoch": 0.30666666666666664,
-      "grad_norm": 0.9844673871994019,
-      "learning_rate": 3.971744216759216e-06,
-      "loss": 1.47,
+      "epoch": 0.3942857142857143,
+      "grad_norm": 1.0831209421157837,
+      "learning_rate": 3.3605858863263274e-06,
+      "loss": 1.4073,
       "step": 13800
     },
     {
-      "epoch": 0.3088888888888889,
-      "grad_norm": 1.184783697128296,
-      "learning_rate": 3.957456805591122e-06,
-      "loss": 1.4476,
+      "epoch": 0.39714285714285713,
+      "grad_norm": 1.1459494829177856,
+      "learning_rate": 3.339269490915223e-06,
+      "loss": 1.4147,
       "step": 13900
     },
     {
-      "epoch": 0.3111111111111111,
-      "grad_norm": 1.0774155855178833,
-      "learning_rate": 3.943096917754647e-06,
-      "loss": 1.4708,
+      "epoch": 0.4,
+      "grad_norm": 1.0614882707595825,
+      "learning_rate": 3.317884104617155e-06,
+      "loss": 1.4089,
       "step": 14000
     },
     {
-      "epoch": 0.3111111111111111,
-      "eval_loss": 1.46890389919281,
-      "eval_runtime": 102.6188,
-      "eval_samples_per_second": 133.455,
-      "eval_steps_per_second": 2.085,
+      "epoch": 0.4,
+      "eval_loss": 1.4181102514266968,
+      "eval_runtime": 99.8701,
+      "eval_samples_per_second": 137.128,
+      "eval_steps_per_second": 2.143,
       "step": 14000
     },
     {
-      "epoch": 0.31333333333333335,
-      "grad_norm": 1.1095852851867676,
-      "learning_rate": 3.928665267340817e-06,
-      "loss": 1.4476,
+      "epoch": 0.40285714285714286,
+      "grad_norm": 1.0587329864501953,
+      "learning_rate": 3.2964314853856593e-06,
+      "loss": 1.3895,
       "step": 14100
     },
     {
-      "epoch": 0.31555555555555553,
-      "grad_norm": 1.0171401500701904,
-      "learning_rate": 3.9141625720092825e-06,
-      "loss": 1.4399,
+      "epoch": 0.4057142857142857,
+      "grad_norm": 1.1020365953445435,
+      "learning_rate": 3.2749133967010545e-06,
+      "loss": 1.4037,
       "step": 14200
     },
     {
-      "epoch": 0.31777777777777777,
-      "grad_norm": 1.0451526641845703,
-      "learning_rate": 3.899589552952628e-06,
-      "loss": 1.463,
+      "epoch": 0.4085714285714286,
+      "grad_norm": 1.1230683326721191,
+      "learning_rate": 3.253331607425475e-06,
+      "loss": 1.4018,
       "step": 14300
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 1.0842229127883911,
-      "learning_rate": 3.8849469348605085e-06,
-      "loss": 1.4435,
+      "epoch": 0.4114285714285714,
+      "grad_norm": 1.0774966478347778,
+      "learning_rate": 3.231687891657469e-06,
+      "loss": 1.4087,
       "step": 14400
     },
     {
-      "epoch": 0.32222222222222224,
-      "grad_norm": 1.0401058197021484,
-      "learning_rate": 3.8702354458836124e-06,
-      "loss": 1.4554,
+      "epoch": 0.4142857142857143,
+      "grad_norm": 1.0514012575149536,
+      "learning_rate": 3.209984028586157e-06,
+      "loss": 1.3861,
       "step": 14500
     },
     {
-      "epoch": 0.3244444444444444,
-      "grad_norm": 1.1635191440582275,
-      "learning_rate": 3.855455817597451e-06,
-      "loss": 1.4612,
+      "epoch": 0.41714285714285715,
+      "grad_norm": 1.1025465726852417,
+      "learning_rate": 3.188221802344978e-06,
+      "loss": 1.4038,
       "step": 14600
     },
     {
-      "epoch": 0.32666666666666666,
-      "grad_norm": 1.0651050806045532,
-      "learning_rate": 3.840608784965979e-06,
-      "loss": 1.4636,
+      "epoch": 0.42,
+      "grad_norm": 1.139419436454773,
+      "learning_rate": 3.16640300186503e-06,
+      "loss": 1.4033,
       "step": 14700
     },
     {
-      "epoch": 0.3288888888888889,
-      "grad_norm": 1.0558030605316162,
-      "learning_rate": 3.825695086305046e-06,
-      "loss": 1.4519,
+      "epoch": 0.4228571428571429,
+      "grad_norm": 1.043289303779602,
+      "learning_rate": 3.1445294207280093e-06,
+      "loss": 1.3867,
       "step": 14800
     },
     {
-      "epoch": 0.33111111111111113,
-      "grad_norm": 1.0227941274642944,
-      "learning_rate": 3.8107154632456845e-06,
-      "loss": 1.4731,
+      "epoch": 0.4257142857142857,
+      "grad_norm": 1.101967453956604,
+      "learning_rate": 3.1226028570187737e-06,
+      "loss": 1.391,
       "step": 14900
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 1.0800951719284058,
-      "learning_rate": 3.7956706606972214e-06,
-      "loss": 1.4547,
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.0626415014266968,
+      "learning_rate": 3.1006251131775342e-06,
+      "loss": 1.3949,
       "step": 15000
     },
     {
-      "epoch": 0.3333333333333333,
-      "eval_loss": 1.4689823389053345,
-      "eval_runtime": 102.3222,
-      "eval_samples_per_second": 133.842,
-      "eval_steps_per_second": 2.091,
+      "epoch": 0.42857142857142855,
+      "eval_loss": 1.4195818901062012,
+      "eval_runtime": 99.5817,
+      "eval_samples_per_second": 137.525,
+      "eval_steps_per_second": 2.149,
       "step": 15000
     },
     {
-      "epoch": 0.33555555555555555,
-      "grad_norm": 1.0613428354263306,
-      "learning_rate": 3.780561426810247e-06,
-      "loss": 1.446,
+      "epoch": 0.43142857142857144,
+      "grad_norm": 1.1212193965911865,
+      "learning_rate": 3.078597995851689e-06,
+      "loss": 1.4007,
       "step": 15100
     },
     {
-      "epoch": 0.3377777777777778,
-      "grad_norm": 1.0628856420516968,
-      "learning_rate": 3.7653885129393996e-06,
-      "loss": 1.4708,
+      "epoch": 0.4342857142857143,
+      "grad_norm": 1.0601767301559448,
+      "learning_rate": 3.056523315747308e-06,
+      "loss": 1.4098,
       "step": 15200
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 1.077598214149475,
-      "learning_rate": 3.7501526736060113e-06,
-      "loss": 1.4434,
+      "epoch": 0.43714285714285717,
+      "grad_norm": 1.0668915510177612,
+      "learning_rate": 3.034402887480287e-06,
+      "loss": 1.3885,
       "step": 15300
     },
     {
-      "epoch": 0.3422222222222222,
-      "grad_norm": 1.0646636486053467,
-      "learning_rate": 3.7348546664605777e-06,
-      "loss": 1.4583,
+      "epoch": 0.44,
+      "grad_norm": 1.0714190006256104,
+      "learning_rate": 3.012238529427181e-06,
+      "loss": 1.4018,
       "step": 15400
     },
     {
-      "epoch": 0.34444444444444444,
-      "grad_norm": 1.07997465133667,
-      "learning_rate": 3.7194952522450905e-06,
-      "loss": 1.4523,
+      "epoch": 0.44285714285714284,
+      "grad_norm": 1.1230597496032715,
+      "learning_rate": 2.9900320635757293e-06,
+      "loss": 1.4086,
       "step": 15500
     },
     {
-      "epoch": 0.3466666666666667,
-      "grad_norm": 1.0550297498703003,
-      "learning_rate": 3.7040751947552013e-06,
-      "loss": 1.4629,
+      "epoch": 0.44571428571428573,
+      "grad_norm": 1.0094853639602661,
+      "learning_rate": 2.9677853153750763e-06,
+      "loss": 1.3801,
       "step": 15600
     },
     {
-      "epoch": 0.3488888888888889,
-      "grad_norm": 1.084794044494629,
-      "learning_rate": 3.68859526080224e-06,
-      "loss": 1.4588,
+      "epoch": 0.44857142857142857,
+      "grad_norm": 1.0972274541854858,
+      "learning_rate": 2.9455001135857194e-06,
+      "loss": 1.3985,
       "step": 15700
     },
     {
-      "epoch": 0.3511111111111111,
-      "grad_norm": 1.0864589214324951,
-      "learning_rate": 3.6730562201750864e-06,
-      "loss": 1.427,
+      "epoch": 0.4514285714285714,
+      "grad_norm": 1.0266581773757935,
+      "learning_rate": 2.9231782901291726e-06,
+      "loss": 1.4124,
       "step": 15800
     },
     {
-      "epoch": 0.35333333333333333,
-      "grad_norm": 1.1361236572265625,
-      "learning_rate": 3.657458845601885e-06,
-      "loss": 1.45,
