diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14137 @@
+{
+  "best_metric": 1.6265630154840407,
+  "best_model_checkpoint": "train/20241111-Compress:256x-Lr:5e-5-Llama3-8B-instruct-GPT2-Large-RAG-no-ft_token-onlySquad-everymem/checkpoint-2000",
+  "epoch": 2.9482218536944904,
+  "eval_steps": 250,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014741109268472453,
+      "grad_norm": 10.542601492422415,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 4.9379,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029482218536944905,
+      "grad_norm": 14.75158889747135,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 5.1736,
+      "step": 2
+    },
+    {
+      "epoch": 0.004422332780541736,
+      "grad_norm": 20.766239292080247,
+      "learning_rate": 1.5e-06,
+      "loss": 4.988,
+      "step": 3
+    },
+    {
+      "epoch": 0.005896443707388981,
+      "grad_norm": 22.19482163527959,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 5.1685,
+      "step": 4
+    },
+    {
+      "epoch": 0.0073705546342362266,
+      "grad_norm": 11.413242986921624,
+      "learning_rate": 2.5e-06,
+      "loss": 5.2236,
+      "step": 5
+    },
+    {
+      "epoch": 0.008844665561083471,
+      "grad_norm": 31.813613722216566,
+      "learning_rate": 3e-06,
+      "loss": 5.0986,
+      "step": 6
+    },
+    {
+      "epoch": 0.010318776487930717,
+      "grad_norm": 11.019348121373584,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 5.3992,
+      "step": 7
+    },
+    {
+      "epoch": 0.011792887414777962,
+      "grad_norm": 13.5247891619426,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 5.0985,
+      "step": 8
+    },
+    {
+      "epoch": 0.013266998341625208,
+      "grad_norm": 14.064692754390002,
+      "learning_rate": 4.5e-06,
+      "loss": 5.1665,
+      "step": 9
+    },
+    {
+      "epoch": 0.014741109268472453,
+      "grad_norm": 19.06002641481889,
+      "learning_rate": 5e-06,
+      "loss": 4.9754,
+      "step": 10
+    },
+    {
+      "epoch": 0.016215220195319697,
+      "grad_norm": 28.46141572477718,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 4.9661,
+      "step": 11
+    },
+    {
+      "epoch": 0.017689331122166942,
+      "grad_norm": 14.223456574227844,
+      "learning_rate": 6e-06,
+      "loss": 5.0477,
+      "step": 12
+    },
+    {
+      "epoch": 0.019163442049014188,
+      "grad_norm": 10.541714955899053,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 4.9983,
+      "step": 13
+    },
+    {
+      "epoch": 0.020637552975861433,
+      "grad_norm": 10.403206961696505,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 4.538,
+      "step": 14
+    },
+    {
+      "epoch": 0.02211166390270868,
+      "grad_norm": 9.667787640349095,
+      "learning_rate": 7.5e-06,
+      "loss": 4.3433,
+      "step": 15
+    },
+    {
+      "epoch": 0.023585774829555924,
+      "grad_norm": 34.74159281731877,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 4.2695,
+      "step": 16
+    },
+    {
+      "epoch": 0.02505988575640317,
+      "grad_norm": 8.544322523623235,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 4.1162,
+      "step": 17
+    },
+    {
+      "epoch": 0.026533996683250415,
+      "grad_norm": 12.138609494258457,
+      "learning_rate": 9e-06,
+      "loss": 3.7167,
+      "step": 18
+    },
+    {
+      "epoch": 0.02800810761009766,
+      "grad_norm": 7.109522555989333,
+      "learning_rate": 9.5e-06,
+      "loss": 3.8958,
+      "step": 19
+    },
+    {
+      "epoch": 0.029482218536944906,
+      "grad_norm": 12.117353722264145,
+      "learning_rate": 1e-05,
+      "loss": 3.6133,
+      "step": 20
+    },
+    {
+      "epoch": 0.03095632946379215,
+      "grad_norm": 8.98839965901471,
+      "learning_rate": 1.05e-05,
+      "loss": 3.7009,
+      "step": 21
+    },
+    {
+      "epoch": 0.032430440390639394,
+      "grad_norm": 27.510298586535054,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 3.5887,
+      "step": 22
+    },
+    {
+      "epoch": 0.03390455131748664,
+      "grad_norm": 9.121228418438392,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 3.5404,
+      "step": 23
+    },
+    {
+      "epoch": 0.035378662244333885,
+      "grad_norm": 30.216555431881524,
+      "learning_rate": 1.2e-05,
+      "loss": 3.5466,
+      "step": 24
+    },
+    {
+      "epoch": 0.03685277317118113,
+      "grad_norm": 40.91363510016404,
+      "learning_rate": 1.25e-05,
+      "loss": 3.2227,
+      "step": 25
+    },
+    {
+      "epoch": 0.038326884098028376,
+      "grad_norm": 14.123070509116236,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 3.2812,
+      "step": 26
+    },
+    {
+      "epoch": 0.03980099502487562,
+      "grad_norm": 28.9019232403691,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 3.257,
+      "step": 27
+    },
+    {
+      "epoch": 0.04127510595172287,
+      "grad_norm": 8.806093741682783,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 3.2704,
+      "step": 28
+    },
+    {
+      "epoch": 0.04274921687857011,
+      "grad_norm": 10.635808495985684,
+      "learning_rate": 1.45e-05,
+      "loss": 3.1303,
+      "step": 29
+    },
+    {
+      "epoch": 0.04422332780541736,
+      "grad_norm": 12.7978741787882,
+      "learning_rate": 1.5e-05,
+      "loss": 2.8083,
+      "step": 30
+    },
+    {
+      "epoch": 0.0456974387322646,
+      "grad_norm": 7.577938242705195,
+      "learning_rate": 1.55e-05,
+      "loss": 2.7483,
+      "step": 31
+    },
+    {
+      "epoch": 0.04717154965911185,
+      "grad_norm": 5.686714537323279,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.8328,
+      "step": 32
+    },
+    {
+      "epoch": 0.048645660585959094,
+      "grad_norm": 5.251588853465192,
+      "learning_rate": 1.65e-05,
+      "loss": 2.4911,
+      "step": 33
+    },
+    {
+      "epoch": 0.05011977151280634,
+      "grad_norm": 5.799491518027747,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 2.6598,
+      "step": 34
+    },
+    {
+      "epoch": 0.051593882439653585,
+      "grad_norm": 49.994274574538785,
+      "learning_rate": 1.75e-05,
+      "loss": 2.5952,
+      "step": 35
+    },
+    {
+      "epoch": 0.05306799336650083,
+      "grad_norm": 3.706841386569642,
+      "learning_rate": 1.8e-05,
+      "loss": 2.3438,
+      "step": 36
+    },
+    {
+      "epoch": 0.054542104293348076,
+      "grad_norm": 3.9279312850784254,
+      "learning_rate": 1.85e-05,
+      "loss": 2.2941,
+      "step": 37
+    },
+    {
+      "epoch": 0.05601621522019532,
+      "grad_norm": 3.305465137801589,
+      "learning_rate": 1.9e-05,
+      "loss": 2.1947,
+      "step": 38
+    },
+    {
+      "epoch": 0.05749032614704257,
+      "grad_norm": 4.047149295421852,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 2.2379,
+      "step": 39
+    },
+    {
+      "epoch": 0.05896443707388981,
+      "grad_norm": 4.849153126192219,
+      "learning_rate": 2e-05,
+      "loss": 1.9444,
+      "step": 40
+    },
+    {
+      "epoch": 0.06043854800073706,
+      "grad_norm": 3.9283157290189785,
+      "learning_rate": 2.05e-05,
+      "loss": 1.9679,
+      "step": 41
+    },
+    {
+      "epoch": 0.0619126589275843,
+      "grad_norm": 3.55358453763254,
+      "learning_rate": 2.1e-05,
+      "loss": 2.0119,
+      "step": 42
+    },
+    {
+      "epoch": 0.06338676985443155,
+      "grad_norm": 3.1384102523002255,
+      "learning_rate": 2.15e-05,
+      "loss": 1.9532,
+      "step": 43
+    },
+    {
+      "epoch": 0.06486088078127879,
+      "grad_norm": 3.728170691671425,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.9814,
+      "step": 44
+    },
+    {
+      "epoch": 0.06633499170812604,
+      "grad_norm": 2.759380467583325,
+      "learning_rate": 2.25e-05,
+      "loss": 1.8839,
+      "step": 45
+    },
+    {
+      "epoch": 0.06780910263497328,
+      "grad_norm": 2.6363244786239415,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.8761,
+      "step": 46
+    },
+    {
+      "epoch": 0.06928321356182053,
+      "grad_norm": 2.371999752900238,
+      "learning_rate": 2.35e-05,
+      "loss": 1.8118,
+      "step": 47
+    },
+    {
+      "epoch": 0.07075732448866777,
+      "grad_norm": 3.5700949304177927,
+      "learning_rate": 2.4e-05,
+      "loss": 1.7577,
+      "step": 48
+    },
+    {
+      "epoch": 0.07223143541551502,
+      "grad_norm": 2.9866840157969983,
+      "learning_rate": 2.45e-05,
+      "loss": 1.9868,
+      "step": 49
+    },
+    {
+      "epoch": 0.07370554634236226,
+      "grad_norm": 2.7323796430573015,
+      "learning_rate": 2.5e-05,
+      "loss": 1.9494,
+      "step": 50
+    },
+    {
+      "epoch": 0.07517965726920951,
+      "grad_norm": 2.4284875979500526,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 1.871,
+      "step": 51
+    },
+    {
+      "epoch": 0.07665376819605675,
+      "grad_norm": 3.2890906253436727,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.8587,
+      "step": 52
+    },
+    {
+      "epoch": 0.078127879122904,
+      "grad_norm": 2.312546497599205,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 1.6874,
+      "step": 53
+    },
+    {
+      "epoch": 0.07960199004975124,
+      "grad_norm": 2.6813059554230123,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.904,
+      "step": 54
+    },
+    {
+      "epoch": 0.0810761009765985,
+      "grad_norm": 2.080791397660076,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 1.8092,
+      "step": 55
+    },
+    {
+      "epoch": 0.08255021190344573,
+      "grad_norm": 2.350351522411094,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.8377,
+      "step": 56
+    },
+    {
+      "epoch": 0.08402432283029299,
+      "grad_norm": 2.374558759157098,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.9329,
+      "step": 57
+    },
+    {
+      "epoch": 0.08549843375714022,
+      "grad_norm": 2.447574532049086,
+      "learning_rate": 2.9e-05,
+      "loss": 1.7067,
+      "step": 58
+    },
+    {
+      "epoch": 0.08697254468398748,
+      "grad_norm": 2.2704482729088618,
+      "learning_rate": 2.95e-05,
+      "loss": 1.6713,
+      "step": 59
+    },
+    {
+      "epoch": 0.08844665561083472,
+      "grad_norm": 2.0926121139115943,
+      "learning_rate": 3e-05,
+      "loss": 1.5644,
+      "step": 60
+    },
+    {
+      "epoch": 0.08992076653768195,
+      "grad_norm": 2.3963622020865016,
+      "learning_rate": 3.05e-05,
+      "loss": 1.8546,
+      "step": 61
+    },
+    {
+      "epoch": 0.0913948774645292,
+      "grad_norm": 2.196258124052599,
+      "learning_rate": 3.1e-05,
+      "loss": 1.841,
+      "step": 62
+    },
+    {
+      "epoch": 0.09286898839137644,
+      "grad_norm": 2.270958876566279,
+      "learning_rate": 3.15e-05,
+      "loss": 1.6501,
+      "step": 63
+    },
+    {
+      "epoch": 0.0943430993182237,
+      "grad_norm": 2.601987603541407,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.532,
+      "step": 64
+    },
+    {
+      "epoch": 0.09581721024507094,
+      "grad_norm": 2.068880327158985,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.7656,
+      "step": 65
+    },
+    {
+      "epoch": 0.09729132117191819,
+      "grad_norm": 2.3192701470932087,
+      "learning_rate": 3.3e-05,
+      "loss": 1.8471,
+      "step": 66
+    },
+    {
+      "epoch": 0.09876543209876543,
+      "grad_norm": 2.412193389541428,
+      "learning_rate": 3.35e-05,
+      "loss": 1.7321,
+      "step": 67
+    },
+    {
+      "epoch": 0.10023954302561268,
+      "grad_norm": 3.005819715767446,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.6546,
+      "step": 68
+    },
+    {
+      "epoch": 0.10171365395245992,
+      "grad_norm": 2.144864082565192,
+      "learning_rate": 3.45e-05,
+      "loss": 1.6561,
+      "step": 69
+    },
+    {
+      "epoch": 0.10318776487930717,
+      "grad_norm": 2.423850158542077,
+      "learning_rate": 3.5e-05,
+      "loss": 1.8283,
+      "step": 70
+    },
+    {
+      "epoch": 0.10466187580615441,
+      "grad_norm": 2.319801145778886,
+      "learning_rate": 3.55e-05,
+      "loss": 1.7837,
+      "step": 71
+    },
+    {
+      "epoch": 0.10613598673300166,
+      "grad_norm": 2.06578138166346,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7814,
+      "step": 72
+    },
+    {
+      "epoch": 0.1076100976598489,
+      "grad_norm": 1.9773213735314357,
+      "learning_rate": 3.65e-05,
+      "loss": 1.777,
+      "step": 73
+    },
+    {
+      "epoch": 0.10908420858669615,
+      "grad_norm": 1.8238764897198516,
+      "learning_rate": 3.7e-05,
+      "loss": 1.6219,
+      "step": 74
+    },
+    {
+      "epoch": 0.11055831951354339,
+      "grad_norm": 2.1520342318708696,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.7022,
+      "step": 75
+    },
+    {
+      "epoch": 0.11203243044039064,
+      "grad_norm": 1.9571210341882685,
+      "learning_rate": 3.8e-05,
+      "loss": 1.6719,
+      "step": 76
+    },
+    {
+      "epoch": 0.11350654136723788,
+      "grad_norm": 2.258248890873436,
+      "learning_rate": 3.85e-05,
+      "loss": 1.6456,
+      "step": 77
+    },
+    {
+      "epoch": 0.11498065229408513,
+      "grad_norm": 1.9365640040624643,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.5789,
+      "step": 78
+    },
+    {
+      "epoch": 0.11645476322093237,
+      "grad_norm": 2.4701209788483003,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 1.9057,
+      "step": 79
+    },
+    {
+      "epoch": 0.11792887414777962,
+      "grad_norm": 2.4384415593077433,
+      "learning_rate": 4e-05,
+      "loss": 1.6999,
+      "step": 80
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 2.487784966483082,
+      "learning_rate": 4.05e-05,
+      "loss": 1.8003,
+      "step": 81
+    },
+    {
+      "epoch": 0.12087709600147412,
+      "grad_norm": 2.184200086671485,
+      "learning_rate": 4.1e-05,
+      "loss": 1.7446,
+      "step": 82
+    },
+    {
+      "epoch": 0.12235120692832135,
+      "grad_norm": 2.1855229844329296,
+      "learning_rate": 4.15e-05,
+      "loss": 1.8108,
+      "step": 83
+    },
+    {
+      "epoch": 0.1238253178551686,
+      "grad_norm": 2.1906026362911493,
+      "learning_rate": 4.2e-05,
+      "loss": 1.6437,
+      "step": 84
+    },
+    {
+      "epoch": 0.12529942878201586,
+      "grad_norm": 2.3199766795860777,
+      "learning_rate": 4.25e-05,
+      "loss": 1.6714,
+      "step": 85
+    },
+    {
+      "epoch": 0.1267735397088631,
+      "grad_norm": 2.343977141499948,
+      "learning_rate": 4.3e-05,
+      "loss": 1.6882,
+      "step": 86
+    },
+    {
+      "epoch": 0.12824765063571034,
+      "grad_norm": 2.201401814430496,
+      "learning_rate": 4.35e-05,
+      "loss": 1.6077,
+      "step": 87
+    },
+    {
+      "epoch": 0.12972176156255757,
+      "grad_norm": 1.8836813540439896,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.4839,
+      "step": 88
+    },
+    {
+      "epoch": 0.13119587248940484,
+      "grad_norm": 2.071602243274418,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 1.6349,
+      "step": 89
+    },
+    {
+      "epoch": 0.13266998341625208,
+      "grad_norm": 2.178031812122008,
+      "learning_rate": 4.5e-05,
+      "loss": 1.508,
+      "step": 90
+    },
+    {
+      "epoch": 0.13414409434309932,
+      "grad_norm": 2.4650430009352697,
+      "learning_rate": 4.55e-05,
+      "loss": 1.5345,
+      "step": 91
+    },
+    {
+      "epoch": 0.13561820526994656,
+      "grad_norm": 2.2376626110126336,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.8138,
+      "step": 92
+    },
+    {
+      "epoch": 0.1370923161967938,
+      "grad_norm": 1.848529568956504,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 1.7583,
+      "step": 93
+    },
+    {
+      "epoch": 0.13856642712364106,
+      "grad_norm": 2.3004335616761007,
+      "learning_rate": 4.7e-05,
+      "loss": 1.6402,
+      "step": 94
+    },
+    {
+      "epoch": 0.1400405380504883,
+      "grad_norm": 2.048771918355873,
+      "learning_rate": 4.75e-05,
+      "loss": 1.5049,
+      "step": 95
+    },
+    {
+      "epoch": 0.14151464897733554,
+      "grad_norm": 2.212046853767831,
+      "learning_rate": 4.8e-05,
+      "loss": 1.709,
+      "step": 96
+    },
+    {
+      "epoch": 0.14298875990418278,
+      "grad_norm": 2.131450287924872,
+      "learning_rate": 4.85e-05,
+      "loss": 1.5567,
+      "step": 97
+    },
+    {
+      "epoch": 0.14446287083103004,
+      "grad_norm": 2.298086249237732,
+      "learning_rate": 4.9e-05,
+      "loss": 1.63,
+      "step": 98
+    },
+    {
+      "epoch": 0.14593698175787728,
+      "grad_norm": 2.1815599718613368,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 1.8168,
+      "step": 99
+    },
+    {
+      "epoch": 0.14741109268472452,
+      "grad_norm": 1.8790682843739575,
+      "learning_rate": 5e-05,
+      "loss": 1.6481,
+      "step": 100
+    },
+    {
+      "epoch": 0.14888520361157176,
+      "grad_norm": 1.8522093081747166,
+      "learning_rate": 4.999999216450553e-05,
+      "loss": 1.689,
+      "step": 101
+    },
+    {
+      "epoch": 0.15035931453841903,
+      "grad_norm": 2.536841445178404,
+      "learning_rate": 4.9999968658027006e-05,
+      "loss": 1.614,
+      "step": 102
+    },
+    {
+      "epoch": 0.15183342546526626,
+      "grad_norm": 2.3043916156729503,
+      "learning_rate": 4.999992948057919e-05,
+      "loss": 1.6039,
+      "step": 103
+    },
+    {
+      "epoch": 0.1533075363921135,
+      "grad_norm": 2.1947569228942054,
+      "learning_rate": 4.999987463218663e-05,
+      "loss": 1.6672,
+      "step": 104
+    },
+    {
+      "epoch": 0.15478164731896074,
+      "grad_norm": 2.1080649369677635,
+      "learning_rate": 4.9999804112883694e-05,
+      "loss": 1.6218,
+      "step": 105
+    },
+    {
+      "epoch": 0.156255758245808,
+      "grad_norm": 2.2353676678305248,
+      "learning_rate": 4.99997179227146e-05,
+      "loss": 1.6265,
+      "step": 106
+    },
+    {
+      "epoch": 0.15772986917265525,
+      "grad_norm": 2.1160965815101758,
+      "learning_rate": 4.999961606173337e-05,
+      "loss": 1.6612,
+      "step": 107
+    },
+    {
+      "epoch": 0.15920398009950248,
+      "grad_norm": 1.8458162787735273,
+      "learning_rate": 4.9999498530003866e-05,
+      "loss": 1.5433,
+      "step": 108
+    },
+    {
+      "epoch": 0.16067809102634972,
+      "grad_norm": 2.0363236664899396,
+      "learning_rate": 4.999936532759974e-05,
+      "loss": 1.7881,
+      "step": 109
+    },
+    {
+      "epoch": 0.162152201953197,
+      "grad_norm": 1.852527801366834,
+      "learning_rate": 4.9999216454604505e-05,
+      "loss": 1.5593,
+      "step": 110
+    },
+    {
+      "epoch": 0.16362631288004423,
+      "grad_norm": 2.230435078862805,
+      "learning_rate": 4.9999051911111484e-05,
+      "loss": 1.5841,
+      "step": 111
+    },
+    {
+      "epoch": 0.16510042380689147,
+      "grad_norm": 2.1978031026660187,
+      "learning_rate": 4.99988716972238e-05,
+      "loss": 1.5882,
+      "step": 112
+    },
+    {
+      "epoch": 0.1665745347337387,
+      "grad_norm": 1.9271774956319956,
+      "learning_rate": 4.999867581305444e-05,
+      "loss": 1.5627,
+      "step": 113
+    },
+    {
+      "epoch": 0.16804864566058597,
+      "grad_norm": 2.062486937510423,
+      "learning_rate": 4.9998464258726174e-05,
+      "loss": 1.7414,
+      "step": 114
+    },
+    {
+      "epoch": 0.1695227565874332,
+      "grad_norm": 1.8279360159980338,
+      "learning_rate": 4.999823703437162e-05,
+      "loss": 1.5979,
+      "step": 115
+    },
+    {
+      "epoch": 0.17099686751428045,
+      "grad_norm": 2.039134407967887,
+      "learning_rate": 4.999799414013322e-05,
+      "loss": 1.6692,
+      "step": 116
+    },
+    {
+      "epoch": 0.1724709784411277,
+      "grad_norm": 1.8610128674110982,
+      "learning_rate": 4.9997735576163215e-05,
+      "loss": 1.4458,
+      "step": 117
+    },
+    {
+      "epoch": 0.17394508936797495,
+      "grad_norm": 2.033024878113718,
+      "learning_rate": 4.9997461342623686e-05,
+      "loss": 1.5339,
+      "step": 118
+    },
+    {
+      "epoch": 0.1754192002948222,
+      "grad_norm": 2.2969587661104565,
+      "learning_rate": 4.999717143968654e-05,
+      "loss": 1.4748,
+      "step": 119
+    },
+    {
+      "epoch": 0.17689331122166943,
+      "grad_norm": 2.0766775496898267,
+      "learning_rate": 4.9996865867533496e-05,
+      "loss": 1.4835,
+      "step": 120
+    },
+    {
+      "epoch": 0.17836742214851667,
+      "grad_norm": 1.9920823611770595,
+      "learning_rate": 4.99965446263561e-05,
+      "loss": 1.5707,
+      "step": 121
+    },
+    {
+      "epoch": 0.1798415330753639,
+      "grad_norm": 1.8183865816598839,
+      "learning_rate": 4.9996207716355726e-05,
+      "loss": 1.5999,
+      "step": 122
+    },
+    {
+      "epoch": 0.18131564400221117,
+      "grad_norm": 2.044071753136424,
+      "learning_rate": 4.999585513774354e-05,
+      "loss": 1.7161,
+      "step": 123
+    },
+    {
+      "epoch": 0.1827897549290584,
+      "grad_norm": 1.9475531485360498,
+      "learning_rate": 4.9995486890740573e-05,
+      "loss": 1.5239,
+      "step": 124
+    },
+    {
+      "epoch": 0.18426386585590565,
+      "grad_norm": 1.8695379171798558,
+      "learning_rate": 4.9995102975577655e-05,
+      "loss": 1.496,
+      "step": 125
+    },
+    {
+      "epoch": 0.1857379767827529,
+      "grad_norm": 2.07468536933046,
+      "learning_rate": 4.999470339249543e-05,
+      "loss": 1.4883,
+      "step": 126
+    },
+    {
+      "epoch": 0.18721208770960016,
+      "grad_norm": 2.057945192732373,
+      "learning_rate": 4.9994288141744374e-05,
+      "loss": 1.4464,
+      "step": 127
+    },
+    {
+      "epoch": 0.1886861986364474,
+      "grad_norm": 2.2233408324512176,
+      "learning_rate": 4.999385722358479e-05,
+      "loss": 1.7203,
+      "step": 128
+    },
+    {
+      "epoch": 0.19016030956329463,
+      "grad_norm": 1.8766122879678175,
+      "learning_rate": 4.999341063828679e-05,
+      "loss": 1.5164,
+      "step": 129
+    },
+    {
+      "epoch": 0.19163442049014187,
+      "grad_norm": 2.0143590692499154,
+      "learning_rate": 4.9992948386130315e-05,
+      "loss": 1.5847,
+      "step": 130
+    },
+    {
+      "epoch": 0.19310853141698914,
+      "grad_norm": 2.2516484050259615,
+      "learning_rate": 4.9992470467405104e-05,
+      "loss": 1.5919,
+      "step": 131
+    },
+    {
+      "epoch": 0.19458264234383638,
+      "grad_norm": 1.8031131763830892,
+      "learning_rate": 4.999197688241076e-05,
+      "loss": 1.6906,
+      "step": 132
+    },
+    {
+      "epoch": 0.19605675327068361,
+      "grad_norm": 2.029461939306091,
+      "learning_rate": 4.999146763145668e-05,
+      "loss": 1.4806,
+      "step": 133
+    },
+    {
+      "epoch": 0.19753086419753085,