+      "epoch": 0.4542857142857143,
+      "grad_norm": 1.138675332069397,
+      "learning_rate": 2.900821679937382e-06,
+      "loss": 1.4173,
       "step": 15900
     },
     {
-      "epoch": 0.35555555555555557,
-      "grad_norm": 1.110909342765808,
-      "learning_rate": 3.6418039127116235e-06,
-      "loss": 1.4784,
+      "epoch": 0.45714285714285713,
+      "grad_norm": 1.1691060066223145,
+      "learning_rate": 2.8784321208018817e-06,
+      "loss": 1.4123,
       "step": 16000
     },
     {
-      "epoch": 0.35555555555555557,
-      "eval_loss": 1.4744384288787842,
-      "eval_runtime": 102.4825,
-      "eval_samples_per_second": 133.633,
-      "eval_steps_per_second": 2.088,
+      "epoch": 0.45714285714285713,
+      "eval_loss": 1.4248454570770264,
+      "eval_runtime": 99.8569,
+      "eval_samples_per_second": 137.146,
+      "eval_steps_per_second": 2.143,
       "step": 16000
     },
     {
-      "epoch": 0.35777777777777775,
-      "grad_norm": 1.0018236637115479,
-      "learning_rate": 3.626092199995557e-06,
-      "loss": 1.4492,
+      "epoch": 0.46,
+      "grad_norm": 1.1149132251739502,
+      "learning_rate": 2.8560114532227262e-06,
+      "loss": 1.4171,
       "step": 16100
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 1.0787028074264526,
-      "learning_rate": 3.6103244887684985e-06,
-      "loss": 1.442,
+      "epoch": 0.46285714285714286,
+      "grad_norm": 1.0276226997375488,
+      "learning_rate": 2.8335615202571927e-06,
+      "loss": 1.4177,
       "step": 16200
     },
     {
-      "epoch": 0.3622222222222222,
-      "grad_norm": 1.0067973136901855,
-      "learning_rate": 3.594501563129966e-06,
-      "loss": 1.448,
+      "epoch": 0.4657142857142857,
+      "grad_norm": 1.0828535556793213,
+      "learning_rate": 2.811084167368276e-06,
+      "loss": 1.3762,
       "step": 16300
     },
     {
-      "epoch": 0.36444444444444446,
-      "grad_norm": 1.009955883026123,
-      "learning_rate": 3.5786242099251895e-06,
-      "loss": 1.4541,
+      "epoch": 0.4685714285714286,
+      "grad_norm": 1.171616554260254,
+      "learning_rate": 2.788581242272983e-06,
+      "loss": 1.3965,
       "step": 16400
     },
     {
-      "epoch": 0.36666666666666664,
-      "grad_norm": 1.0146301984786987,
-      "learning_rate": 3.5626932187059807e-06,
-      "loss": 1.444,
+      "epoch": 0.4714285714285714,
+      "grad_norm": 1.0692201852798462,
+      "learning_rate": 2.7660545947904464e-06,
+      "loss": 1.4066,
       "step": 16500
     },
     {
-      "epoch": 0.3688888888888889,
-      "grad_norm": 1.093392252922058,
-      "learning_rate": 3.5467093816914744e-06,
-      "loss": 1.4682,
+      "epoch": 0.4742857142857143,
+      "grad_norm": 1.1563397645950317,
+      "learning_rate": 2.7435060766898614e-06,
+      "loss": 1.4008,
       "step": 16600
     },
     {
-      "epoch": 0.3711111111111111,
-      "grad_norm": 1.0555778741836548,
-      "learning_rate": 3.530673493728727e-06,
-      "loss": 1.4741,
+      "epoch": 0.47714285714285715,
+      "grad_norm": 1.1032534837722778,
+      "learning_rate": 2.7209375415382655e-06,
+      "loss": 1.3905,
       "step": 16700
     },
     {
-      "epoch": 0.37333333333333335,
-      "grad_norm": 1.0543639659881592,
-      "learning_rate": 3.5145863522531974e-06,
-      "loss": 1.4686,
+      "epoch": 0.48,
+      "grad_norm": 1.1357022523880005,
+      "learning_rate": 2.698350844548168e-06,
+      "loss": 1.406,
       "step": 16800
     },
     {
-      "epoch": 0.37555555555555553,
-      "grad_norm": 1.0254998207092285,
-      "learning_rate": 3.498448757249086e-06,
-      "loss": 1.4523,
+      "epoch": 0.4828571428571429,
+      "grad_norm": 1.0574637651443481,
+      "learning_rate": 2.6757478424250417e-06,
+      "loss": 1.4049,
       "step": 16900
     },
     {
-      "epoch": 0.37777777777777777,
-      "grad_norm": 1.180134892463684,
-      "learning_rate": 3.482261511209556e-06,
-      "loss": 1.4353,
+      "epoch": 0.4857142857142857,
+      "grad_norm": 1.0180025100708008,
+      "learning_rate": 2.653130393214702e-06,
+      "loss": 1.3979,
       "step": 17000
     },
     {
-      "epoch": 0.37777777777777777,
-      "eval_loss": 1.468135952949524,
-      "eval_runtime": 102.9245,
-      "eval_samples_per_second": 133.059,
-      "eval_steps_per_second": 2.079,
+      "epoch": 0.4857142857142857,
+      "eval_loss": 1.4195657968521118,
+      "eval_runtime": 99.9485,
+      "eval_samples_per_second": 137.021,
+      "eval_steps_per_second": 2.141,
       "step": 17000
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 1.0833312273025513,
-      "learning_rate": 3.4660254190968247e-06,
-      "loss": 1.4664,
+      "epoch": 0.48857142857142855,
+      "grad_norm": 1.0153673887252808,
+      "learning_rate": 2.630500356150565e-06,
+      "loss": 1.4138,
       "step": 17100
     },
     {
-      "epoch": 0.38222222222222224,
-      "grad_norm": 1.0613282918930054,
-      "learning_rate": 3.4497412883021375e-06,
-      "loss": 1.46,
+      "epoch": 0.49142857142857144,
+      "grad_norm": 1.0832693576812744,
+      "learning_rate": 2.6078595915008096e-06,
+      "loss": 1.3934,
       "step": 17200
     },
     {
-      "epoch": 0.3844444444444444,
-      "grad_norm": 1.073068380355835,
-      "learning_rate": 3.4334099286056134e-06,
-      "loss": 1.4558,
+      "epoch": 0.4942857142857143,
+      "grad_norm": 1.1552319526672363,
+      "learning_rate": 2.585209960415464e-06,
+      "loss": 1.414,
       "step": 17300
     },
     {
-      "epoch": 0.38666666666666666,
-      "grad_norm": 1.0136057138442993,
-      "learning_rate": 3.4170321521359816e-06,
-      "loss": 1.4775,
+      "epoch": 0.49714285714285716,
+      "grad_norm": 1.1260509490966797,
+      "learning_rate": 2.562553324773404e-06,
+      "loss": 1.3988,
       "step": 17400
     },
     {
-      "epoch": 0.3888888888888889,
-      "grad_norm": 1.0413048267364502,
-      "learning_rate": 3.40060877333019e-06,
-      "loss": 1.4519,
+      "epoch": 0.5,
+      "grad_norm": 1.1187398433685303,
+      "learning_rate": 2.5398915470293077e-06,
+      "loss": 1.4048,
       "step": 17500
     },
     {
-      "epoch": 0.39111111111111113,
-      "grad_norm": 1.090867519378662,
-      "learning_rate": 3.38414060889291e-06,
-      "loss": 1.4653,
+      "epoch": 0.5028571428571429,
+      "grad_norm": 1.0673401355743408,
+      "learning_rate": 2.5172264900605497e-06,
+      "loss": 1.4012,
       "step": 17600
     },
     {
-      "epoch": 0.3933333333333333,
-      "grad_norm": 1.0485255718231201,
-      "learning_rate": 3.367628477755917e-06,
-      "loss": 1.434,
+      "epoch": 0.5057142857142857,
+      "grad_norm": 1.098514199256897,
+      "learning_rate": 2.49456001701407e-06,
+      "loss": 1.4021,
       "step": 17700
     },
     {
-      "epoch": 0.39555555555555555,
-      "grad_norm": 1.0650479793548584,
-      "learning_rate": 3.351073201037374e-06,
-      "loss": 1.4626,
+      "epoch": 0.5085714285714286,
+      "grad_norm": 1.1217247247695923,
+      "learning_rate": 2.471893991153216e-06,
+      "loss": 1.4041,
       "step": 17800
     },
     {
-      "epoch": 0.3977777777777778,
-      "grad_norm": 1.0615607500076294,
-      "learning_rate": 3.3344756020009934e-06,
-      "loss": 1.4471,
+      "epoch": 0.5114285714285715,
+      "grad_norm": 1.1324173212051392,
+      "learning_rate": 2.4492302757045705e-06,
+      "loss": 1.3942,
       "step": 17900
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.9974454045295715,
-      "learning_rate": 3.3178365060150995e-06,
-      "loss": 1.4449,
+      "epoch": 0.5142857142857142,
+      "grad_norm": 1.1281129121780396,
+      "learning_rate": 2.426570733704798e-06,
+      "loss": 1.4046,