+      "grad_norm": 2.2177739750178236,
+      "learning_rate": 4.9990942714862066e-05,
+      "loss": 1.5786,
+      "step": 134
+    },
+    {
+      "epoch": 0.19900497512437812,
+      "grad_norm": 1.8300716110976656,
+      "learning_rate": 4.999040213295597e-05,
+      "loss": 1.4284,
+      "step": 135
+    },
+    {
+      "epoch": 0.20047908605122536,
+      "grad_norm": 2.1550993129566107,
+      "learning_rate": 4.9989845886077246e-05,
+      "loss": 1.6816,
+      "step": 136
+    },
+    {
+      "epoch": 0.2019531969780726,
+      "grad_norm": 1.8417871663069272,
+      "learning_rate": 4.9989273974574566e-05,
+      "loss": 1.3767,
+      "step": 137
+    },
+    {
+      "epoch": 0.20342730790491984,
+      "grad_norm": 2.192083189610869,
+      "learning_rate": 4.998868639880644e-05,
+      "loss": 1.6876,
+      "step": 138
+    },
+    {
+      "epoch": 0.2049014188317671,
+      "grad_norm": 2.3202672883326696,
+      "learning_rate": 4.998808315914117e-05,
+      "loss": 1.6607,
+      "step": 139
+    },
+    {
+      "epoch": 0.20637552975861434,
+      "grad_norm": 2.1634097440339533,
+      "learning_rate": 4.9987464255956894e-05,
+      "loss": 1.4282,
+      "step": 140
+    },
+    {
+      "epoch": 0.20784964068546158,
+      "grad_norm": 2.060724939544029,
+      "learning_rate": 4.9986829689641574e-05,
+      "loss": 1.4832,
+      "step": 141
+    },
+    {
+      "epoch": 0.20932375161230882,
+      "grad_norm": 2.0560524471658854,
+      "learning_rate": 4.998617946059297e-05,
+      "loss": 1.6022,
+      "step": 142
+    },
+    {
+      "epoch": 0.21079786253915608,
+      "grad_norm": 2.5000191687802715,
+      "learning_rate": 4.998551356921868e-05,
+      "loss": 1.7717,
+      "step": 143
+    },
+    {
+      "epoch": 0.21227197346600332,
+      "grad_norm": 2.3369231425139625,
+      "learning_rate": 4.99848320159361e-05,
+      "loss": 1.5764,
+      "step": 144
+    },
+    {
+      "epoch": 0.21374608439285056,
+      "grad_norm": 2.1573902377147407,
+      "learning_rate": 4.9984134801172464e-05,
+      "loss": 1.5507,
+      "step": 145
+    },
+    {
+      "epoch": 0.2152201953196978,
+      "grad_norm": 2.00797826179197,
+      "learning_rate": 4.998342192536482e-05,
+      "loss": 1.422,
+      "step": 146
+    },
+    {
+      "epoch": 0.21669430624654507,
+      "grad_norm": 2.3310063089017783,
+      "learning_rate": 4.998269338896e-05,
+      "loss": 1.6247,
+      "step": 147
+    },
+    {
+      "epoch": 0.2181684171733923,
+      "grad_norm": 1.9432428833058097,
+      "learning_rate": 4.998194919241471e-05,
+      "loss": 1.5054,
+      "step": 148
+    },
+    {
+      "epoch": 0.21964252810023954,
+      "grad_norm": 1.9152441620114702,
+      "learning_rate": 4.9981189336195425e-05,
+      "loss": 1.49,
+      "step": 149
+    },
+    {
+      "epoch": 0.22111663902708678,
+      "grad_norm": 2.026637547852652,
+      "learning_rate": 4.998041382077846e-05,
+      "loss": 1.6851,
+      "step": 150
+    },
+    {
+      "epoch": 0.22259074995393405,
+      "grad_norm": 1.7738780008145096,
+      "learning_rate": 4.9979622646649935e-05,
+      "loss": 1.4543,
+      "step": 151
+    },
+    {
+      "epoch": 0.22406486088078129,
+      "grad_norm": 2.0729515673384666,
+      "learning_rate": 4.997881581430579e-05,
+      "loss": 1.6819,
+      "step": 152
+    },
+    {
+      "epoch": 0.22553897180762852,
+      "grad_norm": 1.7773047138999618,
+      "learning_rate": 4.997799332425178e-05,
+      "loss": 1.6475,
+      "step": 153
+    },
+    {
+      "epoch": 0.22701308273447576,
+      "grad_norm": 1.5539139399810504,
+      "learning_rate": 4.997715517700347e-05,
+      "loss": 1.4423,
+      "step": 154
+    },
+    {
+      "epoch": 0.228487193661323,
+      "grad_norm": 1.920641123104727,
+      "learning_rate": 4.9976301373086254e-05,
+      "loss": 1.5398,
+      "step": 155
+    },
+    {
+      "epoch": 0.22996130458817027,
+      "grad_norm": 2.0395080547318667,
+      "learning_rate": 4.997543191303532e-05,
+      "loss": 1.6276,
+      "step": 156
+    },
+    {
+      "epoch": 0.2314354155150175,
+      "grad_norm": 2.092941925343976,
+      "learning_rate": 4.9974546797395685e-05,
+      "loss": 1.6567,
+      "step": 157
+    },
+    {
+      "epoch": 0.23290952644186474,
+      "grad_norm": 1.9710414206702205,
+      "learning_rate": 4.9973646026722166e-05,
+      "loss": 1.3883,
+      "step": 158
+    },
+    {
+      "epoch": 0.23438363736871198,
+      "grad_norm": 1.964366999096162,
+      "learning_rate": 4.997272960157942e-05,
+      "loss": 1.4694,
+      "step": 159
+    },
+    {
+      "epoch": 0.23585774829555925,
+      "grad_norm": 2.0563750208731255,
+      "learning_rate": 4.997179752254188e-05,
+      "loss": 1.5317,
+      "step": 160
+    },
+    {
+      "epoch": 0.2373318592224065,
+      "grad_norm": 2.0723551336259245,
+      "learning_rate": 4.997084979019382e-05,
+      "loss": 1.6037,
+      "step": 161
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 1.9570127322601956,
+      "learning_rate": 4.996988640512931e-05,
+      "loss": 1.5421,
+      "step": 162
+    },
+    {
+      "epoch": 0.24028008107610097,
+      "grad_norm": 1.898225309342246,
+      "learning_rate": 4.9968907367952245e-05,
+      "loss": 1.6293,
+      "step": 163
+    },
+    {
+      "epoch": 0.24175419200294823,
+      "grad_norm": 1.909058362613203,
+      "learning_rate": 4.9967912679276316e-05,
+      "loss": 1.5222,
+      "step": 164
+    },
+    {
+      "epoch": 0.24322830292979547,
+      "grad_norm": 2.51349298393404,
+      "learning_rate": 4.996690233972505e-05,
+      "loss": 1.6348,
+      "step": 165
+    },
+    {
+      "epoch": 0.2447024138566427,
+      "grad_norm": 2.0716379205494966,
+      "learning_rate": 4.996587634993175e-05,
+      "loss": 1.5544,
+      "step": 166
+    },
+    {
+      "epoch": 0.24617652478348995,
+      "grad_norm": 2.0953407507577113,
+      "learning_rate": 4.996483471053955e-05,
+      "loss": 1.4252,
+      "step": 167
+    },
+    {
+      "epoch": 0.2476506357103372,
+      "grad_norm": 1.858270782079412,
+      "learning_rate": 4.996377742220139e-05,
+      "loss": 1.578,
+      "step": 168
+    },
+    {
+      "epoch": 0.24912474663718445,
+      "grad_norm": 2.348591597546374,
+      "learning_rate": 4.9962704485580034e-05,
+      "loss": 1.5462,
+      "step": 169
+    },
+    {
+      "epoch": 0.2505988575640317,
+      "grad_norm": 1.7413280745047648,
+      "learning_rate": 4.996161590134802e-05,
+      "loss": 1.3747,
+      "step": 170
+    },
+    {
+      "epoch": 0.25207296849087896,
+      "grad_norm": 1.9123069964988997,
+      "learning_rate": 4.996051167018773e-05,
+      "loss": 1.5259,
+      "step": 171
+    },
+    {
+      "epoch": 0.2535470794177262,
+      "grad_norm": 1.9249726429766465,
+      "learning_rate": 4.995939179279134e-05,
+      "loss": 1.4706,
+      "step": 172
+    },
+    {
+      "epoch": 0.25502119034457343,
+      "grad_norm": 1.8405394969827769,
+      "learning_rate": 4.9958256269860826e-05,
+      "loss": 1.7181,
+      "step": 173
+    },
+    {
+      "epoch": 0.2564953012714207,
+      "grad_norm": 2.1502327371646563,
+      "learning_rate": 4.995710510210798e-05,
+      "loss": 1.6783,
+      "step": 174
+    },
+    {
+      "epoch": 0.2579694121982679,
+      "grad_norm": 1.8132622530039457,
+      "learning_rate": 4.9955938290254404e-05,
+      "loss": 1.6482,
+      "step": 175
+    },
+    {
+      "epoch": 0.25944352312511515,
+      "grad_norm": 1.8339603897817565,
+      "learning_rate": 4.99547558350315e-05,
+      "loss": 1.5723,
+      "step": 176
+    },
+    {
+      "epoch": 0.2609176340519624,
+      "grad_norm": 1.804856858006492,
+      "learning_rate": 4.9953557737180477e-05,
+      "loss": 1.4287,
+      "step": 177
+    },
+    {
+      "epoch": 0.2623917449788097,
+      "grad_norm": 1.8453585186936599,
+      "learning_rate": 4.9952343997452355e-05,
+      "loss": 1.4236,
+      "step": 178
+    },
+    {
+      "epoch": 0.2638658559056569,
+      "grad_norm": 2.002195583640961,
+      "learning_rate": 4.995111461660794e-05,
+      "loss": 1.5233,
+      "step": 179
+    },
+    {
+      "epoch": 0.26533996683250416,
+      "grad_norm": 1.8855827497250792,
+      "learning_rate": 4.9949869595417876e-05,
+      "loss": 1.6272,
+      "step": 180
+    },
+    {
+      "epoch": 0.2668140777593514,
+      "grad_norm": 2.0602540011427846,
+      "learning_rate": 4.994860893466258e-05,
+      "loss": 1.4663,
+      "step": 181
+    },
+    {
+      "epoch": 0.26828818868619864,
+      "grad_norm": 1.6948356836843923,
+      "learning_rate": 4.994733263513228e-05,
+      "loss": 1.55,
+      "step": 182
+    },
+    {
+      "epoch": 0.2697622996130459,
+      "grad_norm": 1.83476674512537,
+      "learning_rate": 4.994604069762702e-05,
+      "loss": 1.6474,
+      "step": 183
+    },
+    {
+      "epoch": 0.2712364105398931,
+      "grad_norm": 1.894212644863446,
+      "learning_rate": 4.994473312295663e-05,
+      "loss": 1.667,
+      "step": 184
+    },
+    {
+      "epoch": 0.27271052146674035,
+      "grad_norm": 1.9069100784881219,
+      "learning_rate": 4.994340991194076e-05,
+      "loss": 1.4898,
+      "step": 185
+    },
+    {
+      "epoch": 0.2741846323935876,
+      "grad_norm": 1.8500035801414696,
+      "learning_rate": 4.994207106540884e-05,
+      "loss": 1.7226,
+      "step": 186
+    },
+    {
+      "epoch": 0.2756587433204349,
+      "grad_norm": 2.0568917058720078,
+      "learning_rate": 4.994071658420012e-05,
+      "loss": 1.6814,
+      "step": 187
+    },
+    {
+      "epoch": 0.2771328542472821,
+      "grad_norm": 1.737312166777921,
+      "learning_rate": 4.993934646916364e-05,
+      "loss": 1.6241,
+      "step": 188
+    },
+    {
+      "epoch": 0.27860696517412936,
+      "grad_norm": 1.8974867586822286,
+      "learning_rate": 4.993796072115824e-05,
+      "loss": 1.5502,
+      "step": 189
+    },
+    {
+      "epoch": 0.2800810761009766,
+      "grad_norm": 1.627286109874131,
+      "learning_rate": 4.993655934105256e-05,
+      "loss": 1.512,
+      "step": 190
+    },
+    {
+      "epoch": 0.28155518702782384,
+      "grad_norm": 1.9361902547623564,
+      "learning_rate": 4.993514232972504e-05,
+      "loss": 1.5987,
+      "step": 191
+    },
+    {
+      "epoch": 0.2830292979546711,
+      "grad_norm": 1.8497831913485774,
+      "learning_rate": 4.9933709688063935e-05,
+      "loss": 1.5047,
+      "step": 192
+    },
+    {
+      "epoch": 0.2845034088815183,
+      "grad_norm": 1.9299011112237805,
+      "learning_rate": 4.993226141696726e-05,
+      "loss": 1.4832,
+      "step": 193
+    },
+    {
+      "epoch": 0.28597751980836555,
+      "grad_norm": 1.8794470502453384,
+      "learning_rate": 4.9930797517342853e-05,
+      "loss": 1.5571,
+      "step": 194
+    },
+    {
+      "epoch": 0.28745163073521285,
+      "grad_norm": 1.9405317274889466,
+      "learning_rate": 4.992931799010836e-05,
+      "loss": 1.4643,
+      "step": 195
+    },
+    {
+      "epoch": 0.2889257416620601,
+      "grad_norm": 1.6917059018450584,
+      "learning_rate": 4.992782283619118e-05,
+      "loss": 1.5118,
+      "step": 196
+    },
+    {
+      "epoch": 0.2903998525889073,
+      "grad_norm": 1.7850001828276383,
+      "learning_rate": 4.992631205652857e-05,
+      "loss": 1.6505,
+      "step": 197
+    },
+    {
+      "epoch": 0.29187396351575456,
+      "grad_norm": 2.048264708749373,
+      "learning_rate": 4.992478565206752e-05,
+      "loss": 1.5746,
+      "step": 198
+    },
+    {
+      "epoch": 0.2933480744426018,
+      "grad_norm": 1.8559218967627773,
+      "learning_rate": 4.992324362376484e-05,
+      "loss": 1.7085,
+      "step": 199
+    },
+    {
+      "epoch": 0.29482218536944904,
+      "grad_norm": 1.988831449575493,
+      "learning_rate": 4.992168597258715e-05,
+      "loss": 1.5021,
+      "step": 200
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 1.6964017120185386,
+      "learning_rate": 4.992011269951083e-05,
+      "loss": 1.4622,
+      "step": 201
+    },
+    {
+      "epoch": 0.2977704072231435,
+      "grad_norm": 1.7264181223438744,
+      "learning_rate": 4.991852380552209e-05,
+      "loss": 1.443,
+      "step": 202
+    },
+    {
+      "epoch": 0.2992445181499908,
+      "grad_norm": 1.5860528763017867,
+      "learning_rate": 4.99169192916169e-05,
+      "loss": 1.5066,
+      "step": 203
+    },
+    {
+      "epoch": 0.30071862907683805,
+      "grad_norm": 1.6375511802071567,
+      "learning_rate": 4.991529915880103e-05,
+      "loss": 1.3621,
+      "step": 204
+    },
+    {
+      "epoch": 0.3021927400036853,
+      "grad_norm": 1.620046622111707,
+      "learning_rate": 4.991366340809005e-05,
+      "loss": 1.7273,
+      "step": 205
+    },
+    {
+      "epoch": 0.30366685093053253,
+      "grad_norm": 1.7310077180960322,
+      "learning_rate": 4.99120120405093e-05,
+      "loss": 1.6215,
+      "step": 206
+    },
+    {
+      "epoch": 0.30514096185737977,
+      "grad_norm": 1.7646692440493075,
+      "learning_rate": 4.9910345057093936e-05,
+      "loss": 1.4291,
+      "step": 207
+    },
+    {
+      "epoch": 0.306615072784227,
+      "grad_norm": 1.7385663773792093,
+      "learning_rate": 4.990866245888889e-05,
+      "loss": 1.6184,
+      "step": 208
+    },
+    {
+      "epoch": 0.30808918371107424,
+      "grad_norm": 1.881623965382252,
+      "learning_rate": 4.9906964246948874e-05,
+      "loss": 1.4605,
+      "step": 209
+    },
+    {
+      "epoch": 0.3095632946379215,
+      "grad_norm": 1.6654262774873103,
+      "learning_rate": 4.99052504223384e-05,
+      "loss": 1.5176,
+      "step": 210
+    },
+    {
+      "epoch": 0.3110374055647688,
+      "grad_norm": 1.6322070737859578,
+      "learning_rate": 4.990352098613176e-05,
+      "loss": 1.5362,
+      "step": 211
+    },
+    {
+      "epoch": 0.312511516491616,
+      "grad_norm": 1.5849406299242006,
+      "learning_rate": 4.9901775939413026e-05,
+      "loss": 1.4416,
+      "step": 212
+    },
+    {
+      "epoch": 0.31398562741846325,
+      "grad_norm": 2.052211763910315,
+      "learning_rate": 4.990001528327607e-05,
+      "loss": 1.49,
+      "step": 213
+    },
+    {
+      "epoch": 0.3154597383453105,
+      "grad_norm": 1.7744427491274635,
+      "learning_rate": 4.989823901882454e-05,
+      "loss": 1.3264,
+      "step": 214
+    },
+    {
+      "epoch": 0.31693384927215773,
+      "grad_norm": 2.0492089629439487,
+      "learning_rate": 4.989644714717187e-05,
+      "loss": 1.515,
+      "step": 215
+    },
+    {
+      "epoch": 0.31840796019900497,
+      "grad_norm": 1.8718381925858396,
+      "learning_rate": 4.989463966944127e-05,
+      "loss": 1.5779,
+      "step": 216
+    },
+    {
+      "epoch": 0.3198820711258522,
+      "grad_norm": 1.7506940691450417,
+      "learning_rate": 4.989281658676573e-05,
+      "loss": 1.5214,
+      "step": 217
+    },
+    {
+      "epoch": 0.32135618205269945,
+      "grad_norm": 1.9412881673289646,
+      "learning_rate": 4.989097790028806e-05,
+      "loss": 1.5771,
+      "step": 218
+    },
+    {
+      "epoch": 0.3228302929795467,
+      "grad_norm": 1.8586373669195384,
+      "learning_rate": 4.98891236111608e-05,
+      "loss": 1.595,
+      "step": 219
+    },
+    {
+      "epoch": 0.324304403906394,
+      "grad_norm": 1.9537625302274773,
+      "learning_rate": 4.988725372054629e-05,
+      "loss": 1.5517,
+      "step": 220
+    },
+    {
+      "epoch": 0.3257785148332412,
+      "grad_norm": 1.7472069795055158,
+      "learning_rate": 4.988536822961666e-05,
+      "loss": 1.537,
+      "step": 221
+    },
+    {
+      "epoch": 0.32725262576008846,
+      "grad_norm": 1.9921702664695777,
+      "learning_rate": 4.988346713955381e-05,
+      "loss": 1.5504,
+      "step": 222
+    },
+    {
+      "epoch": 0.3287267366869357,
+      "grad_norm": 2.0329267183860518,
+      "learning_rate": 4.9881550451549405e-05,
+      "loss": 1.6402,
+      "step": 223
+    },
+    {
+      "epoch": 0.33020084761378293,
+      "grad_norm": 1.9075573674845931,
+      "learning_rate": 4.987961816680492e-05,
+      "loss": 1.7153,
+      "step": 224
+    },
+    {
+      "epoch": 0.33167495854063017,
+      "grad_norm": 2.169518598959896,
+      "learning_rate": 4.9877670286531585e-05,
+      "loss": 1.6808,
+      "step": 225
+    },
+    {
+      "epoch": 0.3331490694674774,
+      "grad_norm": 1.6680833120557446,
+      "learning_rate": 4.98757068119504e-05,
+      "loss": 1.5796,
+      "step": 226
+    },
+    {
+      "epoch": 0.33462318039432465,
+      "grad_norm": 1.5567522762754262,
+      "learning_rate": 4.9873727744292144e-05,
+      "loss": 1.6217,
+      "step": 227
+    },
+    {
+      "epoch": 0.33609729132117194,
+      "grad_norm": 2.3513075985056755,
+      "learning_rate": 4.987173308479738e-05,
+      "loss": 1.7312,
+      "step": 228
+    },
+    {
+      "epoch": 0.3375714022480192,
+      "grad_norm": 1.7241615413824978,
+      "learning_rate": 4.9869722834716446e-05,
+      "loss": 1.4205,
+      "step": 229
+    },
+    {
+      "epoch": 0.3390455131748664,
+      "grad_norm": 1.8826174813096306,
+      "learning_rate": 4.9867696995309445e-05,
+      "loss": 1.3508,
+      "step": 230
+    },
+    {
+      "epoch": 0.34051962410171366,
+      "grad_norm": 1.9308716387921656,
+      "learning_rate": 4.986565556784625e-05,
+      "loss": 1.6173,
+      "step": 231
+    },
+    {
+      "epoch": 0.3419937350285609,
+      "grad_norm": 1.879170421024246,
+      "learning_rate": 4.98635985536065e-05,
+      "loss": 1.6777,
+      "step": 232
+    },
+    {
+      "epoch": 0.34346784595540814,
+      "grad_norm": 1.821979790580715,
+      "learning_rate": 4.986152595387963e-05,
+      "loss": 1.4362,
+      "step": 233
+    },
+    {
+      "epoch": 0.3449419568822554,
+      "grad_norm": 1.9441473355444765,
+      "learning_rate": 4.9859437769964815e-05,
+      "loss": 1.6138,
+      "step": 234
+    },
+    {
+      "epoch": 0.3464160678091026,
+      "grad_norm": 1.8543749988685765,
+      "learning_rate": 4.985733400317101e-05,
+      "loss": 1.5919,
+      "step": 235
+    },
+    {
+      "epoch": 0.3478901787359499,
+      "grad_norm": 1.7113855202111685,
+      "learning_rate": 4.985521465481695e-05,
+      "loss": 1.5299,
+      "step": 236
+    },
+    {
+      "epoch": 0.34936428966279715,
+      "grad_norm": 1.9334724157006715,
+      "learning_rate": 4.985307972623112e-05,
+      "loss": 1.5473,
+      "step": 237
+    },
+    {
+      "epoch": 0.3508384005896444,
+      "grad_norm": 1.6254158221590336,
+      "learning_rate": 4.985092921875178e-05,
+      "loss": 1.5182,
+      "step": 238
+    },
+    {
+      "epoch": 0.3523125115164916,
+      "grad_norm": 1.7504192939901426,
+      "learning_rate": 4.984876313372695e-05,
+      "loss": 1.5105,
+      "step": 239
+    },
+    {
+      "epoch": 0.35378662244333886,
+      "grad_norm": 1.8750132242372193,
+      "learning_rate": 4.984658147251442e-05,
+      "loss": 1.5587,
+      "step": 240
+    },
+    {
+      "epoch": 0.3552607333701861,
+      "grad_norm": 1.950406066478905,
+      "learning_rate": 4.984438423648174e-05,
+      "loss": 1.4449,
+      "step": 241
+    },
+    {
+      "epoch": 0.35673484429703334,
+      "grad_norm": 1.9475823453797563,
+      "learning_rate": 4.9842171427006225e-05,
+      "loss": 1.5364,
+      "step": 242
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 1.6947296124965352,
+      "learning_rate": 4.983994304547495e-05,
+      "loss": 1.5385,
+      "step": 243
+    },
+    {
+      "epoch": 0.3596830661507278,
+      "grad_norm": 1.83443633046863,
+      "learning_rate": 4.9837699093284765e-05,
+      "loss": 1.4704,
+      "step": 244
+    },
+    {
+      "epoch": 0.3611571770775751,
+      "grad_norm": 1.9536703340246153,
+      "learning_rate": 4.983543957184224e-05,
+      "loss": 1.6683,
+      "step": 245
+    },
+    {
+      "epoch": 0.36263128800442235,
+      "grad_norm": 1.8570997277690837,
+      "learning_rate": 4.983316448256377e-05,
+      "loss": 1.2434,
+      "step": 246
+    },
+    {
+      "epoch": 0.3641053989312696,
+      "grad_norm": 1.8087943771180426,
+      "learning_rate": 4.983087382687544e-05,
+      "loss": 1.649,
+      "step": 247
+    },
+    {
+      "epoch": 0.3655795098581168,
+      "grad_norm": 1.9300428673326977,
+      "learning_rate": 4.982856760621313e-05,
+      "loss": 1.4011,
+      "step": 248
+    },
+    {
+      "epoch": 0.36705362078496406,
+      "grad_norm": 1.7496896877367119,
+      "learning_rate": 4.9826245822022474e-05,
+      "loss": 1.5923,
+      "step": 249
+    },
+    {
+      "epoch": 0.3685277317118113,
+      "grad_norm": 1.799511244433208,
+      "learning_rate": 4.9823908475758875e-05,
+      "loss": 1.6522,
+      "step": 250
+    },
+    {
+      "epoch": 0.3685277317118113,
+      "eval_bleu": 0.04824338198619645,
+      "eval_bleu_1gram": 0.34108276434595103,
+      "eval_bleu_2gram": 0.11484718043335776,
+      "eval_bleu_3gram": 0.04694283612862056,
+      "eval_bleu_4gram": 0.022946329938694693,
+      "eval_rag_val_loss": 1.5737228063485955,
+      "eval_rouge1": 0.32701224706717685,
+      "eval_rouge2": 0.10811216474930627,
+      "eval_rougeL": 0.3077781137710209,
+      "step": 250
+    },
+    {
+      "epoch": 0.37000184263865854,
+      "grad_norm": 1.6572600829421518,
+      "learning_rate": 4.982155556888745e-05,
+      "loss": 1.5452,
+      "step": 251
+    },
+    {
+      "epoch": 0.3714759535655058,
+      "grad_norm": 1.930245940296186,
+      "learning_rate": 4.981918710288309e-05,
+      "loss": 1.4397,
+      "step": 252
+    },
+    {
+      "epoch": 0.3729500644923531,
+      "grad_norm": 1.938358977856933,
+      "learning_rate": 4.981680307923047e-05,
+      "loss": 1.6566,
+      "step": 253
+    },
+    {
+      "epoch": 0.3744241754192003,
+      "grad_norm": 1.8774043879437272,
+      "learning_rate": 4.981440349942397e-05,
+      "loss": 1.6621,
+      "step": 254
+    },
+    {
+      "epoch": 0.37589828634604755,
+      "grad_norm": 1.861090886146892,
+      "learning_rate": 4.981198836496775e-05,
+      "loss": 1.551,
+      "step": 255
+    },
+    {
+      "epoch": 0.3773723972728948,
+      "grad_norm": 1.802861334234698,
+      "learning_rate": 4.9809557677375704e-05,
+      "loss": 1.5642,
+      "step": 256
+    },
+    {
+      "epoch": 0.378846508199742,
+      "grad_norm": 2.2210855146396966,
+      "learning_rate": 4.98071114381715e-05,
+      "loss": 1.5815,
+      "step": 257
+    },
+    {
+      "epoch": 0.38032061912658927,
+      "grad_norm": 1.818147739177719,
+      "learning_rate": 4.980464964888852e-05,
+      "loss": 1.5253,
+      "step": 258
+    },
+    {
+      "epoch": 0.3817947300534365,
+      "grad_norm": 1.9086978392103475,
+      "learning_rate": 4.980217231106991e-05,
+      "loss": 1.6457,
+      "step": 259
+    },
+    {
+      "epoch": 0.38326884098028374,
+      "grad_norm": 1.8975392166930145,
+      "learning_rate": 4.979967942626858e-05,
+      "loss": 1.5502,
+      "step": 260
+    },
+    {
+      "epoch": 0.38474295190713104,
+      "grad_norm": 2.20955374721703,
+      "learning_rate": 4.979717099604715e-05,
+      "loss": 1.597,
+      "step": 261
+    },
+    {
+      "epoch": 0.3862170628339783,
+      "grad_norm": 1.8116599800082387,
+      "learning_rate": 4.979464702197801e-05,
+      "loss": 1.6264,
+      "step": 262
+    },
+    {
+      "epoch": 0.3876911737608255,
+      "grad_norm": 2.0297578466146042,
+      "learning_rate": 4.9792107505643304e-05,
+      "loss": 1.5653,
+      "step": 263
+    },
+    {
+      "epoch": 0.38916528468767275,
+      "grad_norm": 1.947739768198093,
+      "learning_rate": 4.9789552448634874e-05,
+      "loss": 1.6046,
+      "step": 264
+    },
+    {
+      "epoch": 0.39063939561452,
+      "grad_norm": 1.7200403398396384,
+      "learning_rate": 4.9786981852554346e-05,
+      "loss": 1.4561,
+      "step": 265
+    },
+    {
+      "epoch": 0.39211350654136723,
+      "grad_norm": 1.9104408099008892,
+      "learning_rate": 4.978439571901307e-05,
+      "loss": 1.7935,
+      "step": 266
+    },
+    {
+      "epoch": 0.39358761746821447,
+      "grad_norm": 1.7448903829253266,
+      "learning_rate": 4.9781794049632135e-05,
+      "loss": 1.537,
+      "step": 267
+    },
+    {
+      "epoch": 0.3950617283950617,
+      "grad_norm": 1.6825972792233093,
+      "learning_rate": 4.9779176846042366e-05,
+      "loss": 1.4845,
+      "step": 268
+    },
+    {
+      "epoch": 0.396535839321909,
+      "grad_norm": 1.8188593880703856,
+      "learning_rate": 4.977654410988434e-05,
+      "loss": 1.4426,
+      "step": 269
+    },
+    {
+      "epoch": 0.39800995024875624,
+      "grad_norm": 1.6547712526387357,
+      "learning_rate": 4.977389584280835e-05,
+      "loss": 1.5164,
+      "step": 270
+    },
+    {
+      "epoch": 0.3994840611756035,
+      "grad_norm": 1.6810958281099189,
+      "learning_rate": 4.9771232046474444e-05,
+      "loss": 1.448,
+      "step": 271
+    },
+    {
+      "epoch": 0.4009581721024507,
+      "grad_norm": 1.8850363739511167,
+      "learning_rate": 4.976855272255239e-05,
+      "loss": 1.5656,
+      "step": 272
+    },
+    {
+      "epoch": 0.40243228302929795,
+      "grad_norm": 1.9444085284352308,
+      "learning_rate": 4.976585787272168e-05,
+      "loss": 1.7654,
+      "step": 273
+    },
+    {
+      "epoch": 0.4039063939561452,
+      "grad_norm": 1.8866578402273446,
+      "learning_rate": 4.976314749867158e-05,
+      "loss": 1.6431,
+      "step": 274
+    },
+    {
+      "epoch": 0.40538050488299243,
+      "grad_norm": 1.652201833559983,
+      "learning_rate": 4.976042160210104e-05,
+      "loss": 1.5283,
+      "step": 275
+    },
+    {
+      "epoch": 0.40685461580983967,
+      "grad_norm": 1.7229485296265201,
+      "learning_rate": 4.975768018471877e-05,
+      "loss": 1.4934,
+      "step": 276
+    },
+    {
+      "epoch": 0.4083287267366869,
+      "grad_norm": 1.8622273597605954,
+      "learning_rate": 4.9754923248243195e-05,
+      "loss": 1.6194,
+      "step": 277
+    },
+    {
+      "epoch": 0.4098028376635342,
+      "grad_norm": 1.9324903634486195,
+      "learning_rate": 4.975215079440247e-05,
+      "loss": 1.8159,
+      "step": 278
+    },
+    {
+      "epoch": 0.41127694859038144,
+      "grad_norm": 1.9146934870543342,
+      "learning_rate": 4.974936282493448e-05,
+      "loss": 1.5907,
+      "step": 279
+    },
+    {
+      "epoch": 0.4127510595172287,
+      "grad_norm": 1.7951085571573018,
+      "learning_rate": 4.974655934158684e-05,
+      "loss": 1.5056,
+      "step": 280
+    },
+    {
+      "epoch": 0.4142251704440759,
+      "grad_norm": 1.8439588913338272,
+      "learning_rate": 4.974374034611687e-05,
+      "loss": 1.541,
+      "step": 281
+    },
+    {
+      "epoch": 0.41569928137092316,
+      "grad_norm": 1.7446759161819978,
+      "learning_rate": 4.9740905840291646e-05,
+      "loss": 1.5647,
+      "step": 282
+    },
+    {
+      "epoch": 0.4171733922977704,
+      "grad_norm": 1.893249266743817,
+      "learning_rate": 4.9738055825887936e-05,
+      "loss": 1.6144,
+      "step": 283
+    },
+    {
+      "epoch": 0.41864750322461763,
+      "grad_norm": 1.7681815618999355,
+      "learning_rate": 4.973519030469225e-05,
+      "loss": 1.4294,
+      "step": 284
+    },
+    {
+      "epoch": 0.4201216141514649,
+      "grad_norm": 2.0851913432044213,
+      "learning_rate": 4.97323092785008e-05,
+      "loss": 1.4536,
+      "step": 285
+    },
+    {
+      "epoch": 0.42159572507831217,
+      "grad_norm": 1.984376922366195,
+      "learning_rate": 4.972941274911953e-05,
+      "loss": 1.6432,
+      "step": 286
+    },
+    {
+      "epoch": 0.4230698360051594,
+      "grad_norm": 1.9614436495638687,
+      "learning_rate": 4.97265007183641e-05,
+      "loss": 1.3757,
+      "step": 287
+    },
+    {
+      "epoch": 0.42454394693200664,
+      "grad_norm": 1.7788729282989244,
+      "learning_rate": 4.9723573188059894e-05,
+      "loss": 1.4745,
+      "step": 288
+    },
+    {
+      "epoch": 0.4260180578588539,
+      "grad_norm": 1.825779706491545,
+      "learning_rate": 4.972063016004199e-05,
+      "loss": 1.4818,
+      "step": 289
+    },
+    {
+      "epoch": 0.4274921687857011,
+      "grad_norm": 1.8536216134738772,
+      "learning_rate": 4.971767163615522e-05,
+      "loss": 1.5649,
+      "step": 290
+    },
+    {
+      "epoch": 0.42896627971254836,
+      "grad_norm": 1.70620204592658,
+      "learning_rate": 4.971469761825407e-05,
+      "loss": 1.7124,
+      "step": 291
+    },
+    {
+      "epoch": 0.4304403906393956,
+      "grad_norm": 1.9209948129142063,
+      "learning_rate": 4.971170810820279e-05,
+      "loss": 1.6574,
+      "step": 292
+    },
+    {
+      "epoch": 0.43191450156624284,
+      "grad_norm": 1.8584657898413137,
+      "learning_rate": 4.970870310787532e-05,
+      "loss": 1.5529,
+      "step": 293
+    },
+    {
+      "epoch": 0.43338861249309013,
+      "grad_norm": 1.9514525920867223,
+      "learning_rate": 4.970568261915531e-05,
+      "loss": 1.5712,
+      "step": 294
+    },
+    {
+      "epoch": 0.43486272341993737,
+      "grad_norm": 1.8046863770584642,
+      "learning_rate": 4.970264664393614e-05,
+      "loss": 1.4221,
+      "step": 295
+    },
+    {
+      "epoch": 0.4363368343467846,
+      "grad_norm": 1.594034954850858,
+      "learning_rate": 4.9699595184120853e-05,
+      "loss": 1.5871,
+      "step": 296
+    },
+    {
+      "epoch": 0.43781094527363185,
+      "grad_norm": 1.7385663773792093,
+      "learning_rate": 4.9696528241622244e-05,
+      "loss": 1.6057,
+      "step": 297
+    },
+    {
+      "epoch": 0.4392850562004791,
+      "grad_norm": 1.8332213743298447,
+      "learning_rate": 4.9693445818362783e-05,
+      "loss": 1.4573,
+      "step": 298
+    },
+    {
+      "epoch": 0.4407591671273263,
+      "grad_norm": 1.89901624107547,
+      "learning_rate": 4.969034791627466e-05,
+      "loss": 1.5596,
+      "step": 299
+    },
+    {
+      "epoch": 0.44223327805417356,
+      "grad_norm": 2.0104218978682007,
+      "learning_rate": 4.9687234537299765e-05,
+      "loss": 1.3984,
+      "step": 300
+    },
+    {
+      "epoch": 0.4437073889810208,
+      "grad_norm": 1.843432770933328,
+      "learning_rate": 4.968410568338967e-05,
+      "loss": 1.515,
+      "step": 301
+    },
+    {
+      "epoch": 0.4451814999078681,
+      "grad_norm": 1.819186341159947,
+      "learning_rate": 4.968096135650569e-05,
+      "loss": 1.5026,
+      "step": 302
+    },
+    {
+      "epoch": 0.44665561083471533,
+      "grad_norm": 1.7817738749784682,
+      "learning_rate": 4.9677801558618795e-05,
+      "loss": 1.589,
+      "step": 303
+    },
+    {
+      "epoch": 0.44812972176156257,
+      "grad_norm": 1.8840459052282146,
+      "learning_rate": 4.967462629170969e-05,
+      "loss": 1.5816,
+      "step": 304
+    },
+    {
+      "epoch": 0.4496038326884098,
+      "grad_norm": 1.8997332385669479,
+      "learning_rate": 4.967143555776873e-05,
+      "loss": 1.5773,
+      "step": 305
+    },
+    {
+      "epoch": 0.45107794361525705,
+      "grad_norm": 2.0175375920495884,
+      "learning_rate": 4.9668229358796014e-05,
+      "loss": 1.4526,
+      "step": 306
+    },
+    {
+      "epoch": 0.4525520545421043,
+      "grad_norm": 1.9467586672393884,
+      "learning_rate": 4.966500769680131e-05,
+      "loss": 1.3903,
+      "step": 307
+    },
+    {
+      "epoch": 0.4540261654689515,
+      "grad_norm": 1.8750722871197363,
+      "learning_rate": 4.966177057380409e-05,
+      "loss": 1.5008,
+      "step": 308
+    },
+    {
+      "epoch": 0.45550027639579876,
+      "grad_norm": 1.8761046652318696,
+      "learning_rate": 4.965851799183349e-05,
+      "loss": 1.5039,
+      "step": 309
+    },
+    {
+      "epoch": 0.456974387322646,
+      "grad_norm": 1.8297284259339959,
+      "learning_rate": 4.9655249952928375e-05,
+      "loss": 1.3733,
+      "step": 310
+    },
+    {
+      "epoch": 0.4584484982494933,
+      "grad_norm": 1.609772420163937,
+      "learning_rate": 4.965196645913728e-05,
+      "loss": 1.5001,
+      "step": 311
+    },
+    {
+      "epoch": 0.45992260917634054,
+      "grad_norm": 1.8387292292580308,
+      "learning_rate": 4.964866751251842e-05,
+      "loss": 1.74,
+      "step": 312
+    },
+    {
+      "epoch": 0.4613967201031878,
+      "grad_norm": 1.784231468079857,
+      "learning_rate": 4.964535311513971e-05,
+      "loss": 1.539,
+      "step": 313
+    },
+    {
+      "epoch": 0.462870831030035,
+      "grad_norm": 1.7147122204505345,
+      "learning_rate": 4.9642023269078745e-05,
+      "loss": 1.5192,
+      "step": 314
+    },
+    {
+      "epoch": 0.46434494195688225,
+      "grad_norm": 1.6686955818132867,
+      "learning_rate": 4.963867797642281e-05,
+      "loss": 1.4396,
+      "step": 315
+    },
+    {
+      "epoch": 0.4658190528837295,
+      "grad_norm": 1.7852084696872015,
+      "learning_rate": 4.963531723926885e-05,
+      "loss": 1.6137,
+      "step": 316
+    },
+    {
+      "epoch": 0.46729316381057673,
+      "grad_norm": 1.5856517665890846,
+      "learning_rate": 4.963194105972353e-05,
+      "loss": 1.6777,
+      "step": 317
+    },
+    {
+      "epoch": 0.46876727473742397,
+      "grad_norm": 1.6801835857854193,
+      "learning_rate": 4.962854943990316e-05,
+      "loss": 1.4395,
+      "step": 318
+    },
+    {
+      "epoch": 0.47024138566427126,
+      "grad_norm": 1.7172009510072304,
+      "learning_rate": 4.962514238193375e-05,
+      "loss": 1.2703,
+      "step": 319
+    },
+    {
+      "epoch": 0.4717154965911185,
+      "grad_norm": 1.8608378578241502,
+      "learning_rate": 4.9621719887950966e-05,
+      "loss": 1.711,
+      "step": 320
+    },
+    {
+      "epoch": 0.47318960751796574,
+      "grad_norm": 1.8098349215642842,
+      "learning_rate": 4.9618281960100164e-05,
+      "loss": 1.273,
+      "step": 321
+    },
+    {
+      "epoch": 0.474663718444813,
+      "grad_norm": 1.8268931876760737,
+      "learning_rate": 4.9614828600536386e-05,
+      "loss": 1.5861,
+      "step": 322
+    },
+    {
+      "epoch": 0.4761378293716602,
+      "grad_norm": 1.8848761807763157,
+      "learning_rate": 4.9611359811424324e-05,
+      "loss": 1.6667,
+      "step": 323
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 1.7224359695869051,
+      "learning_rate": 4.960787559493836e-05,
+      "loss": 1.4574,
+      "step": 324
+    },
+    {
+      "epoch": 0.4790860512253547,
+      "grad_norm": 1.8962990604458858,
+      "learning_rate": 4.960437595326253e-05,
+      "loss": 1.5485,
+      "step": 325
+    },
+    {
+      "epoch": 0.48056016215220193,
+      "grad_norm": 1.6479640479656865,
+      "learning_rate": 4.960086088859055e-05,
+      "loss": 1.3061,
+      "step": 326
+    },
+    {
+      "epoch": 0.4820342730790492,
+      "grad_norm": 1.693045229520003,
+      "learning_rate": 4.95973304031258e-05,
+      "loss": 1.5577,
+      "step": 327
+    },
+    {
+      "epoch": 0.48350838400589646,
+      "grad_norm": 1.7807957254002036,
+      "learning_rate": 4.9593784499081336e-05,
+      "loss": 1.5978,
+      "step": 328
+    },
+    {
+      "epoch": 0.4849824949327437,
+      "grad_norm": 1.7486963184507502,
+      "learning_rate": 4.959022317867986e-05,
+      "loss": 1.5676,
+      "step": 329
+    },
+    {
+      "epoch": 0.48645660585959094,
+      "grad_norm": 1.7930202102756085,
+      "learning_rate": 4.9586646444153764e-05,
+      "loss": 1.4347,
+      "step": 330
+    },
+    {
+      "epoch": 0.4879307167864382,
+      "grad_norm": 1.6959663220246723,
+      "learning_rate": 4.958305429774507e-05,
+      "loss": 1.4695,
+      "step": 331
+    },
+    {
+      "epoch": 0.4894048277132854,
+      "grad_norm": 1.7847901350756699,
+      "learning_rate": 4.9579446741705485e-05,
+      "loss": 1.5542,
+      "step": 332
+    },
+    {
+      "epoch": 0.49087893864013266,
+      "grad_norm": 1.736363619773062,
+      "learning_rate": 4.957582377829637e-05,
+      "loss": 1.5498,
+      "step": 333
+    },
+    {
+      "epoch": 0.4923530495669799,
+      "grad_norm": 1.7008940757858666,
+      "learning_rate": 4.957218540978874e-05,
+      "loss": 1.5284,
+      "step": 334
+    },
+    {
+      "epoch": 0.49382716049382713,
+      "grad_norm": 1.5816515808404354,
+      "learning_rate": 4.9568531638463264e-05,
+      "loss": 1.414,
+      "step": 335
+    },
+    {
+      "epoch": 0.4953012714206744,
+      "grad_norm": 1.7616933072234282,
+      "learning_rate": 4.9564862466610284e-05,
+      "loss": 1.5051,
+      "step": 336
+    },
+    {
+      "epoch": 0.49677538234752167,
+      "grad_norm": 1.7645106892530869,
+      "learning_rate": 4.9561177896529764e-05,
+      "loss": 1.5276,
+      "step": 337
+    },
+    {
+      "epoch": 0.4982494932743689,
+      "grad_norm": 1.6571621812083663,
+      "learning_rate": 4.9557477930531346e-05,
+      "loss": 1.6586,
+      "step": 338
+    },
+    {
+      "epoch": 0.49972360420121614,
+      "grad_norm": 1.8880966557152847,
+      "learning_rate": 4.9553762570934314e-05,
+      "loss": 1.5564,
+      "step": 339
+    },
+    {
+      "epoch": 0.5011977151280634,
+      "grad_norm": 1.8661837892738515,
+      "learning_rate": 4.955003182006761e-05,
+      "loss": 1.4869,
+      "step": 340
+    },
+    {
+      "epoch": 0.5026718260549107,
+      "grad_norm": 1.7412277794246471,
+      "learning_rate": 4.954628568026981e-05,
+      "loss": 1.336,
+      "step": 341
+    },
+    {
+      "epoch": 0.5041459369817579,
+      "grad_norm": 1.6618906094555774,
+      "learning_rate": 4.954252415388914e-05,
+      "loss": 1.5603,
+      "step": 342
+    },
+    {
+      "epoch": 0.5056200479086052,
+      "grad_norm": 1.6000290391194125,
+      "learning_rate": 4.953874724328347e-05,
+      "loss": 1.4097,
+      "step": 343
+    },
+    {
+      "epoch": 0.5070941588354524,
+      "grad_norm": 1.5390411220556837,
+      "learning_rate": 4.953495495082032e-05,
+      "loss": 1.4327,
+      "step": 344
+    },
+    {
+      "epoch": 0.5085682697622996,
+      "grad_norm": 1.8537928672356163,
+      "learning_rate": 4.953114727887686e-05,
+      "loss": 1.4635,
+      "step": 345
+    },
+    {
+      "epoch": 0.5100423806891469,
+      "grad_norm": 1.9248772718271365,
+      "learning_rate": 4.952732422983989e-05,
+      "loss": 1.6069,
+      "step": 346
+    },
+    {
+      "epoch": 0.5115164916159941,
+      "grad_norm": 1.8327897378560116,
+      "learning_rate": 4.9523485806105826e-05,
+      "loss": 1.506,
+      "step": 347
+    },
+    {
+      "epoch": 0.5129906025428413,
+      "grad_norm": 1.6867867834621388,
+      "learning_rate": 4.951963201008076e-05,
+      "loss": 1.5001,
+      "step": 348
+    },
+    {
+      "epoch": 0.5144647134696886,
+      "grad_norm": 1.727999818210239,
+      "learning_rate": 4.9515762844180405e-05,
+      "loss": 1.5658,
+      "step": 349
+    },
+    {
+      "epoch": 0.5159388243965358,
+      "grad_norm": 1.792338607523979,
+      "learning_rate": 4.9511878310830106e-05,
+      "loss": 1.4469,
+      "step": 350
+    },
+    {
+      "epoch": 0.5174129353233831,
+      "grad_norm": 1.665252689741385,
+      "learning_rate": 4.950797841246484e-05,
+      "loss": 1.5282,
+      "step": 351
+    },
+    {
+      "epoch": 0.5188870462502303,
+      "grad_norm": 1.7260577124166554,
+      "learning_rate": 4.950406315152921e-05,
+      "loss": 1.4513,
+      "step": 352
+    },
+    {
+      "epoch": 0.5203611571770775,
+      "grad_norm": 1.6363130125730734,
+      "learning_rate": 4.9500132530477475e-05,
+      "loss": 1.5438,
+      "step": 353
+    },
+    {
+      "epoch": 0.5218352681039248,
+      "grad_norm": 1.8963660724800133,
+      "learning_rate": 4.949618655177348e-05,
+      "loss": 1.5857,
+      "step": 354
+    },
+    {
+      "epoch": 0.523309379030772,
+      "grad_norm": 1.8304866957588684,
+      "learning_rate": 4.949222521789074e-05,
+      "loss": 1.5886,
+      "step": 355
+    },
+    {
+      "epoch": 0.5247834899576194,
+      "grad_norm": 1.7552667111914908,
+      "learning_rate": 4.948824853131236e-05,
+      "loss": 1.4921,
+      "step": 356
+    },
+    {
+      "epoch": 0.5262576008844666,
+      "grad_norm": 1.7218471019444106,
+      "learning_rate": 4.948425649453111e-05,
+      "loss": 1.3544,
+      "step": 357
+    },
+    {
+      "epoch": 0.5277317118113138,
+      "grad_norm": 1.8410092041887887,
+      "learning_rate": 4.948024911004933e-05,
+      "loss": 1.5307,
+      "step": 358
+    },
+    {
+      "epoch": 0.5292058227381611,
+      "grad_norm": 1.84916087012853,
+      "learning_rate": 4.9476226380379014e-05,
+      "loss": 1.6957,
+      "step": 359
+    },
+    {
+      "epoch": 0.5306799336650083,
+      "grad_norm": 1.7845316990126243,
+      "learning_rate": 4.947218830804178e-05,
+      "loss": 1.5464,
+      "step": 360
+    },
+    {
+      "epoch": 0.5321540445918556,
+      "grad_norm": 1.9532893607600796,
+      "learning_rate": 4.946813489556883e-05,
+      "loss": 1.54,
+      "step": 361
+    },
+    {
+      "epoch": 0.5336281555187028,
+      "grad_norm": 1.7193653045680455,
+      "learning_rate": 4.946406614550103e-05,
+      "loss": 1.5283,
+      "step": 362
+    },
+    {
+      "epoch": 0.53510226644555,
+      "grad_norm": 1.7851546473255935,
+      "learning_rate": 4.945998206038881e-05,
+      "loss": 1.5738,
+      "step": 363
+    },
+    {
+      "epoch": 0.5365763773723973,
+      "grad_norm": 1.809106082730639,
+      "learning_rate": 4.945588264279225e-05,
+      "loss": 1.6902,
+      "step": 364
+    },
+    {
+      "epoch": 0.5380504882992445,
+      "grad_norm": 1.8137877920828727,
+      "learning_rate": 4.945176789528102e-05,
+      "loss": 1.5106,
+      "step": 365
+    },
+    {
+      "epoch": 0.5395245992260918,
+      "grad_norm": 1.683959213206895,
+      "learning_rate": 4.944763782043441e-05,
+      "loss": 1.4803,
+      "step": 366
+    },
+    {
+      "epoch": 0.540998710152939,
+      "grad_norm": 1.6836850170348148,
+      "learning_rate": 4.944349242084131e-05,
+      "loss": 1.5327,
+      "step": 367
+    },
+    {
+      "epoch": 0.5424728210797862,
+      "grad_norm": 1.7955376333670994,
+      "learning_rate": 4.943933169910023e-05,
+      "loss": 1.6079,
+      "step": 368
+    },
+    {
+      "epoch": 0.5439469320066335,
+      "grad_norm": 1.6751273064131693,
+      "learning_rate": 4.9435155657819266e-05,
+      "loss": 1.3816,
+      "step": 369
+    },
+    {
+      "epoch": 0.5454210429334807,
+      "grad_norm": 1.771809036130391,
+      "learning_rate": 4.9430964299616136e-05,
+      "loss": 1.4085,
+      "step": 370
+    },
+    {
+      "epoch": 0.5468951538603279,
+      "grad_norm": 1.8674027705894036,
+      "learning_rate": 4.942675762711813e-05,
+      "loss": 1.466,
+      "step": 371
+    },
+    {
+      "epoch": 0.5483692647871752,
+      "grad_norm": 1.7510025013982573,
+      "learning_rate": 4.942253564296218e-05,
+      "loss": 1.5686,
+      "step": 372
+    },
+    {
+      "epoch": 0.5498433757140225,
+      "grad_norm": 1.7751545261226611,
+      "learning_rate": 4.9418298349794767e-05,
+      "loss": 1.7306,
+      "step": 373
+    },
+    {
+      "epoch": 0.5513174866408698,
+      "grad_norm": 1.7897797192193394,
+      "learning_rate": 4.941404575027202e-05,
+      "loss": 1.4025,
+      "step": 374
+    },
+    {
+      "epoch": 0.552791597567717,
+      "grad_norm": 1.5912111292883768,
+      "learning_rate": 4.9409777847059625e-05,
+      "loss": 1.5827,
+      "step": 375
+    },
+    {
+      "epoch": 0.5542657084945642,
+      "grad_norm": 1.7964859914554656,
+      "learning_rate": 4.940549464283287e-05,
+      "loss": 1.3956,
+      "step": 376
+    },
+    {
+      "epoch": 0.5557398194214115,
+      "grad_norm": 1.6214447744482985,
+      "learning_rate": 4.940119614027663e-05,
+      "loss": 1.4579,
+      "step": 377
+    },
+    {
+      "epoch": 0.5572139303482587,
+      "grad_norm": 1.7065987813879293,
+      "learning_rate": 4.93968823420854e-05,
+      "loss": 1.6299,
+      "step": 378
+    },
+    {
+      "epoch": 0.558688041275106,
+      "grad_norm": 1.8918716482051499,
+      "learning_rate": 4.9392553250963215e-05,
+      "loss": 1.4715,
+      "step": 379
+    },
+    {
+      "epoch": 0.5601621522019532,
+      "grad_norm": 1.8153975116080192,
+      "learning_rate": 4.9388208869623734e-05,
+      "loss": 1.6462,
+      "step": 380
+    },
+    {
+      "epoch": 0.5616362631288004,
+      "grad_norm": 1.787132388107462,
+      "learning_rate": 4.938384920079019e-05,
+      "loss": 1.2408,
+      "step": 381
+    },
+    {
+      "epoch": 0.5631103740556477,
+      "grad_norm": 1.803564077035902,
+      "learning_rate": 4.937947424719538e-05,
+      "loss": 1.4267,
+      "step": 382
+    },
+    {
+      "epoch": 0.5645844849824949,
+      "grad_norm": 1.9134134866072703,
+      "learning_rate": 4.937508401158171e-05,
+      "loss": 1.5415,
+      "step": 383
+    },
+    {
+      "epoch": 0.5660585959093422,
+      "grad_norm": 1.920379801378615,
+      "learning_rate": 4.937067849670115e-05,
+      "loss": 1.4923,
+      "step": 384
+    },
+    {
+      "epoch": 0.5675327068361894,
+      "grad_norm": 1.7576246373137066,
+      "learning_rate": 4.936625770531525e-05,
+      "loss": 1.4902,
+      "step": 385
+    },
+    {
+      "epoch": 0.5690068177630366,
+      "grad_norm": 1.7083754495530958,
+      "learning_rate": 4.936182164019515e-05,
+      "loss": 1.4744,
+      "step": 386
+    },
+    {
+      "epoch": 0.5704809286898839,