       "step": 18000
     },
     {
-      "epoch": 0.4,
-      "eval_loss": 1.464233636856079,
-      "eval_runtime": 103.2861,
-      "eval_samples_per_second": 132.593,
-      "eval_steps_per_second": 2.072,
+      "epoch": 0.5142857142857142,
+      "eval_loss": 1.4171615839004517,
+      "eval_runtime": 99.7442,
+      "eval_samples_per_second": 137.301,
+      "eval_steps_per_second": 2.145,
       "step": 18000
     },
     {
-      "epoch": 0.4022222222222222,
-      "grad_norm": 1.020620346069336,
-      "learning_rate": 3.301156740511585e-06,
-      "loss": 1.4585,
+      "epoch": 0.5171428571428571,
+      "grad_norm": 1.084283471107483,
+      "learning_rate": 2.4039172278474864e-06,
+      "loss": 1.4183,
       "step": 18100
     },
     {
-      "epoch": 0.40444444444444444,
-      "grad_norm": 1.100151777267456,
-      "learning_rate": 3.2844371349447607e-06,
-      "loss": 1.4553,
+      "epoch": 0.52,
+      "grad_norm": 1.0714788436889648,
+      "learning_rate": 2.381271620330034e-06,
+      "loss": 1.3793,
       "step": 18200
     },
     {
-      "epoch": 0.4066666666666667,
-      "grad_norm": 1.0522452592849731,
-      "learning_rate": 3.267678520750115e-06,
-      "loss": 1.4571,
+      "epoch": 0.5228571428571429,
+      "grad_norm": 1.1440812349319458,
+      "learning_rate": 2.358635772700567e-06,
+      "loss": 1.3765,
       "step": 18300
     },
     {
-      "epoch": 0.4088888888888889,
-      "grad_norm": 1.082814335823059,
-      "learning_rate": 3.2508817313029604e-06,
-      "loss": 1.4408,
+      "epoch": 0.5257142857142857,
+      "grad_norm": 1.0656503438949585,
+      "learning_rate": 2.336011545704916e-06,
+      "loss": 1.4153,
       "step": 18400
     },
     {
-      "epoch": 0.4111111111111111,
-      "grad_norm": 0.9813875555992126,
-      "learning_rate": 3.234047601876999e-06,
-      "loss": 1.45,
+      "epoch": 0.5285714285714286,
+      "grad_norm": 1.1328638792037964,
+      "learning_rate": 2.3134007991336523e-06,
+      "loss": 1.3873,
       "step": 18500
     },
     {
-      "epoch": 0.41333333333333333,
-      "grad_norm": 1.1190879344940186,
-      "learning_rate": 3.2171769696027775e-06,
-      "loss": 1.4611,
+      "epoch": 0.5314285714285715,
+      "grad_norm": 1.0806158781051636,
+      "learning_rate": 2.290805391669212e-06,
+      "loss": 1.3774,
       "step": 18600
     },
     {
-      "epoch": 0.41555555555555557,
-      "grad_norm": 1.1239498853683472,
-      "learning_rate": 3.200270673426068e-06,
-      "loss": 1.4492,
+      "epoch": 0.5342857142857143,
+      "grad_norm": 1.069150686264038,
+      "learning_rate": 2.2682271807331003e-06,
+      "loss": 1.3918,
       "step": 18700
     },
     {
-      "epoch": 0.4177777777777778,
-      "grad_norm": 1.0896953344345093,
-      "learning_rate": 3.183329554066138e-06,
-      "loss": 1.4597,
+      "epoch": 0.5371428571428571,
+      "grad_norm": 1.1267215013504028,
+      "learning_rate": 2.2456680223332103e-06,
+      "loss": 1.3845,
       "step": 18800
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 1.0593230724334717,
-      "learning_rate": 3.1663544539739512e-06,
-      "loss": 1.4318,
+      "epoch": 0.54,
+      "grad_norm": 1.142121434211731,
+      "learning_rate": 2.2231297709112496e-06,
+      "loss": 1.4109,
       "step": 18900
     },
     {
-      "epoch": 0.4222222222222222,
-      "grad_norm": 1.0952677726745605,
-      "learning_rate": 3.1493462172902713e-06,
-      "loss": 1.443,
+      "epoch": 0.5428571428571428,
+      "grad_norm": 1.0814783573150635,
+      "learning_rate": 2.2006142791902957e-06,
+      "loss": 1.4098,
       "step": 19000
     },
     {
-      "epoch": 0.4222222222222222,
-      "eval_loss": 1.4632492065429688,
-      "eval_runtime": 102.9259,
-      "eval_samples_per_second": 133.057,
-      "eval_steps_per_second": 2.079,
+      "epoch": 0.5428571428571428,
+      "eval_loss": 1.416707158088684,
+      "eval_runtime": 100.0528,
+      "eval_samples_per_second": 136.878,
+      "eval_steps_per_second": 2.139,
       "step": 19000
     },
     {
-      "epoch": 0.42444444444444446,
-      "grad_norm": 1.0409597158432007,
-      "learning_rate": 3.132305689803684e-06,
-      "loss": 1.4407,
+      "epoch": 0.5457142857142857,
+      "grad_norm": 1.0706247091293335,
+      "learning_rate": 2.1781233980225035e-06,
+      "loss": 1.4188,
       "step": 19100
     },
     {
-      "epoch": 0.4266666666666667,
-      "grad_norm": 1.0839999914169312,
-      "learning_rate": 3.1152337189085372e-06,
-      "loss": 1.4398,
+      "epoch": 0.5485714285714286,
+      "grad_norm": 1.021273136138916,
+      "learning_rate": 2.1556589762369518e-06,
+      "loss": 1.3989,
       "step": 19200
     },
     {
-      "epoch": 0.4288888888888889,
-      "grad_norm": 1.0945332050323486,
-      "learning_rate": 3.0981311535628024e-06,
-      "loss": 1.4478,
+      "epoch": 0.5514285714285714,
+      "grad_norm": 1.1904112100601196,
+      "learning_rate": 2.133222860487667e-06,
+      "loss": 1.4393,
       "step": 19300
     },
     {
-      "epoch": 0.4311111111111111,
-      "grad_norm": 1.0296911001205444,
-      "learning_rate": 3.0809988442458567e-06,
-      "loss": 1.447,
+      "epoch": 0.5542857142857143,
+      "grad_norm": 1.1062791347503662,
+      "learning_rate": 2.1108168951018186e-06,
+      "loss": 1.4045,
       "step": 19400
     },
     {
-      "epoch": 0.43333333333333335,
-      "grad_norm": 1.044171690940857,
-      "learning_rate": 3.063837642916191e-06,
-      "loss": 1.4706,
+      "epoch": 0.5571428571428572,
+      "grad_norm": 1.1809172630310059,
+      "learning_rate": 2.088442921928113e-06,
+      "loss": 1.3958,
       "step": 19500
     },
     {
-      "epoch": 0.43555555555555553,
-      "grad_norm": 1.0633000135421753,
-      "learning_rate": 3.0466484029690453e-06,
-      "loss": 1.4572,
+      "epoch": 0.56,
+      "grad_norm": 1.0156745910644531,
+      "learning_rate": 2.066102780185383e-06,
+      "loss": 1.398,
       "step": 19600
     },
     {
-      "epoch": 0.43777777777777777,
-      "grad_norm": 0.9554187655448914,
-      "learning_rate": 3.0294319791939653e-06,
-      "loss": 1.4392,
+      "epoch": 0.5628571428571428,
+      "grad_norm": 1.1121779680252075,
+      "learning_rate": 2.0437983063114013e-06,
+      "loss": 1.4122,
       "step": 19700
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 1.1261570453643799,
-      "learning_rate": 3.0121892277323016e-06,
-      "loss": 1.4508,
+      "epoch": 0.5657142857142857,
+      "grad_norm": 1.0523419380187988,
+      "learning_rate": 2.021531333811914e-06,
+      "loss": 1.4063,
       "step": 19800
     },
     {
-      "epoch": 0.44222222222222224,
-      "grad_norm": 1.075859785079956,
-      "learning_rate": 2.9949210060346323e-06,
-      "loss": 1.4525,
+      "epoch": 0.5685714285714286,
+      "grad_norm": 1.099584698677063,
+      "learning_rate": 1.9993036931099265e-06,
+      "loss": 1.409,
       "step": 19900
     },
     {
-      "epoch": 0.4444444444444444,
-      "grad_norm": 1.121369481086731,
-      "learning_rate": 2.977628172818122e-06,
-      "loss": 1.458,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.1999467611312866,
+      "learning_rate": 1.9771172113952327e-06,
+      "loss": 1.4,
       "step": 20000
     },
     {
-      "epoch": 0.4444444444444444,
-      "eval_loss": 1.461082935333252,
-      "eval_runtime": 103.0425,
-      "eval_samples_per_second": 132.906,
-      "eval_steps_per_second": 2.077,
+      "epoch": 0.5714285714285714,
+      "eval_loss": 1.415099024772644,
+      "eval_runtime": 99.755,
+      "eval_samples_per_second": 137.286,
+      "eval_steps_per_second": 2.145,
       "step": 20000
     },
     {
-      "epoch": 0.44666666666666666,