+      "grad_norm": 1.7812743603563133,
+      "learning_rate": 4.935737030412153e-05,
+      "loss": 1.3867,
+      "step": 387
+    },
+    {
+      "epoch": 0.5719550396167311,
+      "grad_norm": 1.7760903434757664,
+      "learning_rate": 4.935290369988468e-05,
+      "loss": 1.5744,
+      "step": 388
+    },
+    {
+      "epoch": 0.5734291505435785,
+      "grad_norm": 1.6242917791685432,
+      "learning_rate": 4.934842183028443e-05,
+      "loss": 1.3365,
+      "step": 389
+    },
+    {
+      "epoch": 0.5749032614704257,
+      "grad_norm": 1.6025642053653,
+      "learning_rate": 4.9343924698130206e-05,
+      "loss": 1.426,
+      "step": 390
+    },
+    {
+      "epoch": 0.5763773723972729,
+      "grad_norm": 1.7101751571314143,
+      "learning_rate": 4.9339412306240984e-05,
+      "loss": 1.6726,
+      "step": 391
+    },
+    {
+      "epoch": 0.5778514833241202,
+      "grad_norm": 1.6439576949688486,
+      "learning_rate": 4.933488465744531e-05,
+      "loss": 1.4645,
+      "step": 392
+    },
+    {
+      "epoch": 0.5793255942509674,
+      "grad_norm": 1.8196544849532825,
+      "learning_rate": 4.933034175458129e-05,
+      "loss": 1.4769,
+      "step": 393
+    },
+    {
+      "epoch": 0.5807997051778147,
+      "grad_norm": 2.076573072052071,
+      "learning_rate": 4.9325783600496596e-05,
+      "loss": 1.4516,
+      "step": 394
+    },
+    {
+      "epoch": 0.5822738161046619,
+      "grad_norm": 1.8776443907458529,
+      "learning_rate": 4.9321210198048465e-05,
+      "loss": 1.4449,
+      "step": 395
+    },
+    {
+      "epoch": 0.5837479270315091,
+      "grad_norm": 1.8768191731394743,
+      "learning_rate": 4.931662155010367e-05,
+      "loss": 1.4863,
+      "step": 396
+    },
+    {
+      "epoch": 0.5852220379583564,
+      "grad_norm": 1.774138795100083,
+      "learning_rate": 4.931201765953858e-05,
+      "loss": 1.4567,
+      "step": 397
+    },
+    {
+      "epoch": 0.5866961488852036,
+      "grad_norm": 2.095018258937029,
+      "learning_rate": 4.9307398529239083e-05,
+      "loss": 1.5368,
+      "step": 398
+    },
+    {
+      "epoch": 0.5881702598120508,
+      "grad_norm": 1.7918840838262935,
+      "learning_rate": 4.930276416210063e-05,
+      "loss": 1.4562,
+      "step": 399
+    },
+    {
+      "epoch": 0.5896443707388981,
+      "grad_norm": 1.8635535850875868,
+      "learning_rate": 4.929811456102824e-05,
+      "loss": 1.6778,
+      "step": 400
+    },
+    {
+      "epoch": 0.5911184816657453,
+      "grad_norm": 1.617357991395137,
+      "learning_rate": 4.929344972893646e-05,
+      "loss": 1.563,
+      "step": 401
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 1.7088147934835842,
+      "learning_rate": 4.928876966874938e-05,
+      "loss": 1.4051,
+      "step": 402
+    },
+    {
+      "epoch": 0.5940667035194398,
+      "grad_norm": 1.6753560129131577,
+      "learning_rate": 4.9284074383400655e-05,
+      "loss": 1.4051,
+      "step": 403
+    },
+    {
+      "epoch": 0.595540814446287,
+      "grad_norm": 1.6029978205332274,
+      "learning_rate": 4.927936387583348e-05,
+      "loss": 1.3202,
+      "step": 404
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 1.7688631055274433,
+      "learning_rate": 4.9274638149000585e-05,
+      "loss": 1.5041,
+      "step": 405
+    },
+    {
+      "epoch": 0.5984890362999816,
+      "grad_norm": 1.5799863809891932,
+      "learning_rate": 4.9269897205864235e-05,
+      "loss": 1.4954,
+      "step": 406
+    },
+    {
+      "epoch": 0.5999631472268289,
+      "grad_norm": 1.6553242093291975,
+      "learning_rate": 4.926514104939625e-05,
+      "loss": 1.3095,
+      "step": 407
+    },
+    {
+      "epoch": 0.6014372581536761,
+      "grad_norm": 1.716742643816914,
+      "learning_rate": 4.9260369682577965e-05,
+      "loss": 1.6561,
+      "step": 408
+    },
+    {
+      "epoch": 0.6029113690805233,
+      "grad_norm": 1.6326547208784874,
+      "learning_rate": 4.9255583108400285e-05,
+      "loss": 1.5655,
+      "step": 409
+    },
+    {
+      "epoch": 0.6043854800073706,
+      "grad_norm": 1.5109955390899672,
+      "learning_rate": 4.9250781329863606e-05,
+      "loss": 1.4361,
+      "step": 410
+    },
+    {
+      "epoch": 0.6058595909342178,
+      "grad_norm": 1.6255240695688515,
+      "learning_rate": 4.924596434997787e-05,
+      "loss": 1.4307,
+      "step": 411
+    },
+    {
+      "epoch": 0.6073337018610651,
+      "grad_norm": 1.5463493301694566,
+      "learning_rate": 4.924113217176256e-05,
+      "loss": 1.5367,
+      "step": 412
+    },
+    {
+      "epoch": 0.6088078127879123,
+      "grad_norm": 1.6327332837906885,
+      "learning_rate": 4.9236284798246666e-05,
+      "loss": 1.5593,
+      "step": 413
+    },
+    {
+      "epoch": 0.6102819237147595,
+      "grad_norm": 2.0993855485710746,
+      "learning_rate": 4.923142223246873e-05,
+      "loss": 1.5256,
+      "step": 414
+    },
+    {
+      "epoch": 0.6117560346416068,
+      "grad_norm": 1.825255269004198,
+      "learning_rate": 4.922654447747679e-05,
+      "loss": 1.5226,
+      "step": 415
+    },
+    {
+      "epoch": 0.613230145568454,
+      "grad_norm": 1.8755328374971283,
+      "learning_rate": 4.922165153632842e-05,
+      "loss": 1.6397,
+      "step": 416
+    },
+    {
+      "epoch": 0.6147042564953012,
+      "grad_norm": 1.8949169277613347,
+      "learning_rate": 4.9216743412090694e-05,
+      "loss": 1.5982,
+      "step": 417
+    },
+    {
+      "epoch": 0.6161783674221485,
+      "grad_norm": 1.8352227950423996,
+      "learning_rate": 4.9211820107840234e-05,
+      "loss": 1.3602,
+      "step": 418
+    },
+    {
+      "epoch": 0.6176524783489957,
+      "grad_norm": 1.9041189570331363,
+      "learning_rate": 4.920688162666316e-05,
+      "loss": 1.5457,
+      "step": 419
+    },
+    {
+      "epoch": 0.619126589275843,
+      "grad_norm": 1.7763910108481864,
+      "learning_rate": 4.920192797165511e-05,
+      "loss": 1.4164,
+      "step": 420
+    },
+    {
+      "epoch": 0.6206007002026902,
+      "grad_norm": 1.7419113829898394,
+      "learning_rate": 4.919695914592122e-05,
+      "loss": 1.5608,
+      "step": 421
+    },
+    {
+      "epoch": 0.6220748111295376,
+      "grad_norm": 1.6244963085382704,
+      "learning_rate": 4.919197515257616e-05,
+      "loss": 1.2675,
+      "step": 422
+    },
+    {
+      "epoch": 0.6235489220563848,
+      "grad_norm": 1.7962890996639524,
+      "learning_rate": 4.9186975994744075e-05,
+      "loss": 1.5134,
+      "step": 423
+    },
+    {
+      "epoch": 0.625023032983232,
+      "grad_norm": 1.7239569426777772,
+      "learning_rate": 4.918196167555866e-05,
+      "loss": 1.5654,
+      "step": 424
+    },
+    {
+      "epoch": 0.6264971439100793,
+      "grad_norm": 1.8657507056948368,
+      "learning_rate": 4.9176932198163074e-05,
+      "loss": 1.3322,
+      "step": 425
+    },
+    {
+      "epoch": 0.6279712548369265,
+      "grad_norm": 1.6874855888599058,
+      "learning_rate": 4.917188756570999e-05,
+      "loss": 1.5147,
+      "step": 426
+    },
+    {
+      "epoch": 0.6294453657637737,
+      "grad_norm": 1.5605202148604673,
+      "learning_rate": 4.9166827781361594e-05,
+      "loss": 1.3247,
+      "step": 427
+    },
+    {
+      "epoch": 0.630919476690621,
+      "grad_norm": 1.7854848346434706,
+      "learning_rate": 4.916175284828955e-05,
+      "loss": 1.5542,
+      "step": 428
+    },
+    {
+      "epoch": 0.6323935876174682,
+      "grad_norm": 1.7684688115453004,
+      "learning_rate": 4.915666276967501e-05,
+      "loss": 1.5501,
+      "step": 429
+    },
+    {
+      "epoch": 0.6338676985443155,
+      "grad_norm": 1.6717944972541314,
+      "learning_rate": 4.9151557548708676e-05,
+      "loss": 1.4231,
+      "step": 430
+    },
+    {
+      "epoch": 0.6353418094711627,
+      "grad_norm": 1.8571704650231153,
+      "learning_rate": 4.9146437188590675e-05,
+      "loss": 1.6042,
+      "step": 431
+    },
+    {
+      "epoch": 0.6368159203980099,
+      "grad_norm": 1.8889238885678714,
+      "learning_rate": 4.914130169253066e-05,
+      "loss": 1.4622,
+      "step": 432
+    },
+    {
+      "epoch": 0.6382900313248572,
+      "grad_norm": 1.8406117126296115,
+      "learning_rate": 4.913615106374777e-05,
+      "loss": 1.6079,
+      "step": 433
+    },
+    {
+      "epoch": 0.6397641422517044,
+      "grad_norm": 1.7939840529593174,
+      "learning_rate": 4.91309853054706e-05,
+      "loss": 1.513,
+      "step": 434
+    },
+    {
+      "epoch": 0.6412382531785517,
+      "grad_norm": 1.7370053527169775,
+      "learning_rate": 4.912580442093727e-05,
+      "loss": 1.5361,
+      "step": 435
+    },
+    {
+      "epoch": 0.6427123641053989,
+      "grad_norm": 1.772219067088753,
+      "learning_rate": 4.9120608413395366e-05,
+      "loss": 1.5908,
+      "step": 436
+    },
+    {
+      "epoch": 0.6441864750322461,
+      "grad_norm": 1.725629114065271,
+      "learning_rate": 4.911539728610194e-05,
+      "loss": 1.4334,
+      "step": 437
+    },
+    {
+      "epoch": 0.6456605859590934,
+      "grad_norm": 1.661961550027967,
+      "learning_rate": 4.9110171042323536e-05,
+      "loss": 1.5997,
+      "step": 438
+    },
+    {
+      "epoch": 0.6471346968859407,
+      "grad_norm": 1.6235969795559844,
+      "learning_rate": 4.910492968533618e-05,
+      "loss": 1.5554,
+      "step": 439
+    },
+    {
+      "epoch": 0.648608807812788,
+      "grad_norm": 2.0796905586340557,
+      "learning_rate": 4.909967321842535e-05,
+      "loss": 1.4449,
+      "step": 440
+    },
+    {
+      "epoch": 0.6500829187396352,
+      "grad_norm": 1.5206619409980802,
+      "learning_rate": 4.9094401644886e-05,
+      "loss": 1.2062,
+      "step": 441
+    },
+    {
+      "epoch": 0.6515570296664824,
+      "grad_norm": 1.8492072211012145,
+      "learning_rate": 4.908911496802257e-05,
+      "loss": 1.5708,
+      "step": 442
+    },
+    {
+      "epoch": 0.6530311405933297,
+      "grad_norm": 1.90030844844969,
+      "learning_rate": 4.908381319114898e-05,
+      "loss": 1.4393,
+      "step": 443
+    },
+    {
+      "epoch": 0.6545052515201769,
+      "grad_norm": 2.1027586982138056,
+      "learning_rate": 4.9078496317588556e-05,
+      "loss": 1.5348,
+      "step": 444
+    },
+    {
+      "epoch": 0.6559793624470242,
+      "grad_norm": 1.774498710229856,
+      "learning_rate": 4.907316435067415e-05,
+      "loss": 1.507,
+      "step": 445
+    },
+    {
+      "epoch": 0.6574534733738714,
+      "grad_norm": 2.028774337387536,
+      "learning_rate": 4.906781729374804e-05,
+      "loss": 1.5131,
+      "step": 446
+    },
+    {
+      "epoch": 0.6589275843007186,
+      "grad_norm": 1.8197254986261917,
+      "learning_rate": 4.906245515016197e-05,
+      "loss": 1.5178,
+      "step": 447
+    },
+    {
+      "epoch": 0.6604016952275659,
+      "grad_norm": 1.821641821358822,
+      "learning_rate": 4.905707792327715e-05,
+      "loss": 1.6951,
+      "step": 448
+    },
+    {
+      "epoch": 0.6618758061544131,
+      "grad_norm": 1.7083524222199116,
+      "learning_rate": 4.9051685616464246e-05,
+      "loss": 1.4012,
+      "step": 449
+    },
+    {
+      "epoch": 0.6633499170812603,
+      "grad_norm": 1.689141428865083,
+      "learning_rate": 4.904627823310335e-05,
+      "loss": 1.3242,
+      "step": 450
+    },
+    {
+      "epoch": 0.6648240280081076,
+      "grad_norm": 1.6191904683833345,
+      "learning_rate": 4.9040855776584035e-05,
+      "loss": 1.438,
+      "step": 451
+    },
+    {
+      "epoch": 0.6662981389349548,
+      "grad_norm": 1.7294841229828566,
+      "learning_rate": 4.9035418250305314e-05,
+      "loss": 1.4644,
+      "step": 452
+    },
+    {
+      "epoch": 0.6677722498618021,
+      "grad_norm": 1.842108432150388,
+      "learning_rate": 4.9029965657675636e-05,
+      "loss": 1.492,
+      "step": 453
+    },
+    {
+      "epoch": 0.6692463607886493,
+      "grad_norm": 1.806699745654287,
+      "learning_rate": 4.9024498002112906e-05,
+      "loss": 1.4552,
+      "step": 454
+    },
+    {
+      "epoch": 0.6707204717154965,
+      "grad_norm": 1.6157211068527249,
+      "learning_rate": 4.901901528704447e-05,
+      "loss": 1.4178,
+      "step": 455
+    },
+    {
+      "epoch": 0.6721945826423439,
+      "grad_norm": 1.6756986584351095,
+      "learning_rate": 4.90135175159071e-05,
+      "loss": 1.4597,
+      "step": 456
+    },
+    {
+      "epoch": 0.6736686935691911,
+      "grad_norm": 1.7297972329884908,
+      "learning_rate": 4.900800469214703e-05,
+      "loss": 1.5601,
+      "step": 457
+    },
+    {
+      "epoch": 0.6751428044960384,
+      "grad_norm": 1.6800806339531662,
+      "learning_rate": 4.900247681921991e-05,
+      "loss": 1.5899,
+      "step": 458
+    },
+    {
+      "epoch": 0.6766169154228856,
+      "grad_norm": 1.6239636858088693,
+      "learning_rate": 4.899693390059082e-05,
+      "loss": 1.513,
+      "step": 459
+    },
+    {
+      "epoch": 0.6780910263497328,
+      "grad_norm": 1.7149385673708424,
+      "learning_rate": 4.89913759397343e-05,
+      "loss": 1.5596,
+      "step": 460
+    },
+    {
+      "epoch": 0.6795651372765801,
+      "grad_norm": 1.7551631374558498,
+      "learning_rate": 4.898580294013428e-05,
+      "loss": 1.5048,
+      "step": 461
+    },
+    {
+      "epoch": 0.6810392482034273,
+      "grad_norm": 1.790992861471002,
+      "learning_rate": 4.898021490528415e-05,
+      "loss": 1.4767,
+      "step": 462
+    },
+    {
+      "epoch": 0.6825133591302746,
+      "grad_norm": 1.832508992775453,
+      "learning_rate": 4.89746118386867e-05,
+      "loss": 1.318,
+      "step": 463
+    },
+    {
+      "epoch": 0.6839874700571218,
+      "grad_norm": 1.5878711056418666,
+      "learning_rate": 4.8968993743854176e-05,
+      "loss": 1.5015,
+      "step": 464
+    },
+    {
+      "epoch": 0.685461580983969,
+      "grad_norm": 1.7137147841115068,
+      "learning_rate": 4.89633606243082e-05,
+      "loss": 1.3526,
+      "step": 465
+    },
+    {
+      "epoch": 0.6869356919108163,
+      "grad_norm": 1.627229628130214,
+      "learning_rate": 4.895771248357983e-05,
+      "loss": 1.4907,
+      "step": 466
+    },
+    {
+      "epoch": 0.6884098028376635,
+      "grad_norm": 1.7391438947550655,
+      "learning_rate": 4.895204932520957e-05,
+      "loss": 1.4798,
+      "step": 467
+    },
+    {
+      "epoch": 0.6898839137645107,
+      "grad_norm": 1.7775534705209048,
+      "learning_rate": 4.8946371152747285e-05,
+      "loss": 1.3824,
+      "step": 468
+    },
+    {
+      "epoch": 0.691358024691358,
+      "grad_norm": 1.8921013734527743,
+      "learning_rate": 4.8940677969752295e-05,
+      "loss": 1.5791,
+      "step": 469
+    },
+    {
+      "epoch": 0.6928321356182052,
+      "grad_norm": 1.8413598683940766,
+      "learning_rate": 4.893496977979331e-05,
+      "loss": 1.6194,
+      "step": 470
+    },
+    {
+      "epoch": 0.6943062465450525,
+      "grad_norm": 1.8377235328533845,
+      "learning_rate": 4.892924658644844e-05,
+      "loss": 1.5055,
+      "step": 471
+    },
+    {
+      "epoch": 0.6957803574718998,
+      "grad_norm": 1.7725471580699306,
+      "learning_rate": 4.892350839330522e-05,
+      "loss": 1.4665,
+      "step": 472
+    },
+    {
+      "epoch": 0.697254468398747,
+      "grad_norm": 1.9568221233659708,
+      "learning_rate": 4.891775520396057e-05,
+      "loss": 1.4196,
+      "step": 473
+    },
+    {
+      "epoch": 0.6987285793255943,
+      "grad_norm": 1.7010672500969553,
+      "learning_rate": 4.8911987022020823e-05,
+      "loss": 1.5434,
+      "step": 474
+    },
+    {
+      "epoch": 0.7002026902524415,
+      "grad_norm": 1.812554523042569,
+      "learning_rate": 4.89062038511017e-05,
+      "loss": 1.4831,
+      "step": 475
+    },
+    {
+      "epoch": 0.7016768011792888,
+      "grad_norm": 1.7774638041098487,
+      "learning_rate": 4.8900405694828313e-05,
+      "loss": 1.3765,
+      "step": 476
+    },
+    {
+      "epoch": 0.703150912106136,
+      "grad_norm": 1.6492734118446895,
+      "learning_rate": 4.8894592556835186e-05,
+      "loss": 1.3158,
+      "step": 477
+    },
+    {
+      "epoch": 0.7046250230329832,
+      "grad_norm": 1.7110334395591669,
+      "learning_rate": 4.8888764440766225e-05,
+      "loss": 1.4758,
+      "step": 478
+    },
+    {
+      "epoch": 0.7060991339598305,
+      "grad_norm": 1.7046997069875438,
+      "learning_rate": 4.888292135027472e-05,
+      "loss": 1.4444,
+      "step": 479
+    },
+    {
+      "epoch": 0.7075732448866777,
+      "grad_norm": 1.7047225039460836,
+      "learning_rate": 4.887706328902335e-05,
+      "loss": 1.5892,
+      "step": 480
+    },
+    {
+      "epoch": 0.709047355813525,
+      "grad_norm": 1.7277817369697133,
+      "learning_rate": 4.8871190260684174e-05,
+      "loss": 1.4221,
+      "step": 481
+    },
+    {
+      "epoch": 0.7105214667403722,
+      "grad_norm": 1.592764905278726,
+      "learning_rate": 4.886530226893865e-05,
+      "loss": 1.4581,
+      "step": 482
+    },
+    {
+      "epoch": 0.7119955776672194,
+      "grad_norm": 2.1578804154588025,
+      "learning_rate": 4.88593993174776e-05,
+      "loss": 1.6157,
+      "step": 483
+    },
+    {
+      "epoch": 0.7134696885940667,
+      "grad_norm": 1.8208716192810515,
+      "learning_rate": 4.885348141000122e-05,
+      "loss": 1.4467,
+      "step": 484
+    },
+    {
+      "epoch": 0.7149437995209139,
+      "grad_norm": 1.7555074540656064,
+      "learning_rate": 4.8847548550219105e-05,
+      "loss": 1.5836,
+      "step": 485
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 1.5560618704334026,
+      "learning_rate": 4.884160074185019e-05,
+      "loss": 1.3614,
+      "step": 486
+    },
+    {
+      "epoch": 0.7178920213746084,
+      "grad_norm": 1.8615176893895358,
+      "learning_rate": 4.8835637988622804e-05,
+      "loss": 1.3898,
+      "step": 487
+    },
+    {
+      "epoch": 0.7193661323014556,
+      "grad_norm": 1.7605236931343822,
+      "learning_rate": 4.8829660294274636e-05,
+      "loss": 1.4536,
+      "step": 488
+    },
+    {
+      "epoch": 0.720840243228303,
+      "grad_norm": 1.6492765198767378,
+      "learning_rate": 4.8823667662552744e-05,
+      "loss": 1.3982,
+      "step": 489
+    },
+    {
+      "epoch": 0.7223143541551502,
+      "grad_norm": 1.8220000732832222,
+      "learning_rate": 4.881766009721354e-05,
+      "loss": 1.4368,
+      "step": 490
+    },
+    {
+      "epoch": 0.7237884650819975,
+      "grad_norm": 1.5817778207747901,
+      "learning_rate": 4.8811637602022806e-05,
+      "loss": 1.5633,
+      "step": 491
+    },
+    {
+      "epoch": 0.7252625760088447,
+      "grad_norm": 1.673540230811404,
+      "learning_rate": 4.8805600180755685e-05,
+      "loss": 1.6391,
+      "step": 492
+    },
+    {
+      "epoch": 0.7267366869356919,
+      "grad_norm": 1.6729552442283937,
+      "learning_rate": 4.8799547837196667e-05,
+      "loss": 1.4103,
+      "step": 493
+    },
+    {
+      "epoch": 0.7282107978625392,
+      "grad_norm": 1.8937503424414397,
+      "learning_rate": 4.87934805751396e-05,
+      "loss": 1.492,
+      "step": 494
+    },
+    {
+      "epoch": 0.7296849087893864,
+      "grad_norm": 1.6432538723385648,
+      "learning_rate": 4.8787398398387684e-05,
+      "loss": 1.5379,
+      "step": 495
+    },
+    {
+      "epoch": 0.7311590197162336,
+      "grad_norm": 1.6910717217086466,
+      "learning_rate": 4.878130131075347e-05,
+      "loss": 1.5918,
+      "step": 496
+    },
+    {
+      "epoch": 0.7326331306430809,
+      "grad_norm": 1.553604975934806,
+      "learning_rate": 4.877518931605885e-05,
+      "loss": 1.32,
+      "step": 497
+    },
+    {
+      "epoch": 0.7341072415699281,
+      "grad_norm": 1.5701799479095186,
+      "learning_rate": 4.8769062418135066e-05,
+      "loss": 1.4937,
+      "step": 498
+    },
+    {
+      "epoch": 0.7355813524967754,
+      "grad_norm": 1.6125995960084167,
+      "learning_rate": 4.8762920620822704e-05,
+      "loss": 1.6703,
+      "step": 499
+    },
+    {
+      "epoch": 0.7370554634236226,
+      "grad_norm": 1.815907859107379,
+      "learning_rate": 4.875676392797168e-05,
+      "loss": 1.5666,
+      "step": 500
+    },
+    {
+      "epoch": 0.7370554634236226,
+      "eval_bleu": 0.049864103853823394,
+      "eval_bleu_1gram": 0.3517426238755257,
+      "eval_bleu_2gram": 0.12146521375505184,
+      "eval_bleu_3gram": 0.04852393719199864,
+      "eval_bleu_4gram": 0.02316378746484796,
+      "eval_rag_val_loss": 1.525708330254401,
+      "eval_rouge1": 0.3319672816480387,
+      "eval_rouge2": 0.11286492939453996,
+      "eval_rougeL": 0.31334997990494057,
+      "step": 500
+    },
+    {
+      "epoch": 0.7385295743504698,
+      "grad_norm": 1.8611478928431302,
+      "learning_rate": 4.875059234344126e-05,
+      "loss": 1.5724,
+      "step": 501
+    },
+    {
+      "epoch": 0.7400036852773171,
+      "grad_norm": 1.6171364154210777,
+      "learning_rate": 4.874440587110003e-05,
+      "loss": 1.5886,
+      "step": 502
+    },
+    {
+      "epoch": 0.7414777962041643,
+      "grad_norm": 1.7119811894105343,
+      "learning_rate": 4.873820451482592e-05,
+      "loss": 1.461,
+      "step": 503
+    },
+    {
+      "epoch": 0.7429519071310116,
+      "grad_norm": 1.6702272370204336,
+      "learning_rate": 4.873198827850618e-05,
+      "loss": 1.5485,
+      "step": 504
+    },
+    {
+      "epoch": 0.7444260180578589,
+      "grad_norm": 1.867967259220181,
+      "learning_rate": 4.872575716603739e-05,
+      "loss": 1.5593,
+      "step": 505
+    },
+    {
+      "epoch": 0.7459001289847061,
+      "grad_norm": 1.8501421152120512,
+      "learning_rate": 4.871951118132547e-05,
+      "loss": 1.4647,
+      "step": 506
+    },
+    {
+      "epoch": 0.7473742399115534,
+      "grad_norm": 1.6515600632114533,
+      "learning_rate": 4.8713250328285654e-05,
+      "loss": 1.5828,
+      "step": 507
+    },
+    {
+      "epoch": 0.7488483508384006,
+      "grad_norm": 1.7049863250944728,
+      "learning_rate": 4.8706974610842474e-05,
+      "loss": 1.4158,
+      "step": 508
+    },
+    {
+      "epoch": 0.7503224617652479,
+      "grad_norm": 1.7348378569063014,
+      "learning_rate": 4.87006840329298e-05,
+      "loss": 1.3832,
+      "step": 509
+    },
+    {
+      "epoch": 0.7517965726920951,
+      "grad_norm": 1.6345373652899626,
+      "learning_rate": 4.8694378598490826e-05,
+      "loss": 1.5045,
+      "step": 510
+    },
+    {
+      "epoch": 0.7532706836189423,
+      "grad_norm": 1.9749966174712368,
+      "learning_rate": 4.868805831147805e-05,
+      "loss": 1.6251,
+      "step": 511
+    },
+    {
+      "epoch": 0.7547447945457896,
+      "grad_norm": 1.646313134607012,
+      "learning_rate": 4.868172317585326e-05,
+      "loss": 1.554,
+      "step": 512
+    },
+    {
+      "epoch": 0.7562189054726368,
+      "grad_norm": 1.6642882144775883,
+      "learning_rate": 4.867537319558758e-05,
+      "loss": 1.4583,
+      "step": 513
+    },
+    {
+      "epoch": 0.757693016399484,
+      "grad_norm": 1.7284106551371836,
+      "learning_rate": 4.866900837466144e-05,
+      "loss": 1.5583,
+      "step": 514
+    },
+    {
+      "epoch": 0.7591671273263313,
+      "grad_norm": 1.6809198870838025,
+      "learning_rate": 4.8662628717064544e-05,
+      "loss": 1.5409,
+      "step": 515
+    },
+    {
+      "epoch": 0.7606412382531785,
+      "grad_norm": 1.562017976796623,
+      "learning_rate": 4.865623422679593e-05,
+      "loss": 1.5286,
+      "step": 516
+    },
+    {
+      "epoch": 0.7621153491800258,
+      "grad_norm": 1.75968284285208,
+      "learning_rate": 4.8649824907863894e-05,
+      "loss": 1.521,
+      "step": 517
+    },
+    {
+      "epoch": 0.763589460106873,
+      "grad_norm": 1.5779583531384793,
+      "learning_rate": 4.864340076428607e-05,
+      "loss": 1.3209,
+      "step": 518
+    },
+    {
+      "epoch": 0.7650635710337202,
+      "grad_norm": 1.7439921658716193,
+      "learning_rate": 4.863696180008937e-05,
+      "loss": 1.5267,
+      "step": 519
+    },
+    {
+      "epoch": 0.7665376819605675,
+      "grad_norm": 1.7213756965833582,
+      "learning_rate": 4.8630508019309976e-05,
+      "loss": 1.51,
+      "step": 520
+    },
+    {
+      "epoch": 0.7680117928874147,
+      "grad_norm": 1.7651257779743919,
+      "learning_rate": 4.8624039425993375e-05,
+      "loss": 1.5558,
+      "step": 521
+    },
+    {
+      "epoch": 0.7694859038142621,
+      "grad_norm": 1.6010956618817953,
+      "learning_rate": 4.861755602419434e-05,
+      "loss": 1.5821,
+      "step": 522
+    },
+    {
+      "epoch": 0.7709600147411093,
+      "grad_norm": 1.4458724303369055,
+      "learning_rate": 4.861105781797692e-05,
+      "loss": 1.4634,
+      "step": 523
+    },
+    {
+      "epoch": 0.7724341256679566,
+      "grad_norm": 1.6344709233276349,
+      "learning_rate": 4.8604544811414465e-05,
+      "loss": 1.4633,
+      "step": 524
+    },
+    {
+      "epoch": 0.7739082365948038,
+      "grad_norm": 1.5475520184082878,
+      "learning_rate": 4.859801700858957e-05,
+      "loss": 1.4808,
+      "step": 525
+    },
+    {
+      "epoch": 0.775382347521651,
+      "grad_norm": 1.7307446262045316,
+      "learning_rate": 4.859147441359412e-05,
+      "loss": 1.576,
+      "step": 526
+    },
+    {
+      "epoch": 0.7768564584484983,
+      "grad_norm": 1.7574041273683978,
+      "learning_rate": 4.858491703052927e-05,
+      "loss": 1.4243,
+      "step": 527
+    },
+    {
+      "epoch": 0.7783305693753455,
+      "grad_norm": 1.5035501430054596,
+      "learning_rate": 4.8578344863505464e-05,
+      "loss": 1.4507,
+      "step": 528
+    },
+    {
+      "epoch": 0.7798046803021927,
+      "grad_norm": 1.5982972144115446,
+      "learning_rate": 4.857175791664238e-05,
+      "loss": 1.3406,
+      "step": 529
+    },
+    {
+      "epoch": 0.78127879122904,
+      "grad_norm": 1.7771944423464143,
+      "learning_rate": 4.856515619406898e-05,
+      "loss": 1.6275,
+      "step": 530
+    },
+    {
+      "epoch": 0.7827529021558872,
+      "grad_norm": 1.6675574624199347,
+      "learning_rate": 4.855853969992349e-05,
+      "loss": 1.4599,
+      "step": 531
+    },
+    {
+      "epoch": 0.7842270130827345,
+      "grad_norm": 1.7613711807223646,
+      "learning_rate": 4.8551908438353374e-05,
+      "loss": 1.4255,
+      "step": 532
+    },
+    {
+      "epoch": 0.7857011240095817,
+      "grad_norm": 1.7896445714252482,
+      "learning_rate": 4.854526241351539e-05,
+      "loss": 1.5616,
+      "step": 533
+    },
+    {
+      "epoch": 0.7871752349364289,
+      "grad_norm": 1.8317942516864372,
+      "learning_rate": 4.853860162957552e-05,
+      "loss": 1.3952,
+      "step": 534
+    },
+    {
+      "epoch": 0.7886493458632762,
+      "grad_norm": 1.8120388233539366,
+      "learning_rate": 4.8531926090709016e-05,
+      "loss": 1.3347,
+      "step": 535
+    },
+    {
+      "epoch": 0.7901234567901234,
+      "grad_norm": 1.7306410314079854,
+      "learning_rate": 4.8525235801100346e-05,
+      "loss": 1.455,
+      "step": 536
+    },
+    {
+      "epoch": 0.7915975677169707,
+      "grad_norm": 1.7714094758103223,
+      "learning_rate": 4.851853076494327e-05,
+      "loss": 1.188,
+      "step": 537
+    },
+    {
+      "epoch": 0.793071678643818,
+      "grad_norm": 1.6633401655134548,
+      "learning_rate": 4.8511810986440766e-05,
+      "loss": 1.563,
+      "step": 538
+    },
+    {
+      "epoch": 0.7945457895706652,
+      "grad_norm": 1.6108013609850818,
+      "learning_rate": 4.8505076469805054e-05,
+      "loss": 1.512,
+      "step": 539
+    },
+    {
+      "epoch": 0.7960199004975125,
+      "grad_norm": 1.76443008904392,
+      "learning_rate": 4.849832721925759e-05,
+      "loss": 1.681,
+      "step": 540
+    },
+    {
+      "epoch": 0.7974940114243597,
+      "grad_norm": 1.8573863189535162,
+      "learning_rate": 4.849156323902908e-05,
+      "loss": 1.4304,
+      "step": 541
+    },
+    {
+      "epoch": 0.798968122351207,
+      "grad_norm": 1.8755339180206456,
+      "learning_rate": 4.848478453335946e-05,
+      "loss": 1.398,
+      "step": 542
+    },
+    {
+      "epoch": 0.8004422332780542,
+      "grad_norm": 1.7021103382497447,
+      "learning_rate": 4.8477991106497874e-05,
+      "loss": 1.5522,
+      "step": 543
+    },
+    {
+      "epoch": 0.8019163442049014,
+      "grad_norm": 1.6021965446536812,
+      "learning_rate": 4.847118296270272e-05,
+      "loss": 1.4176,
+      "step": 544
+    },
+    {
+      "epoch": 0.8033904551317487,
+      "grad_norm": 1.6114137199093335,
+      "learning_rate": 4.8464360106241615e-05,
+      "loss": 1.5274,
+      "step": 545
+    },
+    {
+      "epoch": 0.8048645660585959,
+      "grad_norm": 1.7533287677898846,
+      "learning_rate": 4.845752254139139e-05,
+      "loss": 1.4279,
+      "step": 546
+    },
+    {
+      "epoch": 0.8063386769854431,
+      "grad_norm": 1.923612909817304,
+      "learning_rate": 4.845067027243809e-05,
+      "loss": 1.5484,
+      "step": 547
+    },
+    {
+      "epoch": 0.8078127879122904,
+      "grad_norm": 1.8260488866506253,
+      "learning_rate": 4.844380330367701e-05,
+      "loss": 1.4365,
+      "step": 548
+    },
+    {
+      "epoch": 0.8092868988391376,
+      "grad_norm": 1.7314524728419343,
+      "learning_rate": 4.843692163941264e-05,
+      "loss": 1.6001,
+      "step": 549
+    },
+    {
+      "epoch": 0.8107610097659849,
+      "grad_norm": 1.5860084555401721,
+      "learning_rate": 4.8430025283958645e-05,
+      "loss": 1.564,
+      "step": 550
+    },
+    {
+      "epoch": 0.8122351206928321,
+      "grad_norm": 1.6769805387028527,
+      "learning_rate": 4.842311424163797e-05,
+      "loss": 1.4055,
+      "step": 551
+    },
+    {
+      "epoch": 0.8137092316196793,
+      "grad_norm": 1.6973788417899383,
+      "learning_rate": 4.8416188516782715e-05,
+      "loss": 1.5738,
+      "step": 552
+    },
+    {
+      "epoch": 0.8151833425465266,
+      "grad_norm": 2.0282141667354168,
+      "learning_rate": 4.84092481137342e-05,
+      "loss": 1.5318,
+      "step": 553
+    },
+    {
+      "epoch": 0.8166574534733738,
+      "grad_norm": 1.5529009651691543,
+      "learning_rate": 4.840229303684294e-05,
+      "loss": 1.2693,
+      "step": 554
+    },
+    {
+      "epoch": 0.8181315644002212,
+      "grad_norm": 1.606619466600095,
+      "learning_rate": 4.8395323290468655e-05,
+      "loss": 1.4098,
+      "step": 555
+    },
+    {
+      "epoch": 0.8196056753270684,
+      "grad_norm": 1.6385132050108706,
+      "learning_rate": 4.838833887898026e-05,
+      "loss": 1.5526,
+      "step": 556
+    },
+    {
+      "epoch": 0.8210797862539156,
+      "grad_norm": 1.611953891274566,
+      "learning_rate": 4.838133980675586e-05,
+      "loss": 1.3188,
+      "step": 557
+    },
+    {
+      "epoch": 0.8225538971807629,
+      "grad_norm": 1.697205150655051,
+      "learning_rate": 4.837432607818275e-05,
+      "loss": 1.4741,
+      "step": 558
+    },
+    {
+      "epoch": 0.8240280081076101,
+      "grad_norm": 1.8683644818860858,
+      "learning_rate": 4.836729769765741e-05,
+      "loss": 1.5286,
+      "step": 559
+    },
+    {
+      "epoch": 0.8255021190344574,
+      "grad_norm": 1.6028531712224687,
+      "learning_rate": 4.83602546695855e-05,
+      "loss": 1.4908,
+      "step": 560
+    },
+    {
+      "epoch": 0.8269762299613046,
+      "grad_norm": 1.9011592741517895,
+      "learning_rate": 4.835319699838189e-05,
+      "loss": 1.7057,
+      "step": 561
+    },
+    {
+      "epoch": 0.8284503408881518,
+      "grad_norm": 1.7526635608178835,
+      "learning_rate": 4.834612468847058e-05,
+      "loss": 1.5934,
+      "step": 562
+    },
+    {
+      "epoch": 0.8299244518149991,
+      "grad_norm": 1.7057506375502405,
+      "learning_rate": 4.833903774428481e-05,
+      "loss": 1.3791,
+      "step": 563
+    },
+    {
+      "epoch": 0.8313985627418463,
+      "grad_norm": 1.7633940519083953,
+      "learning_rate": 4.833193617026692e-05,
+      "loss": 1.5183,
+      "step": 564
+    },
+    {
+      "epoch": 0.8328726736686936,
+      "grad_norm": 1.6557604228098732,
+      "learning_rate": 4.8324819970868473e-05,
+      "loss": 1.342,
+      "step": 565
+    },
+    {
+      "epoch": 0.8343467845955408,
+      "grad_norm": 1.6511287325296256,
+      "learning_rate": 4.831768915055019e-05,
+      "loss": 1.5193,
+      "step": 566
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 1.885977399983924,
+      "learning_rate": 4.831054371378194e-05,
+      "loss": 1.4996,
+      "step": 567
+    },
+    {
+      "epoch": 0.8372950064492353,
+      "grad_norm": 1.9637028015020381,
+      "learning_rate": 4.830338366504277e-05,
+      "loss": 1.4268,
+      "step": 568
+    },
+    {
+      "epoch": 0.8387691173760825,
+      "grad_norm": 1.7081776137053164,
+      "learning_rate": 4.829620900882089e-05,
+      "loss": 1.2281,
+      "step": 569
+    },
+    {
+      "epoch": 0.8402432283029297,
+      "grad_norm": 1.642584221246862,
+      "learning_rate": 4.8289019749613645e-05,
+      "loss": 1.4689,
+      "step": 570
+    },
+    {
+      "epoch": 0.8417173392297771,
+      "grad_norm": 1.7788218629411756,
+      "learning_rate": 4.8281815891927554e-05,
+      "loss": 1.4177,
+      "step": 571
+    },
+    {
+      "epoch": 0.8431914501566243,
+      "grad_norm": 1.7466419562686073,
+      "learning_rate": 4.827459744027828e-05,
+      "loss": 1.4491,
+      "step": 572
+    },
+    {
+      "epoch": 0.8446655610834716,
+      "grad_norm": 1.795111279876853,
+      "learning_rate": 4.826736439919063e-05,
+      "loss": 1.4244,
+      "step": 573
+    },
+    {
+      "epoch": 0.8461396720103188,
+      "grad_norm": 1.8952211991700518,
+      "learning_rate": 4.826011677319857e-05,
+      "loss": 1.432,
+      "step": 574
+    },
+    {
+      "epoch": 0.847613782937166,
+      "grad_norm": 1.7548746608742227,
+      "learning_rate": 4.825285456684518e-05,
+      "loss": 1.5606,
+      "step": 575
+    },
+    {
+      "epoch": 0.8490878938640133,
+      "grad_norm": 1.6553174398419468,
+      "learning_rate": 4.824557778468272e-05,
+      "loss": 1.3863,
+      "step": 576
+    },
+    {
+      "epoch": 0.8505620047908605,
+      "grad_norm": 1.7501735601189816,
+      "learning_rate": 4.823828643127255e-05,
+      "loss": 1.5383,
+      "step": 577
+    },
+    {
+      "epoch": 0.8520361157177078,
+      "grad_norm": 1.7159971212305138,
+      "learning_rate": 4.823098051118519e-05,
+      "loss": 1.5674,
+      "step": 578
+    },
+    {
+      "epoch": 0.853510226644555,
+      "grad_norm": 2.047117379084849,
+      "learning_rate": 4.822366002900027e-05,
+      "loss": 1.646,
+      "step": 579
+    },
+    {
+      "epoch": 0.8549843375714022,
+      "grad_norm": 1.8469605376030016,
+      "learning_rate": 4.821632498930656e-05,
+      "loss": 1.5541,
+      "step": 580
+    },
+    {
+      "epoch": 0.8564584484982495,
+      "grad_norm": 1.7193612138978247,
+      "learning_rate": 4.820897539670195e-05,
+      "loss": 1.4,
+      "step": 581
+    },
+    {
+      "epoch": 0.8579325594250967,
+      "grad_norm": 1.6259148296745782,
+      "learning_rate": 4.820161125579347e-05,
+      "loss": 1.5393,
+      "step": 582
+    },
+    {
+      "epoch": 0.859406670351944,
+      "grad_norm": 1.5843744703062477,
+      "learning_rate": 4.819423257119723e-05,
+      "loss": 1.4826,
+      "step": 583
+    },
+    {
+      "epoch": 0.8608807812787912,
+      "grad_norm": 1.8474203527727162,
+      "learning_rate": 4.818683934753851e-05,
+      "loss": 1.6756,
+      "step": 584
+    },
+    {
+      "epoch": 0.8623548922056384,
+      "grad_norm": 1.7197620533319622,
+      "learning_rate": 4.817943158945166e-05,
+      "loss": 1.5244,
+      "step": 585
+    },
+    {
+      "epoch": 0.8638290031324857,
+      "grad_norm": 1.7459539233722814,
+      "learning_rate": 4.817200930158015e-05,
+      "loss": 1.4336,
+      "step": 586
+    },
+    {
+      "epoch": 0.8653031140593329,
+      "grad_norm": 1.606937896435291,
+      "learning_rate": 4.816457248857657e-05,
+      "loss": 1.4565,
+      "step": 587
+    },
+    {
+      "epoch": 0.8667772249861803,
+      "grad_norm": 1.7682005739966837,
+      "learning_rate": 4.815712115510261e-05,
+      "loss": 1.3508,
+      "step": 588
+    },
+    {
+      "epoch": 0.8682513359130275,
+      "grad_norm": 1.4921241467193749,
+      "learning_rate": 4.8149655305829066e-05,
+      "loss": 1.4753,
+      "step": 589
+    },
+    {
+      "epoch": 0.8697254468398747,
+      "grad_norm": 1.633669176736577,
+      "learning_rate": 4.814217494543581e-05,
+      "loss": 1.2383,
+      "step": 590
+    },
+    {
+      "epoch": 0.871199557766722,
+      "grad_norm": 1.488404596779175,
+      "learning_rate": 4.813468007861185e-05,
+      "loss": 1.4448,
+      "step": 591
+    },
+    {
+      "epoch": 0.8726736686935692,
+      "grad_norm": 2.0575587875848593,
+      "learning_rate": 4.812717071005525e-05,
+      "loss": 1.5027,
+      "step": 592
+    },
+    {
+      "epoch": 0.8741477796204165,
+      "grad_norm": 2.0227995938473233,
+      "learning_rate": 4.8119646844473185e-05,
+      "loss": 1.5384,
+      "step": 593
+    },
+    {
+      "epoch": 0.8756218905472637,
+      "grad_norm": 1.9656235226381509,
+      "learning_rate": 4.811210848658191e-05,
+      "loss": 1.3056,
+      "step": 594
+    },
+    {
+      "epoch": 0.8770960014741109,
+      "grad_norm": 1.5471618078198595,
+      "learning_rate": 4.8104555641106766e-05,
+      "loss": 1.4889,
+      "step": 595
+    },
+    {
+      "epoch": 0.8785701124009582,
+      "grad_norm": 1.8372806262418375,
+      "learning_rate": 4.8096988312782174e-05,
+      "loss": 1.6928,
+      "step": 596
+    },
+    {
+      "epoch": 0.8800442233278054,
+      "grad_norm": 1.7147729811262298,
+      "learning_rate": 4.808940650635163e-05,
+      "loss": 1.4582,
+      "step": 597
+    },
+    {
+      "epoch": 0.8815183342546526,
+      "grad_norm": 1.7960392874432884,
+      "learning_rate": 4.8081810226567725e-05,
+      "loss": 1.5862,
+      "step": 598
+    },
+    {
+      "epoch": 0.8829924451814999,
+      "grad_norm": 1.711820331103058,
+      "learning_rate": 4.8074199478192097e-05,
+      "loss": 1.498,
+      "step": 599
+    },
+    {
+      "epoch": 0.8844665561083471,
+      "grad_norm": 1.7276360121758914,
+      "learning_rate": 4.8066574265995464e-05,
+      "loss": 1.473,
+      "step": 600
+    },
+    {
+      "epoch": 0.8859406670351944,
+      "grad_norm": 2.0970992444538403,
+      "learning_rate": 4.805893459475761e-05,
+      "loss": 1.3368,
+      "step": 601
+    },
+    {
+      "epoch": 0.8874147779620416,
+      "grad_norm": 2.1279253178377004,
+      "learning_rate": 4.805128046926739e-05,
+      "loss": 1.6721,
+      "step": 602
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.9262093770312738,
+      "learning_rate": 4.804361189432271e-05,
+      "loss": 1.6607,
+      "step": 603
+    },
+    {
+      "epoch": 0.8903629998157362,
+      "grad_norm": 1.8930343534694576,
+      "learning_rate": 4.803592887473053e-05,
+      "loss": 1.7086,
+      "step": 604
+    },
+    {
+      "epoch": 0.8918371107425834,
+      "grad_norm": 1.632568122061679,
+      "learning_rate": 4.802823141530687e-05,
+      "loss": 1.4234,
+      "step": 605
+    },
+    {
+      "epoch": 0.8933112216694307,
+      "grad_norm": 1.9635123563513197,
+      "learning_rate": 4.8020519520876816e-05,
+      "loss": 1.5608,
+      "step": 606
+    },
+    {
+      "epoch": 0.8947853325962779,
+      "grad_norm": 1.5958031537793698,
+      "learning_rate": 4.801279319627448e-05,
+      "loss": 1.4355,
+      "step": 607
+    },
+    {
+      "epoch": 0.8962594435231251,
+      "grad_norm": 1.5734508525202275,
+      "learning_rate": 4.8005052446343016e-05,
+      "loss": 1.5025,
+      "step": 608
+    },
+    {
+      "epoch": 0.8977335544499724,
+      "grad_norm": 1.590357741890919,
+      "learning_rate": 4.799729727593466e-05,
+      "loss": 1.6171,
+      "step": 609
+    },
+    {
+      "epoch": 0.8992076653768196,
+      "grad_norm": 1.7315146426430756,
+      "learning_rate": 4.798952768991063e-05,
+      "loss": 1.5196,
+      "step": 610
+    },
+    {
+      "epoch": 0.9006817763036669,
+      "grad_norm": 1.573508582834522,
+      "learning_rate": 4.798174369314123e-05,
+      "loss": 1.4365,
+      "step": 611
+    },
+    {
+      "epoch": 0.9021558872305141,
+      "grad_norm": 1.6606401523620964,
+      "learning_rate": 4.7973945290505766e-05,
+      "loss": 1.3367,
+      "step": 612
+    },
+    {
+      "epoch": 0.9036299981573613,
+      "grad_norm": 1.7051436334110197,
+      "learning_rate": 4.796613248689259e-05,
+      "loss": 1.401,
+      "step": 613
+    },
+    {
+      "epoch": 0.9051041090842086,
+      "grad_norm": 1.697160899769141,
+      "learning_rate": 4.795830528719908e-05,
+      "loss": 1.4064,
+      "step": 614
+    },
+    {
+      "epoch": 0.9065782200110558,
+      "grad_norm": 1.6792637490099682,
+      "learning_rate": 4.795046369633163e-05,
+      "loss": 1.4698,
+      "step": 615
+    },
+    {
+      "epoch": 0.908052330937903,
+      "grad_norm": 1.898250366532436,
+      "learning_rate": 4.7942607719205663e-05,
+      "loss": 1.4675,
+      "step": 616
+    },
+    {
+      "epoch": 0.9095264418647503,
+      "grad_norm": 1.6596374419127309,
+      "learning_rate": 4.793473736074561e-05,
+      "loss": 1.3765,
+      "step": 617
+    },
+    {
+      "epoch": 0.9110005527915975,
+      "grad_norm": 1.598639822019987,
+      "learning_rate": 4.792685262588492e-05,
+      "loss": 1.6055,
+      "step": 618
+    },
+    {
+      "epoch": 0.9124746637184448,
+      "grad_norm": 1.7802517419310218,
+      "learning_rate": 4.791895351956607e-05,
+      "loss": 1.6189,
+      "step": 619
+    },
+    {
+      "epoch": 0.913948774645292,
+      "grad_norm": 1.5977536943560384,
+      "learning_rate": 4.791104004674052e-05,
+      "loss": 1.3756,
+      "step": 620
+    },
+    {
+      "epoch": 0.9154228855721394,
+      "grad_norm": 1.5973833592535913,
+      "learning_rate": 4.7903112212368756e-05,
+      "loss": 1.4855,
+      "step": 621
+    },
+    {
+      "epoch": 0.9168969964989866,
+      "grad_norm": 1.5772981649370836,
+      "learning_rate": 4.789517002142026e-05,
+      "loss": 1.2613,
+      "step": 622
+    },
+    {
+      "epoch": 0.9183711074258338,
+      "grad_norm": 1.5425538096633087,
+      "learning_rate": 4.788721347887349e-05,
+      "loss": 1.4042,
+      "step": 623
+    },
+    {
+      "epoch": 0.9198452183526811,
+      "grad_norm": 1.6595327844694368,
+      "learning_rate": 4.7879242589715955e-05,
+      "loss": 1.5202,
+      "step": 624
+    },
+    {
+      "epoch": 0.9213193292795283,
+      "grad_norm": 1.882409966837432,
+      "learning_rate": 4.78712573589441e-05,
+      "loss": 1.8385,
+      "step": 625
+    },
+    {
+      "epoch": 0.9227934402063755,
+      "grad_norm": 1.5409680493339222,
+      "learning_rate": 4.7863257791563384e-05,
+      "loss": 1.5849,
+      "step": 626
+    },
+    {
+      "epoch": 0.9242675511332228,
+      "grad_norm": 1.8439664551999941,
+      "learning_rate": 4.785524389258827e-05,
+      "loss": 1.4637,
+      "step": 627
+    },
+    {
+      "epoch": 0.92574166206007,
+      "grad_norm": 1.6828936827335954,
+      "learning_rate": 4.7847215667042165e-05,
+      "loss": 1.6227,
+      "step": 628
+    },
+    {
+      "epoch": 0.9272157729869173,
+      "grad_norm": 1.6952506199104134,
+      "learning_rate": 4.78391731199575e-05,
+      "loss": 1.5701,
+      "step": 629
+    },
+    {
+      "epoch": 0.9286898839137645,
+      "grad_norm": 1.547078205979934,
+      "learning_rate": 4.7831116256375644e-05,
+      "loss": 1.4696,
+      "step": 630
+    },
+    {
+      "epoch": 0.9301639948406117,
+      "grad_norm": 1.600661531423974,
+      "learning_rate": 4.782304508134696e-05,
+      "loss": 1.5449,
+      "step": 631
+    },
+    {
+      "epoch": 0.931638105767459,
+      "grad_norm": 1.588480371845649,
+      "learning_rate": 4.7814959599930794e-05,
+      "loss": 1.2881,
+      "step": 632
+    },
+    {
+      "epoch": 0.9331122166943062,
+      "grad_norm": 1.7806309075891864,
+      "learning_rate": 4.7806859817195425e-05,
+      "loss": 1.5511,
+      "step": 633
+    },
+    {
+      "epoch": 0.9345863276211535,
+      "grad_norm": 1.5774950337520721,
+      "learning_rate": 4.779874573821814e-05,
+      "loss": 1.3597,
+      "step": 634
+    },
+    {
+      "epoch": 0.9360604385480007,
+      "grad_norm": 1.6792695700996014,
+      "learning_rate": 4.779061736808514e-05,
+      "loss": 1.4927,
+      "step": 635
+    },
+    {
+      "epoch": 0.9375345494748479,
+      "grad_norm": 1.6897696314344495,
+      "learning_rate": 4.778247471189163e-05,
+      "loss": 1.3959,
+      "step": 636
+    },
+    {
+      "epoch": 0.9390086604016952,
+      "grad_norm": 1.6187650201166919,
+      "learning_rate": 4.777431777474174e-05,
+      "loss": 1.4868,
+      "step": 637
+    },
+    {
+      "epoch": 0.9404827713285425,
+      "grad_norm": 1.505513944477025,
+      "learning_rate": 4.776614656174856e-05,
+      "loss": 1.3534,
+      "step": 638
+    },
+    {
+      "epoch": 0.9419568822553898,
+      "grad_norm": 1.4920532805310136,
+      "learning_rate": 4.775796107803413e-05,
+      "loss": 1.4877,
+      "step": 639
+    },
+    {
+      "epoch": 0.943430993182237,
+      "grad_norm": 1.789707650546946,
+      "learning_rate": 4.7749761328729436e-05,
+      "loss": 1.632,
+      "step": 640
+    },
+    {