-      "grad_norm": 1.0296210050582886,
-      "learning_rate": 2.9603115880238228e-06,
-      "loss": 1.4306,
+      "epoch": 0.5742857142857143,
+      "grad_norm": 1.0494403839111328,
+      "learning_rate": 1.9549737124742104e-06,
+      "loss": 1.4095,
       "step": 20100
     },
     {
-      "epoch": 0.4488888888888889,
-      "grad_norm": 1.0637251138687134,
-      "learning_rate": 2.9429721127739095e-06,
-      "loss": 1.4592,
+      "epoch": 0.5771428571428572,
+      "grad_norm": 1.1081063747406006,
+      "learning_rate": 1.9328750166199046e-06,
+      "loss": 1.3992,
       "step": 20200
     },
     {
-      "epoch": 0.45111111111111113,
-      "grad_norm": 1.0740017890930176,
-      "learning_rate": 2.925610609328858e-06,
-      "loss": 1.4439,
+      "epoch": 0.58,
+      "grad_norm": 1.1197865009307861,
+      "learning_rate": 1.91082294042239e-06,
+      "loss": 1.3917,
       "step": 20300
     },
     {
-      "epoch": 0.4533333333333333,
-      "grad_norm": 1.136069893836975,
-      "learning_rate": 2.9082279410445637e-06,
-      "loss": 1.4483,
+      "epoch": 0.5828571428571429,
+      "grad_norm": 1.140148639678955,
+      "learning_rate": 1.8888192966394448e-06,
+      "loss": 1.3907,
       "step": 20400
     },
     {
-      "epoch": 0.45555555555555555,
-      "grad_norm": 1.0300533771514893,
-      "learning_rate": 2.890824972329414e-06,
-      "loss": 1.4462,
+      "epoch": 0.5857142857142857,
+      "grad_norm": 1.0425162315368652,
+      "learning_rate": 1.8668658940475298e-06,
+      "loss": 1.4006,
       "step": 20500
     },
     {
-      "epoch": 0.4577777777777778,
-      "grad_norm": 1.0724245309829712,
-      "learning_rate": 2.8734025686012967e-06,
-      "loss": 1.4387,
+      "epoch": 0.5885714285714285,
+      "grad_norm": 1.1035826206207275,
+      "learning_rate": 1.8449645372931068e-06,
+      "loss": 1.4033,
       "step": 20600
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 1.0145208835601807,
-      "learning_rate": 2.855961596244571e-06,
-      "loss": 1.452,
+      "epoch": 0.5914285714285714,
+      "grad_norm": 1.1139192581176758,
+      "learning_rate": 1.823117026744287e-06,
+      "loss": 1.3964,
       "step": 20700
     },
     {
-      "epoch": 0.4622222222222222,
-      "grad_norm": 1.0359569787979126,
-      "learning_rate": 2.8385029225669757e-06,
-      "loss": 1.4524,
+      "epoch": 0.5942857142857143,
+      "grad_norm": 1.1130657196044922,
+      "learning_rate": 1.8013251583428366e-06,
+      "loss": 1.3972,
       "step": 20800
     },
     {
-      "epoch": 0.46444444444444444,
-      "grad_norm": 1.0701048374176025,
-      "learning_rate": 2.8210274157565078e-06,
-      "loss": 1.4563,
+      "epoch": 0.5971428571428572,
+      "grad_norm": 1.0860106945037842,
+      "learning_rate": 1.7795907234565385e-06,
+      "loss": 1.3931,
       "step": 20900
     },
     {
-      "epoch": 0.4666666666666667,
-      "grad_norm": 1.106961727142334,
-      "learning_rate": 2.803535944838243e-06,
-      "loss": 1.4433,
+      "epoch": 0.6,
+      "grad_norm": 1.05580472946167,
+      "learning_rate": 1.7579155087319443e-06,
+      "loss": 1.3874,
       "step": 21000
     },
     {
-      "epoch": 0.4666666666666667,
-      "eval_loss": 1.4546458721160889,
-      "eval_runtime": 103.0443,
-      "eval_samples_per_second": 132.904,
-      "eval_steps_per_second": 2.077,
+      "epoch": 0.6,
+      "eval_loss": 1.4096276760101318,
+      "eval_runtime": 99.8984,
+      "eval_samples_per_second": 137.089,
+      "eval_steps_per_second": 2.142,
       "step": 21000
     },
     {
-      "epoch": 0.4688888888888889,
-      "grad_norm": 1.019301414489746,
-      "learning_rate": 2.786029379631124e-06,
-      "loss": 1.4205,
+      "epoch": 0.6028571428571429,
+      "grad_norm": 1.1223632097244263,
+      "learning_rate": 1.7363012959475e-06,
+      "loss": 1.3793,
       "step": 21100
     },
     {
-      "epoch": 0.4711111111111111,
-      "grad_norm": 1.132878065109253,
-      "learning_rate": 2.7685085907047038e-06,
-      "loss": 1.4365,
+      "epoch": 0.6057142857142858,
+      "grad_norm": 1.115355372428894,
+      "learning_rate": 1.7147498618670778e-06,
+      "loss": 1.4093,
       "step": 21200
     },
     {
-      "epoch": 0.47333333333333333,
-      "grad_norm": 1.1317287683486938,
-      "learning_rate": 2.7509744493358574e-06,
-      "loss": 1.4391,
+      "epoch": 0.6085714285714285,
+      "grad_norm": 1.0437370538711548,
+      "learning_rate": 1.6932629780939225e-06,
+      "loss": 1.3875,
       "step": 21300
     },
     {
-      "epoch": 0.47555555555555556,
-      "grad_norm": 1.0661753416061401,
-      "learning_rate": 2.7334278274654507e-06,
-      "loss": 1.4392,
+      "epoch": 0.6114285714285714,
+      "grad_norm": 1.0260958671569824,
+      "learning_rate": 1.6718424109250154e-06,
+      "loss": 1.4035,
       "step": 21400
     },
     {
-      "epoch": 0.4777777777777778,
-      "grad_norm": 1.0728217363357544,
-      "learning_rate": 2.7158695976549808e-06,
-      "loss": 1.4284,
+      "epoch": 0.6142857142857143,
+      "grad_norm": 0.9281340837478638,
+      "learning_rate": 1.6504899212058837e-06,
+      "loss": 1.3853,
       "step": 21500
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 1.138945460319519,
-      "learning_rate": 2.6983006330431904e-06,
-      "loss": 1.4318,
+      "epoch": 0.6171428571428571,
+      "grad_norm": 1.1064680814743042,
+      "learning_rate": 1.6292072641858478e-06,
+      "loss": 1.4016,
       "step": 21600
     },
     {
-      "epoch": 0.4822222222222222,
-      "grad_norm": 1.0607870817184448,
-      "learning_rate": 2.6807218073026394e-06,
-      "loss": 1.4507,
+      "epoch": 0.62,
+      "grad_norm": 1.0704963207244873,
+      "learning_rate": 1.6079961893737384e-06,
+      "loss": 1.3848,
       "step": 21700
     },
     {
-      "epoch": 0.48444444444444446,
-      "grad_norm": 1.1019628047943115,
-      "learning_rate": 2.6631339945962682e-06,
-      "loss": 1.4508,
+      "epoch": 0.6228571428571429,
+      "grad_norm": 1.0652328729629517,
+      "learning_rate": 1.5868584403940768e-06,
+      "loss": 1.3749,
       "step": 21800
     },
     {
-      "epoch": 0.4866666666666667,
-      "grad_norm": 1.104874849319458,
-      "learning_rate": 2.6455380695339194e-06,
-      "loss": 1.434,
+      "epoch": 0.6257142857142857,
+      "grad_norm": 1.0742926597595215,
+      "learning_rate": 1.5657957548437447e-06,
+      "loss": 1.404,
       "step": 21900
     },
     {
-      "epoch": 0.4888888888888889,
-      "grad_norm": 1.1439390182495117,
-      "learning_rate": 2.6279349071288506e-06,
-      "loss": 1.4434,
+      "epoch": 0.6285714285714286,
+      "grad_norm": 1.0579770803451538,
+      "learning_rate": 1.5448098641491487e-06,
+      "loss": 1.4036,
       "step": 22000
     },
     {
-      "epoch": 0.4888888888888889,
-      "eval_loss": 1.4595842361450195,
-      "eval_runtime": 103.7757,
-      "eval_samples_per_second": 131.967,
-      "eval_steps_per_second": 2.062,
+      "epoch": 0.6285714285714286,
+      "eval_loss": 1.4158315658569336,
+      "eval_runtime": 100.3868,
+      "eval_samples_per_second": 136.422,
+      "eval_steps_per_second": 2.132,
       "step": 22000
     },
     {
-      "epoch": 0.4911111111111111,
-      "grad_norm": 1.002807378768921,
-      "learning_rate": 2.610325382754217e-06,
-      "loss": 1.4423,
+      "epoch": 0.6314285714285715,
+      "grad_norm": 1.101526141166687,
+      "learning_rate": 1.5239024934238874e-06,
+      "loss": 1.4188,
       "step": 22100
     },
     {
-      "epoch": 0.49333333333333335,
-      "grad_norm": 1.157097578048706,
-      "learning_rate": 2.5927103720995443e-06,
-      "loss": 1.4464,
+      "epoch": 0.6342857142857142,
+      "grad_norm": 1.0752556324005127,
+      "learning_rate": 1.5030753613269455e-06,
+      "loss": 1.3847,
       "step": 22200
     },
     {
-      "epoch": 0.4955555555555556,
-      "grad_norm": 1.0830250978469849,
-      "learning_rate": 2.5750907511271804e-06,
-      "loss": 1.4259,
+      "epoch": 0.6371428571428571,
+      "grad_norm": 1.0786316394805908,
+      "learning_rate": 1.4823301799214101e-06,
+      "loss": 1.3867,
       "step": 22300
     },
     {
-      "epoch": 0.49777777777777776,
-      "grad_norm": 1.1380789279937744,
-      "learning_rate": 2.5574673960287377e-06,
-      "loss": 1.453,
+      "epoch": 0.64,
+      "grad_norm": 1.0339590311050415,
+      "learning_rate": 1.4616686545337374e-06,
+      "loss": 1.3662,
       "step": 22400
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.9866058826446533,
-      "learning_rate": 2.5398411831815173e-06,
-      "loss": 1.4254,
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.0901203155517578,
+      "learning_rate": 1.4410924836135625e-06,
+      "loss": 1.3836,
       "step": 22500
     },
     {
-      "epoch": 0.5022222222222222,
-      "grad_norm": 1.0832710266113281,
-      "learning_rate": 2.522212989104932e-06,
-      "loss": 1.4408,
+      "epoch": 0.6457142857142857,
+      "grad_norm": 1.0557289123535156,
+      "learning_rate": 1.4206033585940895e-06,
+      "loss": 1.375,
       "step": 22600
     },
     {
-      "epoch": 0.5044444444444445,
-      "grad_norm": 1.0661131143569946,
-      "learning_rate": 2.5045836904169164e-06,
-      "loss": 1.4567,
+      "epoch": 0.6485714285714286,
+      "grad_norm": 1.049706220626831,
+      "learning_rate": 1.40020296375304e-06,
+      "loss": 1.3977,
       "step": 22700
     },
     {
-      "epoch": 0.5066666666666667,
-      "grad_norm": 1.0910335779190063,
-      "learning_rate": 2.4869541637903384e-06,
-      "loss": 1.4464,
+      "epoch": 0.6514285714285715,
+      "grad_norm": 1.170900821685791,
+      "learning_rate": 1.379892976074209e-06,
+      "loss": 1.384,
       "step": 22800
     },
     {
-      "epoch": 0.5088888888888888,
-      "grad_norm": 1.051774501800537,
-      "learning_rate": 2.4693252859093982e-06,
-      "loss": 1.4341,
+      "epoch": 0.6542857142857142,
+      "grad_norm": 1.10288667678833,
+      "learning_rate": 1.3596750651096047e-06,
+      "loss": 1.4045,
       "step": 22900
     },
     {
-      "epoch": 0.5111111111111111,
-      "grad_norm": 1.0804002285003662,
-      "learning_rate": 2.4516979334260345e-06,
-      "loss": 1.4552,
+      "epoch": 0.6571428571428571,
+      "grad_norm": 1.0909626483917236,
+      "learning_rate": 1.3395508928422074e-06,
+      "loss": 1.4018,
       "step": 23000
     },
     {
-      "epoch": 0.5111111111111111,
-      "eval_loss": 1.458953619003296,
-      "eval_runtime": 103.1044,
-      "eval_samples_per_second": 132.827,
-      "eval_steps_per_second": 2.076,
+      "epoch": 0.6571428571428571,
+      "eval_loss": 1.4156948328018188,
+      "eval_runtime": 100.678,
+      "eval_samples_per_second": 136.028,
+      "eval_steps_per_second": 2.126,
       "step": 23000
     },
     {
-      "epoch": 0.5133333333333333,
-      "grad_norm": 1.0643597841262817,
-      "learning_rate": 2.4340729829163335e-06,
-      "loss": 1.4319,
+      "epoch": 0.66,
+      "grad_norm": 1.1391985416412354,
+      "learning_rate": 1.3195221135493503e-06,
+      "loss": 1.372,
       "step": 23100
     },
     {
-      "epoch": 0.5155555555555555,
-      "grad_norm": 1.0562705993652344,
-      "learning_rate": 2.416451310836935e-06,
-      "loss": 1.4293,
+      "epoch": 0.6628571428571428,
+      "grad_norm": 1.124377965927124,
+      "learning_rate": 1.2995903736667267e-06,
+      "loss": 1.3998,
       "step": 23200
     },
     {
-      "epoch": 0.5177777777777778,
-      "grad_norm": 1.038696050643921,
-      "learning_rate": 2.3988337934814466e-06,
-      "loss": 1.4542,
+      "epoch": 0.6657142857142857,
+      "grad_norm": 1.1056832075119019,
+      "learning_rate": 1.279757311653056e-06,
+      "loss": 1.3677,
       "step": 23300
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 1.053084373474121,
-      "learning_rate": 2.38122130693687e-06,
-      "loss": 1.4673,
+      "epoch": 0.6685714285714286,
+      "grad_norm": 1.0959793329238892,
+      "learning_rate": 1.2600245578553866e-06,
+      "loss": 1.3801,
       "step": 23400
     },
     {
-      "epoch": 0.5222222222222223,
-      "grad_norm": 1.0920089483261108,
-      "learning_rate": 2.363614727040034e-06,
-      "loss": 1.4402,
+      "epoch": 0.6714285714285714,
+      "grad_norm": 1.0466543436050415,
+      "learning_rate": 1.240393734375086e-06,
+      "loss": 1.3866,
       "step": 23500
     },
     {
-      "epoch": 0.5244444444444445,
-      "grad_norm": 1.0549384355545044,
-      "learning_rate": 2.346014929334041e-06,
-      "loss": 1.4415,
+      "epoch": 0.6742857142857143,
+      "grad_norm": 1.0811994075775146,
+      "learning_rate": 1.2208664549344884e-06,
+      "loss": 1.3885,
       "step": 23600
     },
     {
-      "epoch": 0.5266666666666666,
-      "grad_norm": 1.1243116855621338,
-      "learning_rate": 2.328422789024727e-06,
-      "loss": 1.4323,
+      "epoch": 0.6771428571428572,
+      "grad_norm": 1.1533517837524414,
+      "learning_rate": 1.2014443247442498e-06,
+      "loss": 1.3684,
       "step": 23700
     },
     {
-      "epoch": 0.5288888888888889,
-      "grad_norm": 1.0348533391952515,
-      "learning_rate": 2.3108391809371384e-06,
-      "loss": 1.4137,
+      "epoch": 0.68,
+      "grad_norm": 1.0400276184082031,
+      "learning_rate": 1.1821289403713865e-06,
+      "loss": 1.3733,
       "step": 23800
     },
     {
-      "epoch": 0.5311111111111111,
-      "grad_norm": 1.0663659572601318,
-      "learning_rate": 2.2932649794720327e-06,
-      "loss": 1.4574,
+      "epoch": 0.6828571428571428,
+      "grad_norm": 1.0742322206497192,
+      "learning_rate": 1.1629218896080382e-06,
+      "loss": 1.3884,
       "step": 23900
     },
     {
-      "epoch": 0.5333333333333333,
-      "grad_norm": 1.1175079345703125,
-      "learning_rate": 2.275701058562391e-06,
-      "loss": 1.4467,
+      "epoch": 0.6857142857142857,
+      "grad_norm": 1.0781975984573364,
+      "learning_rate": 1.1438247513409423e-06,
+      "loss": 1.3611,
       "step": 24000
     },
     {
-      "epoch": 0.5333333333333333,
-      "eval_loss": 1.4563220739364624,
-      "eval_runtime": 103.6075,
-      "eval_samples_per_second": 132.182,
-      "eval_steps_per_second": 2.065,
+      "epoch": 0.6857142857142857,
+      "eval_loss": 1.4142818450927734,
+      "eval_runtime": 100.8413,
+      "eval_samples_per_second": 135.807,
+      "eval_steps_per_second": 2.122,
       "step": 24000
     },
     {
-      "epoch": 0.5355555555555556,
-      "grad_norm": 1.1649941205978394,
-      "learning_rate": 2.2581482916299596e-06,
-      "loss": 1.4493,
+      "epoch": 0.6885714285714286,
+      "grad_norm": 1.1411370038986206,
+      "learning_rate": 1.1248390954216437e-06,
+      "loss": 1.3838,
       "step": 24100
     },
     {
-      "epoch": 0.5377777777777778,
-      "grad_norm": 1.0690239667892456,
-      "learning_rate": 2.240607551541823e-06,
-      "loss": 1.4552,
+      "epoch": 0.6914285714285714,
+      "grad_norm": 1.0584548711776733,
+      "learning_rate": 1.1059664825374511e-06,
+      "loss": 1.3482,
       "step": 24200
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 1.074440836906433,
-      "learning_rate": 2.2230797105669876e-06,
-      "loss": 1.4635,
+      "epoch": 0.6942857142857143,
+      "grad_norm": 1.096170425415039,
+      "learning_rate": 1.0872084640831356e-06,
+      "loss": 1.3704,