+      "epoch": 0.9449051041090842,
+      "grad_norm": 1.5988219096316705,
+      "learning_rate": 4.77415473189744e-05,
+      "loss": 1.4742,
+      "step": 641
+    },
+    {
+      "epoch": 0.9463792150359315,
+      "grad_norm": 1.5586449942135832,
+      "learning_rate": 4.77333190539179e-05,
+      "loss": 1.4758,
+      "step": 642
+    },
+    {
+      "epoch": 0.9478533259627787,
+      "grad_norm": 1.5988251902993467,
+      "learning_rate": 4.772507653871773e-05,
+      "loss": 1.3524,
+      "step": 643
+    },
+    {
+      "epoch": 0.949327436889626,
+      "grad_norm": 1.7705209419112953,
+      "learning_rate": 4.7716819778540625e-05,
+      "loss": 1.6272,
+      "step": 644
+    },
+    {
+      "epoch": 0.9508015478164732,
+      "grad_norm": 1.660636706671694,
+      "learning_rate": 4.770854877856225e-05,
+      "loss": 1.4897,
+      "step": 645
+    },
+    {
+      "epoch": 0.9522756587433204,
+      "grad_norm": 1.9015811590127132,
+      "learning_rate": 4.7700263543967195e-05,
+      "loss": 1.4256,
+      "step": 646
+    },
+    {
+      "epoch": 0.9537497696701677,
+      "grad_norm": 1.7306241553384767,
+      "learning_rate": 4.769196407994898e-05,
+      "loss": 1.4095,
+      "step": 647
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 1.6024022575718655,
+      "learning_rate": 4.768365039171002e-05,
+      "loss": 1.5108,
+      "step": 648
+    },
+    {
+      "epoch": 0.9566979915238621,
+      "grad_norm": 1.8686845595725392,
+      "learning_rate": 4.7675322484461674e-05,
+      "loss": 1.5385,
+      "step": 649
+    },
+    {
+      "epoch": 0.9581721024507094,
+      "grad_norm": 1.5783638348530795,
+      "learning_rate": 4.766698036342421e-05,
+      "loss": 1.3522,
+      "step": 650
+    },
+    {
+      "epoch": 0.9596462133775566,
+      "grad_norm": 1.7358905940409899,
+      "learning_rate": 4.765862403382678e-05,
+      "loss": 1.4563,
+      "step": 651
+    },
+    {
+      "epoch": 0.9611203243044039,
+      "grad_norm": 1.6868556169864075,
+      "learning_rate": 4.7650253500907494e-05,
+      "loss": 1.5699,
+      "step": 652
+    },
+    {
+      "epoch": 0.9625944352312511,
+      "grad_norm": 1.576506209337351,
+      "learning_rate": 4.76418687699133e-05,
+      "loss": 1.6541,
+      "step": 653
+    },
+    {
+      "epoch": 0.9640685461580984,
+      "grad_norm": 1.970339239606081,
+      "learning_rate": 4.76334698461001e-05,
+      "loss": 1.4497,
+      "step": 654
+    },
+    {
+      "epoch": 0.9655426570849457,
+      "grad_norm": 1.7663451894931055,
+      "learning_rate": 4.7625056734732654e-05,
+      "loss": 1.417,
+      "step": 655
+    },
+    {
+      "epoch": 0.9670167680117929,
+      "grad_norm": 1.8093780069540335,
+      "learning_rate": 4.7616629441084655e-05,
+      "loss": 1.468,
+      "step": 656
+    },
+    {
+      "epoch": 0.9684908789386402,
+      "grad_norm": 1.506355886575976,
+      "learning_rate": 4.760818797043864e-05,
+      "loss": 1.5133,
+      "step": 657
+    },
+    {
+      "epoch": 0.9699649898654874,
+      "grad_norm": 1.665298432763018,
+      "learning_rate": 4.759973232808609e-05,
+      "loss": 1.474,
+      "step": 658
+    },
+    {
+      "epoch": 0.9714391007923346,
+      "grad_norm": 1.5021839138152469,
+      "learning_rate": 4.75912625193273e-05,
+      "loss": 1.4184,
+      "step": 659
+    },
+    {
+      "epoch": 0.9729132117191819,
+      "grad_norm": 1.7473590541859465,
+      "learning_rate": 4.7582778549471494e-05,
+      "loss": 1.449,
+      "step": 660
+    },
+    {
+      "epoch": 0.9743873226460291,
+      "grad_norm": 1.5233425012948403,
+      "learning_rate": 4.7574280423836776e-05,
+      "loss": 1.2742,
+      "step": 661
+    },
+    {
+      "epoch": 0.9758614335728764,
+      "grad_norm": 1.7013333196929006,
+      "learning_rate": 4.756576814775009e-05,
+      "loss": 1.5258,
+      "step": 662
+    },
+    {
+      "epoch": 0.9773355444997236,
+      "grad_norm": 1.6033550372541576,
+      "learning_rate": 4.7557241726547266e-05,
+      "loss": 1.4548,
+      "step": 663
+    },
+    {
+      "epoch": 0.9788096554265708,
+      "grad_norm": 1.6487359956344683,
+      "learning_rate": 4.7548701165573003e-05,
+      "loss": 1.3467,
+      "step": 664
+    },
+    {
+      "epoch": 0.9802837663534181,
+      "grad_norm": 1.648531364191147,
+      "learning_rate": 4.754014647018088e-05,
+      "loss": 1.5249,
+      "step": 665
+    },
+    {
+      "epoch": 0.9817578772802653,
+      "grad_norm": 1.527489228405722,
+      "learning_rate": 4.75315776457333e-05,
+      "loss": 1.2307,
+      "step": 666
+    },
+    {
+      "epoch": 0.9832319882071126,
+      "grad_norm": 1.6423498623305697,
+      "learning_rate": 4.752299469760154e-05,
+      "loss": 1.4412,
+      "step": 667
+    },
+    {
+      "epoch": 0.9847060991339598,
+      "grad_norm": 1.608062579370135,
+      "learning_rate": 4.751439763116575e-05,
+      "loss": 1.5007,
+      "step": 668
+    },
+    {
+      "epoch": 0.986180210060807,
+      "grad_norm": 1.693547609068286,
+      "learning_rate": 4.750578645181489e-05,
+      "loss": 1.523,
+      "step": 669
+    },
+    {
+      "epoch": 0.9876543209876543,
+      "grad_norm": 1.6393169456496408,
+      "learning_rate": 4.74971611649468e-05,
+      "loss": 1.5944,
+      "step": 670
+    },
+    {
+      "epoch": 0.9891284319145016,
+      "grad_norm": 1.5490394630897009,
+      "learning_rate": 4.748852177596815e-05,
+      "loss": 1.4744,
+      "step": 671
+    },
+    {
+      "epoch": 0.9906025428413489,
+      "grad_norm": 1.7242987105613135,
+      "learning_rate": 4.747986829029445e-05,
+      "loss": 1.6405,
+      "step": 672
+    },
+    {
+      "epoch": 0.9920766537681961,
+      "grad_norm": 1.6003581659453179,
+      "learning_rate": 4.747120071335004e-05,
+      "loss": 1.5092,
+      "step": 673
+    },
+    {
+      "epoch": 0.9935507646950433,
+      "grad_norm": 1.7075237899173015,
+      "learning_rate": 4.746251905056811e-05,
+      "loss": 1.4837,
+      "step": 674
+    },
+    {
+      "epoch": 0.9950248756218906,
+      "grad_norm": 1.6254956149860846,
+      "learning_rate": 4.745382330739067e-05,
+      "loss": 1.5699,
+      "step": 675
+    },
+    {
+      "epoch": 0.9964989865487378,
+      "grad_norm": 1.6152010156583876,
+      "learning_rate": 4.7445113489268544e-05,
+      "loss": 1.3392,
+      "step": 676
+    },
+    {
+      "epoch": 0.997973097475585,
+      "grad_norm": 1.4897348109055284,
+      "learning_rate": 4.74363896016614e-05,
+      "loss": 1.4951,
+      "step": 677
+    },
+    {
+      "epoch": 0.9994472084024323,
+      "grad_norm": 1.6969560663519077,
+      "learning_rate": 4.742765165003772e-05,
+      "loss": 1.4857,
+      "step": 678
+    },
+    {
+      "epoch": 1.0009213193292796,
+      "grad_norm": 1.6645701175908278,
+      "learning_rate": 4.741889963987478e-05,
+      "loss": 1.4045,
+      "step": 679
+    },
+    {
+      "epoch": 1.0023954302561269,
+      "grad_norm": 1.6543320400999817,
+      "learning_rate": 4.741013357665871e-05,
+      "loss": 1.1458,
+      "step": 680
+    },
+    {
+      "epoch": 1.0038695411829741,
+      "grad_norm": 1.8272377534665492,
+      "learning_rate": 4.7401353465884406e-05,
+      "loss": 1.3011,
+      "step": 681
+    },
+    {
+      "epoch": 1.0053436521098214,
+      "grad_norm": 1.554451162499311,
+      "learning_rate": 4.73925593130556e-05,
+      "loss": 1.1387,
+      "step": 682
+    },
+    {
+      "epoch": 1.0068177630366686,
+      "grad_norm": 2.033797910907898,
+      "learning_rate": 4.7383751123684806e-05,
+      "loss": 1.3314,
+      "step": 683
+    },
+    {
+      "epoch": 1.0082918739635158,
+      "grad_norm": 1.7324135488763774,
+      "learning_rate": 4.737492890329335e-05,
+      "loss": 1.1564,
+      "step": 684
+    },
+    {
+      "epoch": 1.009765984890363,
+      "grad_norm": 1.9036313816118642,
+      "learning_rate": 4.736609265741135e-05,
+      "loss": 1.2494,
+      "step": 685
+    },
+    {
+      "epoch": 1.0112400958172103,
+      "grad_norm": 1.6847992937671599,
+      "learning_rate": 4.7357242391577724e-05,
+      "loss": 1.3732,
+      "step": 686
+    },
+    {
+      "epoch": 1.0127142067440575,
+      "grad_norm": 2.0434491557307917,
+      "learning_rate": 4.7348378111340145e-05,
+      "loss": 1.1515,
+      "step": 687
+    },
+    {
+      "epoch": 1.0141883176709048,
+      "grad_norm": 1.8929767956407253,
+      "learning_rate": 4.733949982225511e-05,
+      "loss": 1.184,
+      "step": 688
+    },
+    {
+      "epoch": 1.015662428597752,
+      "grad_norm": 1.8249296645105988,
+      "learning_rate": 4.7330607529887884e-05,
+      "loss": 1.0749,
+      "step": 689
+    },
+    {
+      "epoch": 1.0171365395245993,
+      "grad_norm": 1.8259922857257385,
+      "learning_rate": 4.73217012398125e-05,
+      "loss": 1.2582,
+      "step": 690
+    },
+    {
+      "epoch": 1.0186106504514465,
+      "grad_norm": 1.9125887432577295,
+      "learning_rate": 4.731278095761178e-05,
+      "loss": 1.2178,
+      "step": 691
+    },
+    {
+      "epoch": 1.0200847613782937,
+      "grad_norm": 1.9581689021561286,
+      "learning_rate": 4.73038466888773e-05,
+      "loss": 1.3432,
+      "step": 692
+    },
+    {
+      "epoch": 1.021558872305141,
+      "grad_norm": 2.0120926296608945,
+      "learning_rate": 4.729489843920942e-05,
+      "loss": 1.1121,
+      "step": 693
+    },
+    {
+      "epoch": 1.0230329832319882,
+      "grad_norm": 1.7744387182476278,
+      "learning_rate": 4.728593621421726e-05,
+      "loss": 1.0938,
+      "step": 694
+    },
+    {
+      "epoch": 1.0245070941588355,
+      "grad_norm": 1.9230047487506763,
+      "learning_rate": 4.727696001951869e-05,
+      "loss": 1.0658,
+      "step": 695
+    },
+    {
+      "epoch": 1.0259812050856827,
+      "grad_norm": 1.7635808268519948,
+      "learning_rate": 4.726796986074034e-05,
+      "loss": 1.0238,
+      "step": 696
+    },
+    {
+      "epoch": 1.02745531601253,
+      "grad_norm": 1.8424304831031615,
+      "learning_rate": 4.725896574351763e-05,
+      "loss": 1.1836,
+      "step": 697
+    },
+    {
+      "epoch": 1.0289294269393772,
+      "grad_norm": 1.9853839862238,
+      "learning_rate": 4.7249947673494645e-05,
+      "loss": 1.2393,
+      "step": 698
+    },
+    {
+      "epoch": 1.0304035378662244,
+      "grad_norm": 1.8546668317457866,
+      "learning_rate": 4.72409156563243e-05,
+      "loss": 1.068,
+      "step": 699
+    },
+    {
+      "epoch": 1.0318776487930716,
+      "grad_norm": 1.9263237427019861,
+      "learning_rate": 4.7231869697668214e-05,
+      "loss": 1.2192,
+      "step": 700
+    },
+    {
+      "epoch": 1.0333517597199189,
+      "grad_norm": 1.8961694929400221,
+      "learning_rate": 4.722280980319675e-05,
+      "loss": 1.1651,
+      "step": 701
+    },
+    {
+      "epoch": 1.0348258706467661,
+      "grad_norm": 1.8958722274758995,
+      "learning_rate": 4.7213735978589016e-05,
+      "loss": 1.086,
+      "step": 702
+    },
+    {
+      "epoch": 1.0362999815736134,
+      "grad_norm": 1.920707968581033,
+      "learning_rate": 4.720464822953284e-05,
+      "loss": 1.3001,
+      "step": 703
+    },
+    {
+      "epoch": 1.0377740925004606,
+      "grad_norm": 2.033207463709648,
+      "learning_rate": 4.719554656172478e-05,
+      "loss": 1.3512,
+      "step": 704
+    },
+    {
+      "epoch": 1.0392482034273078,
+      "grad_norm": 1.893497334359798,
+      "learning_rate": 4.7186430980870124e-05,
+      "loss": 1.2213,
+      "step": 705
+    },
+    {
+      "epoch": 1.040722314354155,
+      "grad_norm": 1.9565062894303316,
+      "learning_rate": 4.717730149268287e-05,
+      "loss": 1.115,
+      "step": 706
+    },
+    {
+      "epoch": 1.0421964252810023,
+      "grad_norm": 1.9193517356463834,
+      "learning_rate": 4.716815810288575e-05,
+      "loss": 1.2553,
+      "step": 707
+    },
+    {
+      "epoch": 1.0436705362078496,
+      "grad_norm": 2.0806191757088546,
+      "learning_rate": 4.7159000817210205e-05,
+      "loss": 1.2919,
+      "step": 708
+    },
+    {
+      "epoch": 1.0451446471346968,
+      "grad_norm": 1.9243055651615135,
+      "learning_rate": 4.714982964139639e-05,
+      "loss": 1.1681,
+      "step": 709
+    },
+    {
+      "epoch": 1.046618758061544,
+      "grad_norm": 2.02960584562886,
+      "learning_rate": 4.714064458119314e-05,
+      "loss": 1.1467,
+      "step": 710
+    },
+    {
+      "epoch": 1.0480928689883915,
+      "grad_norm": 2.3131263503261645,
+      "learning_rate": 4.713144564235803e-05,
+      "loss": 1.1478,
+      "step": 711
+    },
+    {
+      "epoch": 1.0495669799152387,
+      "grad_norm": 1.804795034928628,
+      "learning_rate": 4.7122232830657315e-05,
+      "loss": 1.1921,
+      "step": 712
+    },
+    {
+      "epoch": 1.051041090842086,
+      "grad_norm": 1.7319013120978317,
+      "learning_rate": 4.7113006151865944e-05,
+      "loss": 1.1302,
+      "step": 713
+    },
+    {
+      "epoch": 1.0525152017689332,
+      "grad_norm": 1.7901973539915657,
+      "learning_rate": 4.710376561176758e-05,
+      "loss": 1.1353,
+      "step": 714
+    },
+    {
+      "epoch": 1.0539893126957804,
+      "grad_norm": 1.87534379986065,
+      "learning_rate": 4.7094511216154546e-05,
+      "loss": 1.1556,
+      "step": 715
+    },
+    {
+      "epoch": 1.0554634236226277,
+      "grad_norm": 1.8132493015878601,
+      "learning_rate": 4.708524297082786e-05,
+      "loss": 1.1175,
+      "step": 716
+    },
+    {
+      "epoch": 1.056937534549475,
+      "grad_norm": 1.9548926327386598,
+      "learning_rate": 4.7075960881597236e-05,
+      "loss": 1.3925,
+      "step": 717
+    },
+    {
+      "epoch": 1.0584116454763222,
+      "grad_norm": 1.9201881014553503,
+      "learning_rate": 4.706666495428105e-05,
+      "loss": 1.1691,
+      "step": 718
+    },
+    {
+      "epoch": 1.0598857564031694,
+      "grad_norm": 2.100252899200614,
+      "learning_rate": 4.705735519470636e-05,
+      "loss": 1.3896,
+      "step": 719
+    },
+    {
+      "epoch": 1.0613598673300166,
+      "grad_norm": 1.897210057940877,
+      "learning_rate": 4.7048031608708876e-05,
+      "loss": 1.2165,
+      "step": 720
+    },
+    {
+      "epoch": 1.0628339782568639,
+      "grad_norm": 1.797016702123261,
+      "learning_rate": 4.703869420213301e-05,
+      "loss": 1.0157,
+      "step": 721
+    },
+    {
+      "epoch": 1.0643080891837111,
+      "grad_norm": 1.954950075024271,
+      "learning_rate": 4.702934298083181e-05,
+      "loss": 1.0934,
+      "step": 722
+    },
+    {
+      "epoch": 1.0657822001105584,
+      "grad_norm": 1.8716187825564312,
+      "learning_rate": 4.701997795066699e-05,
+      "loss": 1.4161,
+      "step": 723
+    },
+    {
+      "epoch": 1.0672563110374056,
+      "grad_norm": 1.891808194763231,
+      "learning_rate": 4.701059911750893e-05,
+      "loss": 1.1178,
+      "step": 724
+    },
+    {
+      "epoch": 1.0687304219642528,
+      "grad_norm": 1.7294349079350462,
+      "learning_rate": 4.7001206487236644e-05,
+      "loss": 1.0219,
+      "step": 725
+    },
+    {
+      "epoch": 1.0702045328911,
+      "grad_norm": 1.9835866127922441,
+      "learning_rate": 4.69918000657378e-05,
+      "loss": 1.1748,
+      "step": 726
+    },
+    {
+      "epoch": 1.0716786438179473,
+      "grad_norm": 1.8108946168971511,
+      "learning_rate": 4.698237985890873e-05,
+      "loss": 1.2905,
+      "step": 727
+    },
+    {
+      "epoch": 1.0731527547447945,
+      "grad_norm": 2.110975322483375,
+      "learning_rate": 4.697294587265438e-05,
+      "loss": 1.1964,
+      "step": 728
+    },
+    {
+      "epoch": 1.0746268656716418,
+      "grad_norm": 1.8274964776376994,
+      "learning_rate": 4.696349811288836e-05,
+      "loss": 1.0506,
+      "step": 729
+    },
+    {
+      "epoch": 1.076100976598489,
+      "grad_norm": 1.78960567041504,
+      "learning_rate": 4.695403658553288e-05,
+      "loss": 1.3826,
+      "step": 730
+    },
+    {
+      "epoch": 1.0775750875253363,
+      "grad_norm": 1.7964804838220931,
+      "learning_rate": 4.6944561296518816e-05,
+      "loss": 1.0888,
+      "step": 731
+    },
+    {
+      "epoch": 1.0790491984521835,
+      "grad_norm": 2.012961113271312,
+      "learning_rate": 4.693507225178564e-05,
+      "loss": 1.175,
+      "step": 732
+    },
+    {
+      "epoch": 1.0805233093790307,
+      "grad_norm": 1.7533405300664726,
+      "learning_rate": 4.692556945728147e-05,
+      "loss": 1.0948,
+      "step": 733
+    },
+    {
+      "epoch": 1.081997420305878,
+      "grad_norm": 1.969127679353585,
+      "learning_rate": 4.691605291896304e-05,
+      "loss": 1.1976,
+      "step": 734
+    },
+    {
+      "epoch": 1.0834715312327252,
+      "grad_norm": 1.9246303996197247,
+      "learning_rate": 4.690652264279567e-05,
+      "loss": 1.3948,
+      "step": 735
+    },
+    {
+      "epoch": 1.0849456421595725,
+      "grad_norm": 1.9228475947457793,
+      "learning_rate": 4.689697863475334e-05,
+      "loss": 1.209,
+      "step": 736
+    },
+    {
+      "epoch": 1.0864197530864197,
+      "grad_norm": 1.8442747371545725,
+      "learning_rate": 4.688742090081859e-05,
+      "loss": 1.2617,
+      "step": 737
+    },
+    {
+      "epoch": 1.087893864013267,
+      "grad_norm": 1.7129984248655685,
+      "learning_rate": 4.68778494469826e-05,
+      "loss": 1.0915,
+      "step": 738
+    },
+    {
+      "epoch": 1.0893679749401142,
+      "grad_norm": 1.6886787536184504,
+      "learning_rate": 4.686826427924514e-05,
+      "loss": 1.3351,
+      "step": 739
+    },
+    {
+      "epoch": 1.0908420858669614,
+      "grad_norm": 1.9253643582812818,
+      "learning_rate": 4.685866540361456e-05,
+      "loss": 1.2054,
+      "step": 740
+    },
+    {
+      "epoch": 1.0923161967938086,
+      "grad_norm": 1.833177090258177,
+      "learning_rate": 4.684905282610781e-05,
+      "loss": 1.1527,
+      "step": 741
+    },
+    {
+      "epoch": 1.0937903077206559,
+      "grad_norm": 1.878769962816404,
+      "learning_rate": 4.6839426552750454e-05,
+      "loss": 1.2708,
+      "step": 742
+    },
+    {
+      "epoch": 1.0952644186475031,
+      "grad_norm": 1.9389312748919887,
+      "learning_rate": 4.6829786589576604e-05,
+      "loss": 1.1904,
+      "step": 743
+    },
+    {
+      "epoch": 1.0967385295743504,
+      "grad_norm": 1.7420598141532473,
+      "learning_rate": 4.6820132942628974e-05,
+      "loss": 1.186,
+      "step": 744
+    },
+    {
+      "epoch": 1.0982126405011976,
+      "grad_norm": 1.8555039573139989,
+      "learning_rate": 4.6810465617958856e-05,
+      "loss": 1.1122,
+      "step": 745
+    },
+    {
+      "epoch": 1.099686751428045,
+      "grad_norm": 1.8615491961941468,
+      "learning_rate": 4.680078462162611e-05,
+      "loss": 1.2685,
+      "step": 746
+    },
+    {
+      "epoch": 1.1011608623548923,
+      "grad_norm": 2.0443377894878516,
+      "learning_rate": 4.679108995969917e-05,
+      "loss": 1.3216,
+      "step": 747
+    },
+    {
+      "epoch": 1.1026349732817395,
+      "grad_norm": 1.821300320424321,
+      "learning_rate": 4.678138163825503e-05,
+      "loss": 1.2258,
+      "step": 748
+    },
+    {
+      "epoch": 1.1041090842085868,
+      "grad_norm": 1.9161189029090677,
+      "learning_rate": 4.677165966337924e-05,
+      "loss": 1.3147,
+      "step": 749
+    },
+    {
+      "epoch": 1.105583195135434,
+      "grad_norm": 1.9361667968328764,
+      "learning_rate": 4.676192404116594e-05,
+      "loss": 1.1274,
+      "step": 750
+    },
+    {
+      "epoch": 1.105583195135434,
+      "eval_bleu": 0.04832359604321219,
+      "eval_bleu_1gram": 0.3489221720041889,
+      "eval_bleu_2gram": 0.1226059538293485,
+      "eval_bleu_3gram": 0.048325292345045893,
+      "eval_bleu_4gram": 0.022483994131201883,
+      "eval_rag_val_loss": 1.5553608203446994,
+      "eval_rouge1": 0.3373041450590792,
+      "eval_rouge2": 0.11489670979442308,
+      "eval_rougeL": 0.31697945556853585,
+      "step": 750
+    },
+    {
+      "epoch": 1.1070573060622813,
+      "grad_norm": 2.059973469720756,
+      "learning_rate": 4.6752174777717786e-05,
+      "loss": 1.2282,
+      "step": 751
+    },
+    {
+      "epoch": 1.1085314169891285,
+      "grad_norm": 1.9499485889406722,
+      "learning_rate": 4.674241187914601e-05,
+      "loss": 0.99,
+      "step": 752
+    },
+    {
+      "epoch": 1.1100055279159757,
+      "grad_norm": 1.960007283528088,
+      "learning_rate": 4.673263535157038e-05,
+      "loss": 1.1945,
+      "step": 753
+    },
+    {
+      "epoch": 1.111479638842823,
+      "grad_norm": 2.15022541904847,
+      "learning_rate": 4.6722845201119214e-05,
+      "loss": 1.2765,
+      "step": 754
+    },
+    {
+      "epoch": 1.1129537497696702,
+      "grad_norm": 2.098606228382695,
+      "learning_rate": 4.671304143392936e-05,
+      "loss": 1.2809,
+      "step": 755
+    },
+    {
+      "epoch": 1.1144278606965174,
+      "grad_norm": 2.1353745495124086,
+      "learning_rate": 4.670322405614621e-05,
+      "loss": 1.2348,
+      "step": 756
+    },
+    {
+      "epoch": 1.1159019716233647,
+      "grad_norm": 1.8928509054055485,
+      "learning_rate": 4.6693393073923686e-05,
+      "loss": 1.2209,
+      "step": 757
+    },
+    {
+      "epoch": 1.117376082550212,
+      "grad_norm": 1.8769153984160498,
+      "learning_rate": 4.6683548493424236e-05,
+      "loss": 1.1185,
+      "step": 758
+    },
+    {
+      "epoch": 1.1188501934770592,
+      "grad_norm": 1.861674257580004,
+      "learning_rate": 4.667369032081883e-05,
+      "loss": 1.2134,
+      "step": 759
+    },
+    {
+      "epoch": 1.1203243044039064,
+      "grad_norm": 1.9099952302244056,
+      "learning_rate": 4.666381856228697e-05,
+      "loss": 1.2536,
+      "step": 760
+    },
+    {
+      "epoch": 1.1217984153307536,
+      "grad_norm": 1.9062018466558288,
+      "learning_rate": 4.665393322401664e-05,
+      "loss": 1.411,
+      "step": 761
+    },
+    {
+      "epoch": 1.1232725262576009,
+      "grad_norm": 1.9084961587109184,
+      "learning_rate": 4.6644034312204387e-05,
+      "loss": 1.0589,
+      "step": 762
+    },
+    {
+      "epoch": 1.1247466371844481,
+      "grad_norm": 1.9482587082241267,
+      "learning_rate": 4.6634121833055235e-05,
+      "loss": 1.1527,
+      "step": 763
+    },
+    {
+      "epoch": 1.1262207481112954,
+      "grad_norm": 1.9062931493659574,
+      "learning_rate": 4.662419579278271e-05,
+      "loss": 1.2479,
+      "step": 764
+    },
+    {
+      "epoch": 1.1276948590381426,
+      "grad_norm": 1.9283854304040413,
+      "learning_rate": 4.6614256197608855e-05,
+      "loss": 1.2503,
+      "step": 765
+    },
+    {
+      "epoch": 1.1291689699649898,
+      "grad_norm": 1.9222112842143262,
+      "learning_rate": 4.660430305376419e-05,
+      "loss": 1.323,
+      "step": 766
+    },
+    {
+      "epoch": 1.130643080891837,
+      "grad_norm": 1.8120554674850413,
+      "learning_rate": 4.659433636748775e-05,
+      "loss": 1.3234,
+      "step": 767
+    },
+    {
+      "epoch": 1.1321171918186843,
+      "grad_norm": 1.71110610479528,
+      "learning_rate": 4.658435614502705e-05,
+      "loss": 1.0505,
+      "step": 768
+    },
+    {
+      "epoch": 1.1335913027455315,
+      "grad_norm": 2.0124255670086892,
+      "learning_rate": 4.657436239263808e-05,
+      "loss": 1.177,
+      "step": 769
+    },
+    {
+      "epoch": 1.1350654136723788,
+      "grad_norm": 1.8902943219499422,
+      "learning_rate": 4.6564355116585325e-05,
+      "loss": 1.3793,
+      "step": 770
+    },
+    {
+      "epoch": 1.136539524599226,
+      "grad_norm": 1.7362883727390206,
+      "learning_rate": 4.655433432314174e-05,
+      "loss": 1.0214,
+      "step": 771
+    },
+    {
+      "epoch": 1.1380136355260733,
+      "grad_norm": 1.8006550047472214,
+      "learning_rate": 4.654430001858874e-05,
+      "loss": 1.1815,
+      "step": 772
+    },
+    {
+      "epoch": 1.1394877464529205,
+      "grad_norm": 1.9270018139342167,
+      "learning_rate": 4.653425220921626e-05,
+      "loss": 1.3309,
+      "step": 773
+    },
+    {
+      "epoch": 1.1409618573797677,
+      "grad_norm": 2.1138547447054794,
+      "learning_rate": 4.6524190901322626e-05,
+      "loss": 1.2242,
+      "step": 774
+    },
+    {
+      "epoch": 1.142435968306615,
+      "grad_norm": 2.1271683344980046,
+      "learning_rate": 4.651411610121469e-05,
+      "loss": 1.2117,
+      "step": 775
+    },
+    {
+      "epoch": 1.1439100792334624,
+      "grad_norm": 2.04938114873831,
+      "learning_rate": 4.650402781520772e-05,
+      "loss": 1.3106,
+      "step": 776
+    },
+    {
+      "epoch": 1.1453841901603097,
+      "grad_norm": 1.8633284652773754,
+      "learning_rate": 4.649392604962546e-05,
+      "loss": 1.2068,
+      "step": 777
+    },
+    {
+      "epoch": 1.146858301087157,
+      "grad_norm": 2.0148619161010983,
+      "learning_rate": 4.648381081080009e-05,
+      "loss": 1.1792,
+      "step": 778
+    },
+    {
+      "epoch": 1.1483324120140042,
+      "grad_norm": 1.9219231948392115,
+      "learning_rate": 4.647368210507225e-05,
+      "loss": 1.319,
+      "step": 779
+    },
+    {
+      "epoch": 1.1498065229408514,
+      "grad_norm": 2.0746685912426517,
+      "learning_rate": 4.6463539938791e-05,
+      "loss": 1.1628,
+      "step": 780
+    },
+    {
+      "epoch": 1.1512806338676986,
+      "grad_norm": 1.9249777829758352,
+      "learning_rate": 4.645338431831388e-05,
+      "loss": 1.1656,
+      "step": 781
+    },
+    {
+      "epoch": 1.1527547447945459,
+      "grad_norm": 1.9588541461106448,
+      "learning_rate": 4.6443215250006806e-05,
+      "loss": 1.1347,
+      "step": 782
+    },
+    {
+      "epoch": 1.154228855721393,
+      "grad_norm": 2.213609462522597,
+      "learning_rate": 4.643303274024416e-05,
+      "loss": 1.241,
+      "step": 783
+    },
+    {
+      "epoch": 1.1557029666482403,
+      "grad_norm": 2.24719720686905,
+      "learning_rate": 4.642283679540874e-05,
+      "loss": 1.3281,
+      "step": 784
+    },
+    {
+      "epoch": 1.1571770775750876,
+      "grad_norm": 2.119060968404964,
+      "learning_rate": 4.641262742189178e-05,
+      "loss": 1.2635,
+      "step": 785
+    },
+    {
+      "epoch": 1.1586511885019348,
+      "grad_norm": 2.0360543581258246,
+      "learning_rate": 4.640240462609291e-05,
+      "loss": 1.3259,
+      "step": 786
+    },
+    {
+      "epoch": 1.160125299428782,
+      "grad_norm": 2.0091196992569613,
+      "learning_rate": 4.639216841442018e-05,
+      "loss": 1.295,
+      "step": 787
+    },
+    {
+      "epoch": 1.1615994103556293,
+      "grad_norm": 2.04665589069839,
+      "learning_rate": 4.6381918793290055e-05,
+      "loss": 1.4083,
+      "step": 788
+    },
+    {
+      "epoch": 1.1630735212824765,
+      "grad_norm": 1.9770604889665038,
+      "learning_rate": 4.6371655769127396e-05,
+      "loss": 1.3114,
+      "step": 789
+    },
+    {
+      "epoch": 1.1645476322093238,
+      "grad_norm": 2.0549801230107407,
+      "learning_rate": 4.63613793483655e-05,
+      "loss": 1.2201,
+      "step": 790
+    },
+    {
+      "epoch": 1.166021743136171,
+      "grad_norm": 2.0593650574509255,
+      "learning_rate": 4.6351089537446e-05,
+      "loss": 1.3657,
+      "step": 791
+    },
+    {
+      "epoch": 1.1674958540630183,
+      "grad_norm": 2.0118875080915544,
+      "learning_rate": 4.6340786342818964e-05,
+      "loss": 1.3112,
+      "step": 792
+    },
+    {
+      "epoch": 1.1689699649898655,
+      "grad_norm": 1.852055994896926,
+      "learning_rate": 4.633046977094286e-05,
+      "loss": 1.232,
+      "step": 793
+    },
+    {
+      "epoch": 1.1704440759167127,
+      "grad_norm": 1.8902511227002372,
+      "learning_rate": 4.632013982828451e-05,
+      "loss": 1.3732,
+      "step": 794
+    },
+    {
+      "epoch": 1.17191818684356,
+      "grad_norm": 1.8677073756518854,
+      "learning_rate": 4.630979652131913e-05,
+      "loss": 1.1752,
+      "step": 795
+    },
+    {
+      "epoch": 1.1733922977704072,
+      "grad_norm": 1.7973951830697428,
+      "learning_rate": 4.629943985653032e-05,
+      "loss": 1.3248,
+      "step": 796
+    },
+    {
+      "epoch": 1.1748664086972544,
+      "grad_norm": 1.838999171693998,
+      "learning_rate": 4.6289069840410036e-05,
+      "loss": 1.2396,
+      "step": 797
+    },
+    {
+      "epoch": 1.1763405196241017,
+      "grad_norm": 1.9341370715115471,
+      "learning_rate": 4.627868647945863e-05,
+      "loss": 1.0613,
+      "step": 798
+    },
+    {
+      "epoch": 1.177814630550949,
+      "grad_norm": 1.7367682226413044,
+      "learning_rate": 4.62682897801848e-05,
+      "loss": 0.9865,
+      "step": 799
+    },
+    {
+      "epoch": 1.1792887414777962,
+      "grad_norm": 2.019427593824958,
+      "learning_rate": 4.625787974910559e-05,
+      "loss": 1.196,
+      "step": 800
+    },
+    {
+      "epoch": 1.1807628524046434,
+      "grad_norm": 1.945706614820701,
+      "learning_rate": 4.6247456392746444e-05,
+      "loss": 1.2637,
+      "step": 801
+    },
+    {
+      "epoch": 1.1822369633314906,
+      "grad_norm": 1.7717963872354054,
+      "learning_rate": 4.623701971764112e-05,
+      "loss": 1.1619,
+      "step": 802
+    },
+    {
+      "epoch": 1.1837110742583379,
+      "grad_norm": 1.92765521937285,
+      "learning_rate": 4.622656973033174e-05,
+      "loss": 1.0535,
+      "step": 803
+    },
+    {
+      "epoch": 1.1851851851851851,
+      "grad_norm": 1.7733314247168879,
+      "learning_rate": 4.621610643736878e-05,
+      "loss": 1.146,
+      "step": 804
+    },
+    {
+      "epoch": 1.1866592961120324,
+      "grad_norm": 1.7575950657622053,
+      "learning_rate": 4.620562984531103e-05,
+      "loss": 1.1746,
+      "step": 805
+    },
+    {
+      "epoch": 1.1881334070388796,
+      "grad_norm": 1.8443727249860549,
+      "learning_rate": 4.619513996072564e-05,
+      "loss": 1.2879,
+      "step": 806
+    },
+    {
+      "epoch": 1.1896075179657268,
+      "grad_norm": 1.8136864429277426,
+      "learning_rate": 4.618463679018808e-05,
+      "loss": 1.3109,
+      "step": 807
+    },
+    {
+      "epoch": 1.191081628892574,
+      "grad_norm": 1.9022187303511535,
+      "learning_rate": 4.617412034028217e-05,
+      "loss": 1.242,
+      "step": 808
+    },
+    {
+      "epoch": 1.1925557398194213,
+      "grad_norm": 1.7985484416293065,
+      "learning_rate": 4.616359061760001e-05,
+      "loss": 1.2317,
+      "step": 809
+    },
+    {
+      "epoch": 1.1940298507462686,
+      "grad_norm": 1.964728168098802,
+      "learning_rate": 4.6153047628742066e-05,
+      "loss": 1.1521,
+      "step": 810
+    },
+    {
+      "epoch": 1.1955039616731158,
+      "grad_norm": 1.9894042675131691,
+      "learning_rate": 4.61424913803171e-05,
+      "loss": 1.216,
+      "step": 811
+    },
+    {
+      "epoch": 1.196978072599963,
+      "grad_norm": 1.9904342416241996,
+      "learning_rate": 4.613192187894218e-05,
+      "loss": 1.3258,
+      "step": 812
+    },
+    {
+      "epoch": 1.1984521835268105,
+      "grad_norm": 2.06252612473097,
+      "learning_rate": 4.612133913124268e-05,
+      "loss": 1.2279,
+      "step": 813
+    },
+    {
+      "epoch": 1.1999262944536577,
+      "grad_norm": 2.0443174968342244,
+      "learning_rate": 4.61107431438523e-05,
+      "loss": 1.3923,
+      "step": 814
+    },
+    {
+      "epoch": 1.201400405380505,
+      "grad_norm": 2.037292647781029,
+      "learning_rate": 4.610013392341301e-05,
+      "loss": 1.1695,
+      "step": 815
+    },
+    {
+      "epoch": 1.2028745163073522,
+      "grad_norm": 1.9747739396914579,
+      "learning_rate": 4.608951147657511e-05,
+      "loss": 1.2224,
+      "step": 816
+    },
+    {
+      "epoch": 1.2043486272341994,
+      "grad_norm": 1.928814648121447,
+      "learning_rate": 4.607887580999715e-05,
+      "loss": 1.3066,
+      "step": 817
+    },
+    {
+      "epoch": 1.2058227381610467,
+      "grad_norm": 1.7628306328987635,
+      "learning_rate": 4.6068226930345995e-05,
+      "loss": 1.0662,
+      "step": 818
+    },
+    {
+      "epoch": 1.207296849087894,
+      "grad_norm": 1.8021481305712592,
+      "learning_rate": 4.605756484429678e-05,
+      "loss": 1.2606,
+      "step": 819
+    },
+    {
+      "epoch": 1.2087709600147412,
+      "grad_norm": 2.047069744135142,
+      "learning_rate": 4.604688955853293e-05,
+      "loss": 1.1361,
+      "step": 820
+    },
+    {
+      "epoch": 1.2102450709415884,
+      "grad_norm": 1.8754182349097253,
+      "learning_rate": 4.603620107974612e-05,
+      "loss": 1.1022,
+      "step": 821
+    },
+    {
+      "epoch": 1.2117191818684356,
+      "grad_norm": 1.8822805832531442,
+      "learning_rate": 4.602549941463633e-05,
+      "loss": 1.2389,
+      "step": 822
+    },
+    {
+      "epoch": 1.2131932927952829,
+      "grad_norm": 1.8911851415713685,
+      "learning_rate": 4.601478456991178e-05,
+      "loss": 1.257,
+      "step": 823
+    },
+    {
+      "epoch": 1.2146674037221301,
+      "grad_norm": 1.8257667785776173,
+      "learning_rate": 4.6004056552288956e-05,
+      "loss": 1.0839,
+      "step": 824
+    },
+    {
+      "epoch": 1.2161415146489774,
+      "grad_norm": 1.9537110937031328,
+      "learning_rate": 4.5993315368492603e-05,
+      "loss": 1.0454,
+      "step": 825
+    },
+    {
+      "epoch": 1.2176156255758246,
+      "grad_norm": 1.7445804600964923,
+      "learning_rate": 4.5982561025255726e-05,
+      "loss": 1.2837,
+      "step": 826
+    },
+    {
+      "epoch": 1.2190897365026718,
+      "grad_norm": 1.9185719211708803,
+      "learning_rate": 4.5971793529319576e-05,
+      "loss": 1.159,
+      "step": 827
+    },
+    {
+      "epoch": 1.220563847429519,
+      "grad_norm": 1.8274022165170416,
+      "learning_rate": 4.596101288743362e-05,
+      "loss": 1.2292,
+      "step": 828
+    },
+    {
+      "epoch": 1.2220379583563663,
+      "grad_norm": 2.229485985057623,
+      "learning_rate": 4.595021910635563e-05,
+      "loss": 1.3312,
+      "step": 829
+    },
+    {
+      "epoch": 1.2235120692832135,
+      "grad_norm": 2.0491974445941397,
+      "learning_rate": 4.5939412192851535e-05,
+      "loss": 1.2012,
+      "step": 830
+    },
+    {
+      "epoch": 1.2249861802100608,
+      "grad_norm": 2.0764561886408117,
+      "learning_rate": 4.592859215369557e-05,
+      "loss": 1.2366,
+      "step": 831
+    },
+    {
+      "epoch": 1.226460291136908,
+      "grad_norm": 2.0749286363578348,
+      "learning_rate": 4.591775899567015e-05,
+      "loss": 1.3437,
+      "step": 832
+    },
+    {
+      "epoch": 1.2279344020637553,
+      "grad_norm": 2.0806212383315326,
+      "learning_rate": 4.590691272556592e-05,
+      "loss": 1.0536,
+      "step": 833
+    },
+    {
+      "epoch": 1.2294085129906025,
+      "grad_norm": 2.138625052997736,
+      "learning_rate": 4.589605335018176e-05,
+      "loss": 1.2359,
+      "step": 834
+    },
+    {
+      "epoch": 1.2308826239174497,
+      "grad_norm": 2.017168267451644,
+      "learning_rate": 4.588518087632475e-05,
+      "loss": 1.3051,
+      "step": 835
+    },
+    {
+      "epoch": 1.232356734844297,
+      "grad_norm": 1.9112500138716264,
+      "learning_rate": 4.587429531081019e-05,
+      "loss": 1.1611,
+      "step": 836
+    },
+    {
+      "epoch": 1.2338308457711442,
+      "grad_norm": 1.9438705922095745,
+      "learning_rate": 4.5863396660461575e-05,
+      "loss": 1.2222,
+      "step": 837
+    },
+    {
+      "epoch": 1.2353049566979915,
+      "grad_norm": 1.9247965742353426,
+      "learning_rate": 4.585248493211063e-05,
+      "loss": 1.1865,
+      "step": 838
+    },
+    {
+      "epoch": 1.2367790676248387,
+      "grad_norm": 1.938434560018029,
+      "learning_rate": 4.5841560132597244e-05,
+      "loss": 1.0139,
+      "step": 839
+    },
+    {
+      "epoch": 1.238253178551686,
+      "grad_norm": 1.9023332223463831,
+      "learning_rate": 4.583062226876952e-05,
+      "loss": 1.3509,
+      "step": 840
+    },
+    {
+      "epoch": 1.2397272894785332,
+      "grad_norm": 1.8888812259755496,
+      "learning_rate": 4.5819671347483725e-05,
+      "loss": 1.0203,
+      "step": 841
+    },
+    {
+      "epoch": 1.2412014004053806,
+      "grad_norm": 2.008767935976047,
+      "learning_rate": 4.580870737560435e-05,
+      "loss": 1.3465,
+      "step": 842
+    },
+    {
+      "epoch": 1.2426755113322279,
+      "grad_norm": 1.771548706367048,
+      "learning_rate": 4.579773036000405e-05,
+      "loss": 1.4631,
+      "step": 843
+    },
+    {
+      "epoch": 1.244149622259075,
+      "grad_norm": 2.1056041780134684,
+      "learning_rate": 4.5786740307563636e-05,
+      "loss": 1.1632,
+      "step": 844
+    },
+    {
+      "epoch": 1.2456237331859223,
+      "grad_norm": 1.7614772318121938,
+      "learning_rate": 4.577573722517211e-05,
+      "loss": 1.2406,
+      "step": 845
+    },
+    {
+      "epoch": 1.2470978441127696,
+      "grad_norm": 1.941060824145296,
+      "learning_rate": 4.5764721119726653e-05,
+      "loss": 1.1045,
+      "step": 846
+    },
+    {
+      "epoch": 1.2485719550396168,
+      "grad_norm": 1.9177119749394835,
+      "learning_rate": 4.575369199813258e-05,
+      "loss": 1.3186,
+      "step": 847
+    },
+    {
+      "epoch": 1.250046065966464,
+      "grad_norm": 1.9702454594120657,
+      "learning_rate": 4.5742649867303386e-05,
+      "loss": 1.0918,
+      "step": 848
+    },
+    {
+      "epoch": 1.2515201768933113,
+      "grad_norm": 2.2033457543989172,
+      "learning_rate": 4.573159473416072e-05,
+      "loss": 1.1995,
+      "step": 849
+    },
+    {
+      "epoch": 1.2529942878201585,
+      "grad_norm": 1.9168505234923359,
+      "learning_rate": 4.572052660563437e-05,
+      "loss": 1.2621,
+      "step": 850
+    },
+    {
+      "epoch": 1.2544683987470058,
+      "grad_norm": 1.8475782446494462,
+      "learning_rate": 4.570944548866228e-05,
+      "loss": 1.2181,
+      "step": 851
+    },
+    {
+      "epoch": 1.255942509673853,
+      "grad_norm": 1.8534563254064578,
+      "learning_rate": 4.569835139019054e-05,
+      "loss": 1.2162,
+      "step": 852
+    },
+    {
+      "epoch": 1.2574166206007003,
+      "grad_norm": 1.688457641142536,
+      "learning_rate": 4.5687244317173356e-05,
+      "loss": 1.1438,
+      "step": 853
+    },
+    {
+      "epoch": 1.2588907315275475,
+      "grad_norm": 1.9298995669822576,
+      "learning_rate": 4.567612427657308e-05,
+      "loss": 1.3113,
+      "step": 854
+    },
+    {
+      "epoch": 1.2603648424543947,
+      "grad_norm": 1.877831482761959,
+      "learning_rate": 4.566499127536021e-05,
+      "loss": 1.2634,
+      "step": 855
+    },
+    {
+      "epoch": 1.261838953381242,
+      "grad_norm": 2.169264287736473,
+      "learning_rate": 4.565384532051335e-05,
+      "loss": 1.3957,
+      "step": 856
+    },
+    {
+      "epoch": 1.2633130643080892,
+      "grad_norm": 1.7663846702556634,
+      "learning_rate": 4.56426864190192e-05,
+      "loss": 1.1191,
+      "step": 857
+    },
+    {
+      "epoch": 1.2647871752349364,
+      "grad_norm": 1.8045734848403157,
+      "learning_rate": 4.563151457787263e-05,
+      "loss": 1.4029,
+      "step": 858
+    },
+    {
+      "epoch": 1.2662612861617837,
+      "grad_norm": 2.0040378341453624,
+      "learning_rate": 4.562032980407658e-05,
+      "loss": 1.2542,
+      "step": 859
+    },
+    {
+      "epoch": 1.267735397088631,
+      "grad_norm": 1.7774041134151946,
+      "learning_rate": 4.56091321046421e-05,
+      "loss": 1.0325,
+      "step": 860
+    },
+    {
+      "epoch": 1.2692095080154782,
+      "grad_norm": 1.9178436922219209,
+      "learning_rate": 4.5597921486588366e-05,
+      "loss": 1.2749,
+      "step": 861
+    },
+    {
+      "epoch": 1.2706836189423254,
+      "grad_norm": 2.0261190313034447,
+      "learning_rate": 4.558669795694263e-05,
+      "loss": 1.3292,
+      "step": 862
+    },
+    {
+      "epoch": 1.2721577298691726,
+      "grad_norm": 1.8865470097253247,
+      "learning_rate": 4.557546152274025e-05,
+      "loss": 1.2121,
+      "step": 863
+    },
+    {
+      "epoch": 1.2736318407960199,
+      "grad_norm": 2.0123974886547016,
+      "learning_rate": 4.556421219102466e-05,
+      "loss": 1.1027,
+      "step": 864
+    },
+    {
+      "epoch": 1.2751059517228671,
+      "grad_norm": 1.8472406996692792,
+      "learning_rate": 4.555294996884738e-05,
+      "loss": 1.1813,
+      "step": 865
+    },
+    {
+      "epoch": 1.2765800626497144,
+      "grad_norm": 1.8678430021136798,
+      "learning_rate": 4.5541674863268035e-05,
+      "loss": 1.2206,
+      "step": 866
+    },
+    {
+      "epoch": 1.2780541735765616,
+      "grad_norm": 1.7360457887016547,
+      "learning_rate": 4.553038688135429e-05,
+      "loss": 1.3746,
+      "step": 867
+    },
+    {
+      "epoch": 1.2795282845034088,
+      "grad_norm": 1.8732700632594796,
+      "learning_rate": 4.551908603018191e-05,
+      "loss": 1.1941,
+      "step": 868
+    },
+    {
+      "epoch": 1.281002395430256,
+      "grad_norm": 1.8819306393813404,
+      "learning_rate": 4.5507772316834715e-05,
+      "loss": 1.1475,
+      "step": 869
+    },
+    {
+      "epoch": 1.2824765063571033,
+      "grad_norm": 1.9574279383114594,
+      "learning_rate": 4.549644574840458e-05,
+      "loss": 1.2588,
+      "step": 870
+    },
+    {
+      "epoch": 1.2839506172839505,
+      "grad_norm": 1.7583345443335727,
+      "learning_rate": 4.5485106331991446e-05,
+      "loss": 1.0202,
+      "step": 871
+    },
+    {
+      "epoch": 1.2854247282107978,
+      "grad_norm": 1.8985904584888604,
+      "learning_rate": 4.5473754074703324e-05,
+      "loss": 1.2881,
+      "step": 872
+    },
+    {
+      "epoch": 1.286898839137645,
+      "grad_norm": 1.9169067425955675,
+      "learning_rate": 4.546238898365623e-05,
+      "loss": 1.1355,
+      "step": 873
+    },
+    {
+      "epoch": 1.2883729500644923,
+      "grad_norm": 1.8695511162894336,
+      "learning_rate": 4.545101106597428e-05,
+      "loss": 1.2505,
+      "step": 874
+    },
+    {
+      "epoch": 1.2898470609913395,
+      "grad_norm": 2.2799967628589872,
+      "learning_rate": 4.5439620328789593e-05,
+      "loss": 1.2138,
+      "step": 875
+    },
+    {
+      "epoch": 1.2913211719181867,
+      "grad_norm": 1.8974143202665794,
+      "learning_rate": 4.5428216779242336e-05,
+      "loss": 1.1693,
+      "step": 876
+    },
+    {
+      "epoch": 1.292795282845034,
+      "grad_norm": 1.8682219381186351,
+      "learning_rate": 4.541680042448069e-05,
+      "loss": 1.155,
+      "step": 877
+    },
+    {
+      "epoch": 1.2942693937718812,
+      "grad_norm": 1.9171227727905555,
+      "learning_rate": 4.540537127166089e-05,
+      "loss": 1.1816,
+      "step": 878
+    },
+    {
+      "epoch": 1.2957435046987285,
+      "grad_norm": 2.1015021446665445,
+      "learning_rate": 4.5393929327947195e-05,
+      "loss": 1.1996,
+      "step": 879
+    },
+    {
+      "epoch": 1.2972176156255757,
+      "grad_norm": 1.8735547535706223,
+      "learning_rate": 4.538247460051184e-05,
+      "loss": 1.1889,
+      "step": 880
+    },
+    {
+      "epoch": 1.298691726552423,
+      "grad_norm": 2.198880916686168,
+      "learning_rate": 4.537100709653512e-05,
+      "loss": 1.2868,
+      "step": 881
+    },
+    {
+      "epoch": 1.3001658374792704,
+      "grad_norm": 1.9357337284066887,
+      "learning_rate": 4.535952682320531e-05,
+      "loss": 1.3053,
+      "step": 882
+    },
+    {
+      "epoch": 1.3016399484061176,
+      "grad_norm": 1.9950794248139392,
+      "learning_rate": 4.534803378771871e-05,
+      "loss": 1.1592,
+      "step": 883
+    },
+    {
+      "epoch": 1.3031140593329649,
+      "grad_norm": 2.0115906313496157,
+      "learning_rate": 4.53365279972796e-05,
+      "loss": 1.2114,
+      "step": 884
+    },
+    {
+      "epoch": 1.304588170259812,
+      "grad_norm": 1.9877671811917321,
+      "learning_rate": 4.532500945910026e-05,
+      "loss": 1.3775,
+      "step": 885
+    },
+    {
+      "epoch": 1.3060622811866593,
+      "grad_norm": 2.00708410209064,
+      "learning_rate": 4.5313478180400995e-05,
+      "loss": 1.3223,
+      "step": 886
+    },
+    {
+      "epoch": 1.3075363921135066,
+      "grad_norm": 1.9213141886921883,
+      "learning_rate": 4.530193416841003e-05,
+      "loss": 1.26,
+      "step": 887
+    },
+    {
+      "epoch": 1.3090105030403538,
+      "grad_norm": 2.202929548968211,
+      "learning_rate": 4.529037743036362e-05,
+      "loss": 1.4045,
+      "step": 888
+    },
+    {
+      "epoch": 1.310484613967201,
+      "grad_norm": 2.1333296711214738,
+      "learning_rate": 4.5278807973506e-05,
+      "loss": 1.2379,
+      "step": 889
+    },
+    {
+      "epoch": 1.3119587248940483,
+      "grad_norm": 1.930311835599059,
+      "learning_rate": 4.526722580508934e-05,
+      "loss": 1.1965,
+      "step": 890
+    },
+    {
+      "epoch": 1.3134328358208955,
+      "grad_norm": 2.031047047966508,
+      "learning_rate": 4.525563093237383e-05,
+      "loss": 1.2721,
+      "step": 891
+    },
+    {
+      "epoch": 1.3149069467477428,
+      "grad_norm": 1.9557458002337205,
+      "learning_rate": 4.524402336262756e-05,
+      "loss": 1.188,
+      "step": 892
+    },
+    {
+      "epoch": 1.31638105767459,
+      "grad_norm": 1.8648089338104161,
+      "learning_rate": 4.523240310312664e-05,
+      "loss": 1.2553,
+      "step": 893
+    },
+    {
+      "epoch": 1.3178551686014373,
+      "grad_norm": 1.9969554616271883,
+      "learning_rate": 4.522077016115511e-05,
+      "loss": 1.2092,