       "step": 24300
     },
     {
-      "epoch": 0.5422222222222223,
-      "grad_norm": 1.0503307580947876,
-      "learning_rate": 2.205565640333014e-06,
-      "loss": 1.424,
+      "epoch": 0.6971428571428572,
+      "grad_norm": 1.0241279602050781,
+      "learning_rate": 1.068566582033411e-06,
+      "loss": 1.3735,
       "step": 24400
     },
     {
-      "epoch": 0.5444444444444444,
-      "grad_norm": 1.0343286991119385,
-      "learning_rate": 2.188066211782666e-06,
-      "loss": 1.448,
+      "epoch": 0.7,
+      "grad_norm": 1.0666210651397705,
+      "learning_rate": 1.050042368816168e-06,
+      "loss": 1.3893,
       "step": 24500
     },
     {
-      "epoch": 0.5466666666666666,
-      "grad_norm": 1.054713249206543,
-      "learning_rate": 2.1705822951306043e-06,
-      "loss": 1.4259,
+      "epoch": 0.7028571428571428,
+      "grad_norm": 1.0070935487747192,
+      "learning_rate": 1.0316373471865108e-06,
+      "loss": 1.3807,
       "step": 24600
     },
     {
-      "epoch": 0.5488888888888889,
-      "grad_norm": 1.0660899877548218,
-      "learning_rate": 2.153114759820113e-06,
-      "loss": 1.4519,
+      "epoch": 0.7057142857142857,
+      "grad_norm": 1.0485628843307495,
+      "learning_rate": 1.013353030101576e-06,
+      "loss": 1.3817,
       "step": 24700
     },
     {
-      "epoch": 0.5511111111111111,
-      "grad_norm": 1.0503852367401123,
-      "learning_rate": 2.1356644744798592e-06,
-      "loss": 1.4447,
+      "epoch": 0.7085714285714285,
+      "grad_norm": 0.9520274996757507,
+      "learning_rate": 9.951909205961665e-07,
+      "loss": 1.3201,
       "step": 24800
     },
     {
-      "epoch": 0.5533333333333333,
-      "grad_norm": 1.0827716588974,
-      "learning_rate": 2.1182323068807016e-06,
-      "loss": 1.4503,
+      "epoch": 0.7114285714285714,
+      "grad_norm": 1.0479100942611694,
+      "learning_rate": 9.77152511659194e-07,
+      "loss": 1.2627,
       "step": 24900
     },
     {
-      "epoch": 0.5555555555555556,
-      "grad_norm": 1.128435730934143,
-      "learning_rate": 2.1008191238925376e-06,
-      "loss": 1.4327,
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.0204826593399048,
+      "learning_rate": 9.59239286110952e-07,
+      "loss": 1.2352,
       "step": 25000
     },
     {
-      "epoch": 0.5555555555555556,
-      "eval_loss": 1.449695348739624,
-      "eval_runtime": 102.9301,
-      "eval_samples_per_second": 133.051,
-      "eval_steps_per_second": 2.079,
+      "epoch": 0.7142857142857143,
+      "eval_loss": 1.4112086296081543,
+      "eval_runtime": 101.0868,
+      "eval_samples_per_second": 135.478,
+      "eval_steps_per_second": 2.117,
       "step": 25000
-    },
-    {
-      "epoch": 0.5577777777777778,
-      "grad_norm": 1.0508419275283813,
-      "learning_rate": 2.083425791441193e-06,
-      "loss": 1.4257,
-      "step": 25100
-    },
-    {
-      "epoch": 0.56,
-      "grad_norm": 1.131657600402832,
-      "learning_rate": 2.066053174465362e-06,
-      "loss": 1.4478,
-      "step": 25200
-    },
-    {
-      "epoch": 0.5622222222222222,
-      "grad_norm": 1.082889199256897,
-      "learning_rate": 2.0487021368736002e-06,
-      "loss": 1.4415,
-      "step": 25300
-    },
-    {
-      "epoch": 0.5644444444444444,
-      "grad_norm": 1.1404916048049927,
-      "learning_rate": 2.0313735415013548e-06,
-      "loss": 1.4375,
-      "step": 25400
-    },
-    {
-      "epoch": 0.5666666666666667,
-      "grad_norm": 1.0749248266220093,
-      "learning_rate": 2.0140682500680656e-06,
-      "loss": 1.4283,
-      "step": 25500
-    },
-    {
-      "epoch": 0.5688888888888889,
-      "grad_norm": 1.1225584745407104,
-      "learning_rate": 1.9967871231343077e-06,
-      "loss": 1.4431,
-      "step": 25600
-    },
-    {
-      "epoch": 0.5711111111111111,
-      "grad_norm": 1.1027264595031738,
-      "learning_rate": 1.9795310200590015e-06,
-      "loss": 1.4287,
-      "step": 25700
-    },
-    {
-      "epoch": 0.5733333333333334,
-      "grad_norm": 1.0218228101730347,
-      "learning_rate": 1.962300798956676e-06,
-      "loss": 1.4234,
-      "step": 25800
-    },
-    {
-      "epoch": 0.5755555555555556,
-      "grad_norm": 1.0777612924575806,
-      "learning_rate": 1.9450973166547965e-06,
-      "loss": 1.4519,
-      "step": 25900
-    },
-    {
-      "epoch": 0.5777777777777777,
-      "grad_norm": 1.067936658859253,
-      "learning_rate": 1.9279214286511566e-06,
-      "loss": 1.4201,
-      "step": 26000
-    },
-    {
-      "epoch": 0.5777777777777777,
-      "eval_loss": 1.45020592212677,
-      "eval_runtime": 103.4404,
-      "eval_samples_per_second": 132.395,
-      "eval_steps_per_second": 2.069,
-      "step": 26000
-    },
-    {
-      "epoch": 0.58,
-      "grad_norm": 1.053848147392273,
-      "learning_rate": 1.910773989071337e-06,
-      "loss": 1.4288,
-      "step": 26100
-    },
-    {
-      "epoch": 0.5822222222222222,
-      "grad_norm": 1.0881634950637817,
-      "learning_rate": 1.8936558506262287e-06,
-      "loss": 1.4404,
-      "step": 26200
-    },
-    {
-      "epoch": 0.5844444444444444,
-      "grad_norm": 1.0691543817520142,
-      "learning_rate": 1.87656786456963e-06,
-      "loss": 1.4286,
-      "step": 26300
-    },
-    {
-      "epoch": 0.5866666666666667,
-      "grad_norm": 1.0484671592712402,
-      "learning_rate": 1.8595108806559193e-06,
-      "loss": 1.4273,
-      "step": 26400
-    },
-    {
-      "epoch": 0.5888888888888889,
-      "grad_norm": 1.0702567100524902,
-      "learning_rate": 1.8424857470977914e-06,
-      "loss": 1.427,
-      "step": 26500
-    },
-    {
-      "epoch": 0.5911111111111111,
-      "grad_norm": 1.0914722681045532,
-      "learning_rate": 1.8254933105240832e-06,
-      "loss": 1.4324,
-      "step": 26600
-    },
-    {
-      "epoch": 0.5933333333333334,
-      "grad_norm": 1.0758668184280396,
-      "learning_rate": 1.8085344159376694e-06,
-      "loss": 1.4302,
-      "step": 26700
-    },
-    {
-      "epoch": 0.5955555555555555,
-      "grad_norm": 1.0367869138717651,
-      "learning_rate": 1.791609906673442e-06,
-      "loss": 1.4322,
-      "step": 26800
-    },
-    {
-      "epoch": 0.5977777777777777,
-      "grad_norm": 1.142140507698059,
-      "learning_rate": 1.7747206243563758e-06,
-      "loss": 1.4228,
-      "step": 26900
-    },
-    {
-      "epoch": 0.6,
-      "grad_norm": 1.112724781036377,
-      "learning_rate": 1.757867408859672e-06,
-      "loss": 1.436,
-      "step": 27000
-    },
-    {
-      "epoch": 0.6,
-      "eval_loss": 1.4449330568313599,
-      "eval_runtime": 103.9341,
-      "eval_samples_per_second": 131.766,
-      "eval_steps_per_second": 2.059,
-      "step": 27000
-    },
-    {
-      "epoch": 0.6022222222222222,
-      "grad_norm": 1.027860403060913,
-      "learning_rate": 1.7410510982629953e-06,
-      "loss": 1.4382,
-      "step": 27100
-    },
-    {
-      "epoch": 0.6044444444444445,
-      "grad_norm": 1.035631775856018,
-      "learning_rate": 1.7242725288107959e-06,
-      "loss": 1.4201,
-      "step": 27200
-    },
-    {
-      "epoch": 0.6066666666666667,
-      "grad_norm": 1.1015000343322754,
-      "learning_rate": 1.7075325348707283e-06,
-      "loss": 1.4371,
-      "step": 27300
-    },
-    {
-      "epoch": 0.6088888888888889,
-      "grad_norm": 1.0760738849639893,
-      "learning_rate": 1.6908319488921538e-06,
-      "loss": 1.4333,
-      "step": 27400
-    },
-    {
-      "epoch": 0.6111111111111112,
-      "grad_norm": 1.0578911304473877,
-      "learning_rate": 1.6741716013647512e-06,
-      "loss": 1.4551,
-      "step": 27500
-    },
-    {
-      "epoch": 0.6133333333333333,
-      "grad_norm": 1.0643107891082764,
-      "learning_rate": 1.657552320777212e-06,
-      "loss": 1.4336,
-      "step": 27600
-    },
-    {
-      "epoch": 0.6155555555555555,
-      "grad_norm": 1.124281406402588,
-      "learning_rate": 1.6409749335760447e-06,
-      "loss": 1.4425,
-      "step": 27700
-    },
-    {
-      "epoch": 0.6177777777777778,
-      "grad_norm": 1.1095889806747437,
-      "learning_rate": 1.6244402641244755e-06,
-      "loss": 1.4472,
-      "step": 27800
-    },
-    {
-      "epoch": 0.62,
-      "grad_norm": 1.05001962184906,
-      "learning_rate": 1.6079491346614557e-06,
-      "loss": 1.4329,
-      "step": 27900
-    },
-    {
-      "epoch": 0.6222222222222222,
-      "grad_norm": 1.0860042572021484,
-      "learning_rate": 1.5915023652607709e-06,
-      "loss": 1.4324,
-      "step": 28000
-    },
-    {
-      "epoch": 0.6222222222222222,
-      "eval_loss": 1.4518882036209106,
-      "eval_runtime": 104.2615,
-      "eval_samples_per_second": 131.352,
-      "eval_steps_per_second": 2.053,
-      "step": 28000
-    },
-    {
-      "epoch": 0.6244444444444445,
-      "grad_norm": 1.1069949865341187,
-      "learning_rate": 1.5751007737902634e-06,
-      "loss": 1.4316,
-      "step": 28100
-    },
-    {
-      "epoch": 0.6266666666666667,
-      "grad_norm": 1.0824823379516602,
-      "learning_rate": 1.5587451758711589e-06,
-      "loss": 1.4155,
-      "step": 28200
-    },
-    {
-      "epoch": 0.6288888888888889,
-      "grad_norm": 1.1121022701263428,
-      "learning_rate": 1.542436384837507e-06,
-      "loss": 1.453,
-      "step": 28300
-    },
-    {
-      "epoch": 0.6311111111111111,
-      "grad_norm": 1.0836877822875977,
-      "learning_rate": 1.526175211695736e-06,
-      "loss": 1.43,
-      "step": 28400
-    },
-    {
-      "epoch": 0.6333333333333333,
-      "grad_norm": 1.1165013313293457,
-      "learning_rate": 1.5099624650843264e-06,
-      "loss": 1.4568,
-      "step": 28500
-    },
-    {
-      "epoch": 0.6355555555555555,
-      "grad_norm": 1.1114059686660767,
-      "learning_rate": 1.4937989512335923e-06,
-      "loss": 1.4432,
-      "step": 28600
-    },
-    {
-      "epoch": 0.6377777777777778,
-      "grad_norm": 1.0963460206985474,
-      "learning_rate": 1.4776854739255933e-06,
-      "loss": 1.4282,
-      "step": 28700
-    },
-    {
-      "epoch": 0.64,
-      "grad_norm": 1.0809634923934937,
-      "learning_rate": 1.4616228344541627e-06,
-      "loss": 1.4282,
-      "step": 28800
-    },
-    {
-      "epoch": 0.6422222222222222,
-      "grad_norm": 1.084708333015442,
-      "learning_rate": 1.4456118315850618e-06,
-      "loss": 1.4525,
-      "step": 28900
-    },
-    {
-      "epoch": 0.6444444444444445,
-      "grad_norm": 1.0788947343826294,
-      "learning_rate": 1.4296532615162588e-06,
-      "loss": 1.432,
-      "step": 29000
-    },
-    {
-      "epoch": 0.6444444444444445,
-      "eval_loss": 1.4509940147399902,
-      "eval_runtime": 104.7265,
-      "eval_samples_per_second": 130.769,
-      "eval_steps_per_second": 2.043,
-      "step": 29000
-    },
-    {
-      "epoch": 0.6466666666666666,
-      "grad_norm": 1.082764744758606,
-      "learning_rate": 1.4137479178383307e-06,
-      "loss": 1.4224,
-      "step": 29100
-    },
-    {
-      "epoch": 0.6488888888888888,
-      "grad_norm": 1.0896496772766113,
-      "learning_rate": 1.3978965914950065e-06,
-      "loss": 1.4193,
-      "step": 29200
-    },
-    {
-      "epoch": 0.6511111111111111,
-      "grad_norm": 1.1591044664382935,
-      "learning_rate": 1.3821000707438325e-06,
-      "loss": 1.4246,
-      "step": 29300
-    },
-    {
-      "epoch": 0.6533333333333333,
-      "grad_norm": 1.067596673965454,
-      "learning_rate": 1.366359141116968e-06,
-      "loss": 1.4308,
-      "step": 29400
-    },
-    {
-      "epoch": 0.6555555555555556,
-      "grad_norm": 1.0923844575881958,
-      "learning_rate": 1.3506745853821318e-06,
-      "loss": 1.4376,
-      "step": 29500
-    },
-    {
-      "epoch": 0.6577777777777778,
-      "grad_norm": 0.9823111891746521,
-      "learning_rate": 1.3350471835036702e-06,
-      "loss": 1.4265,
-      "step": 29600
-    },
-    {
-      "epoch": 0.66,
-      "grad_norm": 1.122013807296753,
-      "learning_rate": 1.3194777126037726e-06,
-      "loss": 1.4407,
-      "step": 29700
-    },
-    {
-      "epoch": 0.6622222222222223,
-      "grad_norm": 1.2428700923919678,
-      "learning_rate": 1.303966946923827e-06,
-      "loss": 1.4516,
-      "step": 29800
-    },
-    {
-      "epoch": 0.6644444444444444,
-      "grad_norm": 1.0549967288970947,
-      "learning_rate": 1.2885156577859159e-06,
-      "loss": 1.4232,
-      "step": 29900
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 1.0705960988998413,
-      "learning_rate": 1.273124613554464e-06,
-      "loss": 1.4486,
-      "step": 30000
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "eval_loss": 1.4483963251113892,
-      "eval_runtime": 103.4915,
-      "eval_samples_per_second": 132.33,
-      "eval_steps_per_second": 2.068,
-      "step": 30000
-    },
-    {
-      "epoch": 0.6688888888888889,
-      "grad_norm": 1.1372791528701782,
-      "learning_rate": 1.2577945795980268e-06,
-      "loss": 1.4192,
-      "step": 30100
-    },
-    {
-      "epoch": 0.6711111111111111,
-      "grad_norm": 1.0534753799438477,
-      "learning_rate": 1.2425263182512276e-06,
-      "loss": 1.4325,
-      "step": 30200
-    },
-    {
-      "epoch": 0.6733333333333333,
-      "grad_norm": 1.0935174226760864,
-      "learning_rate": 1.2273205887768544e-06,
-      "loss": 1.4232,
-      "step": 30300
-    },
-    {
-      "epoch": 0.6755555555555556,
-      "grad_norm": 1.0945768356323242,
-      "learning_rate": 1.2121781473280947e-06,
-      "loss": 1.4281,
-      "step": 30400
-    },
-    {
-      "epoch": 0.6777777777777778,
-      "grad_norm": 1.0865614414215088,
-      "learning_rate": 1.1970997469109412e-06,
-      "loss": 1.4343,
-      "step": 30500
-    },
-    {
-      "epoch": 0.68,
-      "grad_norm": 1.0924773216247559,
-      "learning_rate": 1.1820861373467426e-06,
-      "loss": 1.4389,
-      "step": 30600
-    },
-    {
-      "epoch": 0.6822222222222222,
-      "grad_norm": 1.1250981092453003,
-      "learning_rate": 1.1671380652349156e-06,
-      "loss": 1.4456,
-      "step": 30700
-    },
-    {
-      "epoch": 0.6844444444444444,
-      "grad_norm": 1.097658634185791,
-      "learning_rate": 1.1522562739158195e-06,
-      "loss": 1.426,
-      "step": 30800
-    },
-    {
-      "epoch": 0.6866666666666666,
-      "grad_norm": 1.1177273988723755,
-      "learning_rate": 1.1374415034337912e-06,
-      "loss": 1.4314,
-      "step": 30900
-    },
-    {
-      "epoch": 0.6888888888888889,
-      "grad_norm": 1.0766289234161377,
-      "learning_rate": 1.122694490500342e-06,
-      "loss": 1.4218,
-      "step": 31000
-    },
-    {
-      "epoch": 0.6888888888888889,
-      "eval_loss": 1.4482488632202148,
-      "eval_runtime": 104.3689,
-      "eval_samples_per_second": 131.217,
-      "eval_steps_per_second": 2.05,
-      "step": 31000
-    },
-    {
-      "epoch": 0.6911111111111111,
-      "grad_norm": 1.099563717842102,
-      "learning_rate": 1.1080159684575258e-06,
-      "loss": 1.4393,
-      "step": 31100
-    },
-    {
-      "epoch": 0.6933333333333334,
-      "grad_norm": 1.13261878490448,
-      "learning_rate": 1.0934066672414676e-06,
-      "loss": 1.4581,
-      "step": 31200
-    },
-    {
-      "epoch": 0.6955555555555556,
-      "grad_norm": 1.0697300434112549,
-      "learning_rate": 1.078867313346068e-06,
-      "loss": 1.4266,
-      "step": 31300
-    },
-    {
-      "epoch": 0.6977777777777778,
-      "grad_norm": 1.0712445974349976,
-      "learning_rate": 1.0643986297868766e-06,
-      "loss": 1.4267,
-      "step": 31400
-    },
-    {
-      "epoch": 0.7,
-      "grad_norm": 1.0139946937561035,
-      "learning_rate": 1.0500013360651318e-06,
-      "loss": 1.4228,
-      "step": 31500
-    },
-    {
-      "epoch": 0.7022222222222222,
-      "grad_norm": 1.0279037952423096,
-      "learning_rate": 1.0356761481319897e-06,
-      "loss": 1.4385,
-      "step": 31600
-    },
-    {
-      "epoch": 0.7044444444444444,
-      "grad_norm": 1.133392572402954,
-      "learning_rate": 1.021423778352915e-06,
-      "loss": 1.4099,
-      "step": 31700
-    },
-    {
-      "epoch": 0.7066666666666667,
-      "grad_norm": 1.0647848844528198,
-      "learning_rate": 1.0072449354722591e-06,
-      "loss": 1.4229,
-      "step": 31800
-    },
-    {
-      "epoch": 0.7088888888888889,
-      "grad_norm": 1.0599104166030884,
-      "learning_rate": 9.931403245780157e-07,
-      "loss": 1.4366,
-      "step": 31900
-    },
-    {
-      "epoch": 0.7111111111111111,
-      "grad_norm": 1.0552103519439697,
-      "learning_rate": 9.791106470667545e-07,
-      "loss": 1.4239,
-      "step": 32000
-    },
-    {
-      "epoch": 0.7111111111111111,
-      "eval_loss": 1.4449114799499512,
-      "eval_runtime": 104.4483,
-      "eval_samples_per_second": 131.118,
-      "eval_steps_per_second": 2.049,
-      "step": 32000
-    },
-    {
-      "epoch": 0.7133333333333334,
-      "grad_norm": 1.0780894756317139,
-      "learning_rate": 9.651566006087473e-07,
-      "loss": 1.4223,
-      "step": 32100
-    },
-    {
-      "epoch": 0.7155555555555555,
-      "grad_norm": 1.077383041381836,
-      "learning_rate": 9.512788791132715e-07,
-      "loss": 1.437,
-      "step": 32200
-    },
-    {
-      "epoch": 0.7177777777777777,
-      "grad_norm": 1.125698208808899,
-      "learning_rate": 9.37478172694101e-07,
-      "loss": 1.4324,
-      "step": 32300
-    },
-    {
-      "epoch": 0.72,
-      "grad_norm": 1.0730408430099487,
-      "learning_rate": 9.237551676351927e-07,
-      "loss": 1.4393,
-      "step": 32400
-    },
-    {
-      "epoch": 0.7222222222222222,
-      "grad_norm": 1.0632646083831787,
-      "learning_rate": 9.101105463565572e-07,
-      "loss": 1.446,
-      "step": 32500
-    },
-    {
-      "epoch": 0.7244444444444444,
-      "grad_norm": 1.049178957939148,
-      "learning_rate": 8.965449873803192e-07,
-      "loss": 1.4301,
-      "step": 32600
-    },
-    {
-      "epoch": 0.7266666666666667,
-      "grad_norm": 1.1717880964279175,
-      "learning_rate": 8.830591652969827e-07,
-      "loss": 1.4345,
-      "step": 32700
-    },
-    {
-      "epoch": 0.7288888888888889,
-      "grad_norm": 1.0648304224014282,
-      "learning_rate": 8.696537507318803e-07,
-      "loss": 1.4278,
-      "step": 32800
-    },
-    {
-      "epoch": 0.7311111111111112,
-      "grad_norm": 1.0418579578399658,
-      "learning_rate": 8.563294103118252e-07,
-      "loss": 1.4365,
-      "step": 32900
-    },
-    {
-      "epoch": 0.7333333333333333,
-      "grad_norm": 1.035050630569458,
-      "learning_rate": 8.430868066319625e-07,
-      "loss": 1.41,
-      "step": 33000
-    },
-    {
-      "epoch": 0.7333333333333333,
-      "eval_loss": 1.446353554725647,
-      "eval_runtime": 105.0303,
-      "eval_samples_per_second": 130.391,
-      "eval_steps_per_second": 2.038,
-      "step": 33000
-    },
-    {
-      "epoch": 0.7355555555555555,
-      "grad_norm": 1.071336030960083,
-      "learning_rate": 8.299265982228155e-07,
-      "loss": 1.439,
-      "step": 33100
-    },
-    {
-      "epoch": 0.7377777777777778,
-      "grad_norm": 1.0849003791809082,
-      "learning_rate": 8.168494395175436e-07,
-      "loss": 1.4172,
-      "step": 33200
-    },
-    {
-      "epoch": 0.74,
-      "grad_norm": 1.086184024810791,
-      "learning_rate": 8.038559808193958e-07,
-      "loss": 1.4232,
-      "step": 33300
-    },
-    {
-      "epoch": 0.7422222222222222,
-      "grad_norm": 1.0608620643615723,
-      "learning_rate": 7.909468682693699e-07,
-      "loss": 1.4154,
-      "step": 33400
-    },
-    {
-      "epoch": 0.7444444444444445,
-      "grad_norm": 1.0586633682250977,
-      "learning_rate": 7.781227438140868e-07,
-      "loss": 1.4354,
-      "step": 33500
-    },
-    {
-      "epoch": 0.7466666666666667,
-      "grad_norm": 1.0531774759292603,
-      "learning_rate": 7.65384245173863e-07,
-      "loss": 1.4394,
-      "step": 33600
-    },
-    {
-      "epoch": 0.7488888888888889,
-      "grad_norm": 1.134748935699463,
-      "learning_rate": 7.527320058109985e-07,
-      "loss": 1.4063,
-      "step": 33700
-    },
-    {
-      "epoch": 0.7511111111111111,
-      "grad_norm": 1.0640100240707397,
-      "learning_rate": 7.401666548982786e-07,
-      "loss": 1.4378,
-      "step": 33800
-    },
-    {
-      "epoch": 0.7533333333333333,
-      "grad_norm": 1.1123132705688477,
-      "learning_rate": 7.276888172876831e-07,
-      "loss": 1.4395,
-      "step": 33900
-    },
-    {
-      "epoch": 0.7555555555555555,
-      "grad_norm": 1.028074026107788,
-      "learning_rate": 7.152991134793158e-07,
-      "loss": 1.4091,
-      "step": 34000
-    },
-    {
-      "epoch": 0.7555555555555555,
-      "eval_loss": 1.4453147649765015,
-      "eval_runtime": 104.0869,
-      "eval_samples_per_second": 131.573,
-      "eval_steps_per_second": 2.056,
-      "step": 34000
-    },
-    {
-      "epoch": 0.7577777777777778,
-      "grad_norm": 1.0832644701004028,
-      "learning_rate": 7.029981595905477e-07,
-      "loss": 1.4245,
-      "step": 34100
-    },
-    {
-      "epoch": 0.76,
-      "grad_norm": 1.0776340961456299,
-      "learning_rate": 6.907865673253758e-07,
-      "loss": 1.4116,
-      "step": 34200
-    },
-    {
-      "epoch": 0.7622222222222222,
-      "grad_norm": 1.1135038137435913,
-      "learning_rate": 6.786649439440091e-07,
-      "loss": 1.4335,
-      "step": 34300
-    },
-    {
-      "epoch": 0.7644444444444445,
-      "grad_norm": 1.097848892211914,
-      "learning_rate": 6.666338922326676e-07,
-      "loss": 1.4422,
-      "step": 34400
-    },
-    {
-      "epoch": 0.7666666666666667,
-      "grad_norm": 1.1417701244354248,
-      "learning_rate": 6.546940104736058e-07,
-      "loss": 1.4445,
-      "step": 34500
-    },
-    {
-      "epoch": 0.7688888888888888,
-      "grad_norm": 1.0211708545684814,
-      "learning_rate": 6.428458924153644e-07,
-      "loss": 1.4341,
-      "step": 34600
-    },
-    {
-      "epoch": 0.7711111111111111,
-      "grad_norm": 1.049847960472107,
-      "learning_rate": 6.310901272432433e-07,
-      "loss": 1.4531,
-      "step": 34700
-    },
-    {
-      "epoch": 0.7733333333333333,
-      "grad_norm": 1.156039834022522,
-      "learning_rate": 6.194272995499986e-07,
-      "loss": 1.4254,
-      "step": 34800
-    },
-    {
-      "epoch": 0.7755555555555556,
-      "grad_norm": 1.1008870601654053,
-      "learning_rate": 6.078579893067787e-07,
-      "loss": 1.4354,
-      "step": 34900
-    },
-    {
-      "epoch": 0.7777777777777778,
-      "grad_norm": 1.064918041229248,
-      "learning_rate": 5.96382771834278e-07,
-      "loss": 1.4495,
-      "step": 35000
-    },
-    {
-      "epoch": 0.7777777777777778,
-      "eval_loss": 1.450618863105774,
-      "eval_runtime": 104.1774,
-      "eval_samples_per_second": 131.458,
-      "eval_steps_per_second": 2.054,
-      "step": 35000
     }
   ],
   "logging_steps": 100,
-  "max_steps": 45000,
+  "max_steps": 35000,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
   "save_steps": 5000,
@@ -2764,7 +1984,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.73851590868992e+18,
+  "total_flos": 3.3846542204928e+18,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null