+      "step": 894
+    },
+    {
+      "epoch": 1.3193292795282845,
+      "grad_norm": 2.0407491555642694,
+      "learning_rate": 4.520912454400494e-05,
+      "loss": 1.3227,
+      "step": 895
+    },
+    {
+      "epoch": 1.3208033904551317,
+      "grad_norm": 1.833479644254606,
+      "learning_rate": 4.519746625897607e-05,
+      "loss": 1.2718,
+      "step": 896
+    },
+    {
+      "epoch": 1.322277501381979,
+      "grad_norm": 1.9786800455534768,
+      "learning_rate": 4.518579531337638e-05,
+      "loss": 1.4308,
+      "step": 897
+    },
+    {
+      "epoch": 1.3237516123088262,
+      "grad_norm": 2.0547855484344177,
+      "learning_rate": 4.5174111714521685e-05,
+      "loss": 1.212,
+      "step": 898
+    },
+    {
+      "epoch": 1.3252257232356734,
+      "grad_norm": 1.9233808743084766,
+      "learning_rate": 4.516241546973571e-05,
+      "loss": 1.219,
+      "step": 899
+    },
+    {
+      "epoch": 1.3266998341625207,
+      "grad_norm": 1.7080239349180115,
+      "learning_rate": 4.515070658635013e-05,
+      "loss": 1.0944,
+      "step": 900
+    },
+    {
+      "epoch": 1.328173945089368,
+      "grad_norm": 2.0443339408966077,
+      "learning_rate": 4.5138985071704546e-05,
+      "loss": 1.2423,
+      "step": 901
+    },
+    {
+      "epoch": 1.3296480560162152,
+      "grad_norm": 1.7154840130017959,
+      "learning_rate": 4.512725093314645e-05,
+      "loss": 1.1359,
+      "step": 902
+    },
+    {
+      "epoch": 1.3311221669430624,
+      "grad_norm": 2.2840776033328805,
+      "learning_rate": 4.5115504178031285e-05,
+      "loss": 1.4354,
+      "step": 903
+    },
+    {
+      "epoch": 1.3325962778699096,
+      "grad_norm": 2.1892841147970574,
+      "learning_rate": 4.5103744813722374e-05,
+      "loss": 1.294,
+      "step": 904
+    },
+    {
+      "epoch": 1.3340703887967569,
+      "grad_norm": 1.9591072325437795,
+      "learning_rate": 4.509197284759094e-05,
+      "loss": 1.2564,
+      "step": 905
+    },
+    {
+      "epoch": 1.3355444997236043,
+      "grad_norm": 1.9277819904912872,
+      "learning_rate": 4.508018828701612e-05,
+      "loss": 1.1511,
+      "step": 906
+    },
+    {
+      "epoch": 1.3370186106504516,
+      "grad_norm": 1.898052726025306,
+      "learning_rate": 4.506839113938496e-05,
+      "loss": 1.2168,
+      "step": 907
+    },
+    {
+      "epoch": 1.3384927215772988,
+      "grad_norm": 1.8898282224309781,
+      "learning_rate": 4.505658141209237e-05,
+      "loss": 1.209,
+      "step": 908
+    },
+    {
+      "epoch": 1.339966832504146,
+      "grad_norm": 2.1069520951715197,
+      "learning_rate": 4.504475911254115e-05,
+      "loss": 1.3438,
+      "step": 909
+    },
+    {
+      "epoch": 1.3414409434309933,
+      "grad_norm": 2.0220314354495197,
+      "learning_rate": 4.503292424814198e-05,
+      "loss": 1.2104,
+      "step": 910
+    },
+    {
+      "epoch": 1.3429150543578405,
+      "grad_norm": 2.1640977546767326,
+      "learning_rate": 4.502107682631343e-05,
+      "loss": 1.2863,
+      "step": 911
+    },
+    {
+      "epoch": 1.3443891652846878,
+      "grad_norm": 1.9003034299192694,
+      "learning_rate": 4.500921685448193e-05,
+      "loss": 1.2467,
+      "step": 912
+    },
+    {
+      "epoch": 1.345863276211535,
+      "grad_norm": 1.9945559914380864,
+      "learning_rate": 4.499734434008178e-05,
+      "loss": 1.3747,
+      "step": 913
+    },
+    {
+      "epoch": 1.3473373871383822,
+      "grad_norm": 2.2528308438302878,
+      "learning_rate": 4.498545929055515e-05,
+      "loss": 1.2159,
+      "step": 914
+    },
+    {
+      "epoch": 1.3488114980652295,
+      "grad_norm": 2.368546803800168,
+      "learning_rate": 4.497356171335204e-05,
+      "loss": 1.3446,
+      "step": 915
+    },
+    {
+      "epoch": 1.3502856089920767,
+      "grad_norm": 1.9593938701883025,
+      "learning_rate": 4.496165161593035e-05,
+      "loss": 1.3796,
+      "step": 916
+    },
+    {
+      "epoch": 1.351759719918924,
+      "grad_norm": 1.903468745214265,
+      "learning_rate": 4.4949729005755765e-05,
+      "loss": 1.2523,
+      "step": 917
+    },
+    {
+      "epoch": 1.3532338308457712,
+      "grad_norm": 2.134018330074337,
+      "learning_rate": 4.493779389030187e-05,
+      "loss": 1.3163,
+      "step": 918
+    },
+    {
+      "epoch": 1.3547079417726184,
+      "grad_norm": 1.8943798515939838,
+      "learning_rate": 4.492584627705008e-05,
+      "loss": 1.2499,
+      "step": 919
+    },
+    {
+      "epoch": 1.3561820526994657,
+      "grad_norm": 2.088743572670381,
+      "learning_rate": 4.491388617348959e-05,
+      "loss": 1.4958,
+      "step": 920
+    },
+    {
+      "epoch": 1.357656163626313,
+      "grad_norm": 1.8504505279038082,
+      "learning_rate": 4.490191358711751e-05,
+      "loss": 1.2194,
+      "step": 921
+    },
+    {
+      "epoch": 1.3591302745531602,
+      "grad_norm": 2.0754973355586097,
+      "learning_rate": 4.488992852543871e-05,
+      "loss": 1.3853,
+      "step": 922
+    },
+    {
+      "epoch": 1.3606043854800074,
+      "grad_norm": 1.9777146551093516,
+      "learning_rate": 4.4877930995965905e-05,
+      "loss": 1.1802,
+      "step": 923
+    },
+    {
+      "epoch": 1.3620784964068546,
+      "grad_norm": 1.8829521032472305,
+      "learning_rate": 4.486592100621961e-05,
+      "loss": 1.3003,
+      "step": 924
+    },
+    {
+      "epoch": 1.3635526073337019,
+      "grad_norm": 1.8847925688245744,
+      "learning_rate": 4.4853898563728184e-05,
+      "loss": 1.2712,
+      "step": 925
+    },
+    {
+      "epoch": 1.365026718260549,
+      "grad_norm": 1.8728280521372789,
+      "learning_rate": 4.484186367602775e-05,
+      "loss": 1.2012,
+      "step": 926
+    },
+    {
+      "epoch": 1.3665008291873963,
+      "grad_norm": 1.6873998965489518,
+      "learning_rate": 4.482981635066227e-05,
+      "loss": 1.1333,
+      "step": 927
+    },
+    {
+      "epoch": 1.3679749401142436,
+      "grad_norm": 1.948134371037688,
+      "learning_rate": 4.481775659518346e-05,
+      "loss": 1.1469,
+      "step": 928
+    },
+    {
+      "epoch": 1.3694490510410908,
+      "grad_norm": 1.850175426516826,
+      "learning_rate": 4.480568441715086e-05,
+      "loss": 1.2145,
+      "step": 929
+    },
+    {
+      "epoch": 1.370923161967938,
+      "grad_norm": 2.0001653364505714,
+      "learning_rate": 4.479359982413181e-05,
+      "loss": 1.384,
+      "step": 930
+    },
+    {
+      "epoch": 1.3723972728947853,
+      "grad_norm": 1.9367454659303158,
+      "learning_rate": 4.478150282370138e-05,
+      "loss": 1.1829,
+      "step": 931
+    },
+    {
+      "epoch": 1.3738713838216325,
+      "grad_norm": 1.901426372184927,
+      "learning_rate": 4.476939342344246e-05,
+      "loss": 1.1795,
+      "step": 932
+    },
+    {
+      "epoch": 1.3753454947484798,
+      "grad_norm": 1.7712354390495477,
+      "learning_rate": 4.475727163094572e-05,
+      "loss": 1.2497,
+      "step": 933
+    },
+    {
+      "epoch": 1.376819605675327,
+      "grad_norm": 2.003916719940972,
+      "learning_rate": 4.474513745380955e-05,
+      "loss": 1.2073,
+      "step": 934
+    },
+    {
+      "epoch": 1.3782937166021743,
+      "grad_norm": 1.8773015202071874,
+      "learning_rate": 4.473299089964015e-05,
+      "loss": 1.1917,
+      "step": 935
+    },
+    {
+      "epoch": 1.3797678275290215,
+      "grad_norm": 1.894054738318195,
+      "learning_rate": 4.472083197605146e-05,
+      "loss": 1.1564,
+      "step": 936
+    },
+    {
+      "epoch": 1.3812419384558687,
+      "grad_norm": 1.970212968071443,
+      "learning_rate": 4.470866069066516e-05,
+      "loss": 1.3188,
+      "step": 937
+    },
+    {
+      "epoch": 1.382716049382716,
+      "grad_norm": 1.682057965780056,
+      "learning_rate": 4.4696477051110705e-05,
+      "loss": 1.1682,
+      "step": 938
+    },
+    {
+      "epoch": 1.3841901603095632,
+      "grad_norm": 2.0254033381652596,
+      "learning_rate": 4.468428106502528e-05,
+      "loss": 1.1606,
+      "step": 939
+    },
+    {
+      "epoch": 1.3856642712364104,
+      "grad_norm": 2.0392637190063274,
+      "learning_rate": 4.4672072740053816e-05,
+      "loss": 1.2856,
+      "step": 940
+    },
+    {
+      "epoch": 1.3871383821632577,
+      "grad_norm": 1.9208541888292932,
+      "learning_rate": 4.4659852083848975e-05,
+      "loss": 1.1485,
+      "step": 941
+    },
+    {
+      "epoch": 1.388612493090105,
+      "grad_norm": 2.2490284199562156,
+      "learning_rate": 4.464761910407113e-05,
+      "loss": 1.2585,
+      "step": 942
+    },
+    {
+      "epoch": 1.3900866040169522,
+      "grad_norm": 1.7282960913951813,
+      "learning_rate": 4.463537380838841e-05,
+      "loss": 1.1456,
+      "step": 943
+    },
+    {
+      "epoch": 1.3915607149437994,
+      "grad_norm": 1.7542988565009476,
+      "learning_rate": 4.462311620447666e-05,
+      "loss": 1.2246,
+      "step": 944
+    },
+    {
+      "epoch": 1.3930348258706466,
+      "grad_norm": 1.9156444629470781,
+      "learning_rate": 4.461084630001942e-05,
+      "loss": 1.2611,
+      "step": 945
+    },
+    {
+      "epoch": 1.3945089367974939,
+      "grad_norm": 1.8049839321277235,
+      "learning_rate": 4.459856410270795e-05,
+      "loss": 1.2373,
+      "step": 946
+    },
+    {
+      "epoch": 1.3959830477243411,
+      "grad_norm": 1.9517555013104875,
+      "learning_rate": 4.4586269620241216e-05,
+      "loss": 1.1427,
+      "step": 947
+    },
+    {
+      "epoch": 1.3974571586511886,
+      "grad_norm": 1.7879347932646847,
+      "learning_rate": 4.457396286032589e-05,
+      "loss": 1.1145,
+      "step": 948
+    },
+    {
+      "epoch": 1.3989312695780358,
+      "grad_norm": 1.886347321463787,
+      "learning_rate": 4.4561643830676336e-05,
+      "loss": 1.3078,
+      "step": 949
+    },
+    {
+      "epoch": 1.400405380504883,
+      "grad_norm": 2.0960460997236554,
+      "learning_rate": 4.454931253901461e-05,
+      "loss": 1.2028,
+      "step": 950
+    },
+    {
+      "epoch": 1.4018794914317303,
+      "grad_norm": 2.0038057119866637,
+      "learning_rate": 4.453696899307045e-05,
+      "loss": 1.3936,
+      "step": 951
+    },
+    {
+      "epoch": 1.4033536023585775,
+      "grad_norm": 1.9078422525321308,
+      "learning_rate": 4.4524613200581284e-05,
+      "loss": 1.1678,
+      "step": 952
+    },
+    {
+      "epoch": 1.4048277132854248,
+      "grad_norm": 1.9148368670348805,
+      "learning_rate": 4.4512245169292206e-05,
+      "loss": 1.2803,
+      "step": 953
+    },
+    {
+      "epoch": 1.406301824212272,
+      "grad_norm": 1.8665223792487269,
+      "learning_rate": 4.449986490695599e-05,
+      "loss": 1.2145,
+      "step": 954
+    },
+    {
+      "epoch": 1.4077759351391192,
+      "grad_norm": 1.9907902623538376,
+      "learning_rate": 4.4487472421333074e-05,
+      "loss": 1.2263,
+      "step": 955
+    },
+    {
+      "epoch": 1.4092500460659665,
+      "grad_norm": 2.0034240975450066,
+      "learning_rate": 4.447506772019155e-05,
+      "loss": 1.1615,
+      "step": 956
+    },
+    {
+      "epoch": 1.4107241569928137,
+      "grad_norm": 2.0719946595126815,
+      "learning_rate": 4.44626508113072e-05,
+      "loss": 1.4633,
+      "step": 957
+    },
+    {
+      "epoch": 1.412198267919661,
+      "grad_norm": 1.9689547114435981,
+      "learning_rate": 4.445022170246341e-05,
+      "loss": 1.3024,
+      "step": 958
+    },
+    {
+      "epoch": 1.4136723788465082,
+      "grad_norm": 2.102384385276534,
+      "learning_rate": 4.443778040145124e-05,
+      "loss": 1.1678,
+      "step": 959
+    },
+    {
+      "epoch": 1.4151464897733554,
+      "grad_norm": 1.9068690061021774,
+      "learning_rate": 4.44253269160694e-05,
+      "loss": 1.1425,
+      "step": 960
+    },
+    {
+      "epoch": 1.4166206007002027,
+      "grad_norm": 1.8930244667459024,
+      "learning_rate": 4.441286125412422e-05,
+      "loss": 1.1605,
+      "step": 961
+    },
+    {
+      "epoch": 1.41809471162705,
+      "grad_norm": 2.028689017140227,
+      "learning_rate": 4.440038342342967e-05,
+      "loss": 1.2434,
+      "step": 962
+    },
+    {
+      "epoch": 1.4195688225538972,
+      "grad_norm": 1.911879809181173,
+      "learning_rate": 4.4387893431807344e-05,
+      "loss": 1.2782,
+      "step": 963
+    },
+    {
+      "epoch": 1.4210429334807444,
+      "grad_norm": 1.7739733588014863,
+      "learning_rate": 4.437539128708647e-05,
+      "loss": 1.1386,
+      "step": 964
+    },
+    {
+      "epoch": 1.4225170444075916,
+      "grad_norm": 2.116439405641339,
+      "learning_rate": 4.4362876997103885e-05,
+      "loss": 1.2379,
+      "step": 965
+    },
+    {
+      "epoch": 1.4239911553344389,
+      "grad_norm": 1.8639994582637933,
+      "learning_rate": 4.4350350569704045e-05,
+      "loss": 1.2856,
+      "step": 966
+    },
+    {
+      "epoch": 1.4254652662612861,
+      "grad_norm": 2.031336151643573,
+      "learning_rate": 4.4337812012738996e-05,
+      "loss": 1.1877,
+      "step": 967
+    },
+    {
+      "epoch": 1.4269393771881334,
+      "grad_norm": 2.046614535700179,
+      "learning_rate": 4.4325261334068426e-05,
+      "loss": 1.2791,
+      "step": 968
+    },
+    {
+      "epoch": 1.4284134881149806,
+      "grad_norm": 1.9761423259938107,
+      "learning_rate": 4.431269854155957e-05,
+      "loss": 1.1401,
+      "step": 969
+    },
+    {
+      "epoch": 1.4298875990418278,
+      "grad_norm": 1.893309334828956,
+      "learning_rate": 4.4300123643087304e-05,
+      "loss": 1.3055,
+      "step": 970
+    },
+    {
+      "epoch": 1.431361709968675,
+      "grad_norm": 1.7685626412628297,
+      "learning_rate": 4.428753664653406e-05,
+      "loss": 1.3326,
+      "step": 971
+    },
+    {
+      "epoch": 1.4328358208955223,
+      "grad_norm": 1.8284935212994555,
+      "learning_rate": 4.427493755978987e-05,
+      "loss": 1.2479,
+      "step": 972
+    },
+    {
+      "epoch": 1.4343099318223698,
+      "grad_norm": 2.0735294324535105,
+      "learning_rate": 4.426232639075234e-05,
+      "loss": 1.2229,
+      "step": 973
+    },
+    {
+      "epoch": 1.435784042749217,
+      "grad_norm": 1.8559715473156344,
+      "learning_rate": 4.424970314732664e-05,
+      "loss": 1.1806,
+      "step": 974
+    },
+    {
+      "epoch": 1.4372581536760642,
+      "grad_norm": 1.9585716291957957,
+      "learning_rate": 4.423706783742554e-05,
+      "loss": 1.3054,
+      "step": 975
+    },
+    {
+      "epoch": 1.4387322646029115,
+      "grad_norm": 1.8091937196671901,
+      "learning_rate": 4.422442046896933e-05,
+      "loss": 1.1424,
+      "step": 976
+    },
+    {
+      "epoch": 1.4402063755297587,
+      "grad_norm": 1.7330097633126418,
+      "learning_rate": 4.421176104988589e-05,
+      "loss": 1.1876,
+      "step": 977
+    },
+    {
+      "epoch": 1.441680486456606,
+      "grad_norm": 1.7391180531693347,
+      "learning_rate": 4.419908958811064e-05,
+      "loss": 1.1685,
+      "step": 978
+    },
+    {
+      "epoch": 1.4431545973834532,
+      "grad_norm": 1.7289370955082806,
+      "learning_rate": 4.418640609158656e-05,
+      "loss": 1.3844,
+      "step": 979
+    },
+    {
+      "epoch": 1.4446287083103004,
+      "grad_norm": 1.666172892223675,
+      "learning_rate": 4.417371056826417e-05,
+      "loss": 1.2546,
+      "step": 980
+    },
+    {
+      "epoch": 1.4461028192371477,
+      "grad_norm": 1.9808952880444375,
+      "learning_rate": 4.4161003026101525e-05,
+      "loss": 1.1787,
+      "step": 981
+    },
+    {
+      "epoch": 1.447576930163995,
+      "grad_norm": 2.0488065967574265,
+      "learning_rate": 4.41482834730642e-05,
+      "loss": 1.2623,
+      "step": 982
+    },
+    {
+      "epoch": 1.4490510410908422,
+      "grad_norm": 1.7999039041511262,
+      "learning_rate": 4.4135551917125334e-05,
+      "loss": 1.0902,
+      "step": 983
+    },
+    {
+      "epoch": 1.4505251520176894,
+      "grad_norm": 1.8675776756300442,
+      "learning_rate": 4.4122808366265556e-05,
+      "loss": 1.1667,
+      "step": 984
+    },
+    {
+      "epoch": 1.4519992629445366,
+      "grad_norm": 1.7077469013036894,
+      "learning_rate": 4.411005282847304e-05,
+      "loss": 1.2685,
+      "step": 985
+    },
+    {
+      "epoch": 1.4534733738713839,
+      "grad_norm": 1.7941980739359924,
+      "learning_rate": 4.409728531174345e-05,
+      "loss": 1.0802,
+      "step": 986
+    },
+    {
+      "epoch": 1.454947484798231,
+      "grad_norm": 1.9702490896961755,
+      "learning_rate": 4.4084505824079975e-05,
+      "loss": 1.3053,
+      "step": 987
+    },
+    {
+      "epoch": 1.4564215957250783,
+      "grad_norm": 1.8792361408030633,
+      "learning_rate": 4.40717143734933e-05,
+      "loss": 1.1681,
+      "step": 988
+    },
+    {
+      "epoch": 1.4578957066519256,
+      "grad_norm": 1.7822119808906913,
+      "learning_rate": 4.405891096800162e-05,
+      "loss": 1.1079,
+      "step": 989
+    },
+    {
+      "epoch": 1.4593698175787728,
+      "grad_norm": 1.8977830802839433,
+      "learning_rate": 4.404609561563062e-05,
+      "loss": 1.0291,
+      "step": 990
+    },
+    {
+      "epoch": 1.46084392850562,
+      "grad_norm": 1.9218700998135256,
+      "learning_rate": 4.403326832441345e-05,
+      "loss": 1.233,
+      "step": 991
+    },
+    {
+      "epoch": 1.4623180394324673,
+      "grad_norm": 2.093780517355723,
+      "learning_rate": 4.402042910239078e-05,
+      "loss": 1.0966,
+      "step": 992
+    },
+    {
+      "epoch": 1.4637921503593145,
+      "grad_norm": 1.9256528614260946,
+      "learning_rate": 4.400757795761074e-05,
+      "loss": 1.3607,
+      "step": 993
+    },
+    {
+      "epoch": 1.4652662612861618,
+      "grad_norm": 1.9386013961300534,
+      "learning_rate": 4.399471489812893e-05,
+      "loss": 1.2029,
+      "step": 994
+    },
+    {
+      "epoch": 1.466740372213009,
+      "grad_norm": 2.0512529684413745,
+      "learning_rate": 4.398183993200843e-05,
+      "loss": 1.1685,
+      "step": 995
+    },
+    {
+      "epoch": 1.4682144831398563,
+      "grad_norm": 2.0936053781238435,
+      "learning_rate": 4.3968953067319777e-05,
+      "loss": 1.2021,
+      "step": 996
+    },
+    {
+      "epoch": 1.4696885940667035,
+      "grad_norm": 1.8396869438718744,
+      "learning_rate": 4.395605431214096e-05,
+      "loss": 1.3155,
+      "step": 997
+    },
+    {
+      "epoch": 1.4711627049935507,
+      "grad_norm": 1.9608838199395964,
+      "learning_rate": 4.394314367455744e-05,
+      "loss": 1.0922,
+      "step": 998
+    },
+    {
+      "epoch": 1.472636815920398,
+      "grad_norm": 1.7586355359448314,
+      "learning_rate": 4.393022116266212e-05,
+      "loss": 1.2193,
+      "step": 999
+    },
+    {
+      "epoch": 1.4741109268472452,
+      "grad_norm": 2.2619578180806923,
+      "learning_rate": 4.3917286784555325e-05,
+      "loss": 1.4428,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4741109268472452,
+      "eval_bleu": 0.04710557283526075,
+      "eval_bleu_1gram": 0.3537460085813653,
+      "eval_bleu_2gram": 0.11992925427956323,
+      "eval_bleu_3gram": 0.04622731006726815,
+      "eval_bleu_4gram": 0.021123928680788814,
+      "eval_rag_val_loss": 1.5481538125263747,
+      "eval_rouge1": 0.33311218076912447,
+      "eval_rouge2": 0.11163593556247434,
+      "eval_rougeL": 0.3133783621457725,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4755850377740924,
+      "grad_norm": 1.7852993496691465,
+      "learning_rate": 4.390434054834483e-05,
+      "loss": 1.2443,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4770591487009397,
+      "grad_norm": 1.9408911898702972,
+      "learning_rate": 4.389138246214588e-05,
+      "loss": 1.2617,
+      "step": 1002
+    },
+    {
+      "epoch": 1.478533259627787,
+      "grad_norm": 1.8804633336609262,
+      "learning_rate": 4.387841253408109e-05,
+      "loss": 1.2545,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4800073705546342,
+      "grad_norm": 1.807517800672155,
+      "learning_rate": 4.386543077228053e-05,
+      "loss": 1.0795,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 1.8226115888805021,
+      "learning_rate": 4.3852437184881687e-05,
+      "loss": 1.2368,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4829555924083286,
+      "grad_norm": 1.7759731499515286,
+      "learning_rate": 4.383943178002944e-05,
+      "loss": 1.1635,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4844297033351759,
+      "grad_norm": 2.0251827298830354,
+      "learning_rate": 4.382641456587611e-05,
+      "loss": 1.2174,
+      "step": 1007
+    },
+    {
+      "epoch": 1.4859038142620231,
+      "grad_norm": 1.8529512372175647,
+      "learning_rate": 4.38133855505814e-05,
+      "loss": 1.2467,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4873779251888704,
+      "grad_norm": 1.8583765635002416,
+      "learning_rate": 4.3800344742312396e-05,
+      "loss": 1.2368,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4888520361157176,
+      "grad_norm": 1.6766595561721727,
+      "learning_rate": 4.3787292149243605e-05,
+      "loss": 1.2179,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4903261470425648,
+      "grad_norm": 1.887057887965862,
+      "learning_rate": 4.3774227779556906e-05,
+      "loss": 1.0748,
+      "step": 1011
+    },
+    {
+      "epoch": 1.491800257969412,
+      "grad_norm": 1.8740406125043485,
+      "learning_rate": 4.376115164144157e-05,
+      "loss": 1.2463,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4932743688962593,
+      "grad_norm": 1.8362573872541625,
+      "learning_rate": 4.374806374309421e-05,
+      "loss": 1.0845,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4947484798231068,
+      "grad_norm": 1.9110518464182196,
+      "learning_rate": 4.3734964092718885e-05,
+      "loss": 1.2809,
+      "step": 1014
+    },
+    {
+      "epoch": 1.496222590749954,
+      "grad_norm": 2.0270872449951485,
+      "learning_rate": 4.372185269852693e-05,
+      "loss": 1.3253,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4976967016768012,
+      "grad_norm": 1.8056851886608907,
+      "learning_rate": 4.370872956873712e-05,
+      "loss": 1.2459,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4991708126036485,
+      "grad_norm": 1.7809119322012972,
+      "learning_rate": 4.369559471157552e-05,
+      "loss": 1.2797,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5006449235304957,
+      "grad_norm": 2.111677368526342,
+      "learning_rate": 4.36824481352756e-05,
+      "loss": 1.3684,
+      "step": 1018
+    },
+    {
+      "epoch": 1.502119034457343,
+      "grad_norm": 1.8136400386358484,
+      "learning_rate": 4.366928984807815e-05,
+      "loss": 1.3088,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5035931453841902,
+      "grad_norm": 1.6579132365865314,
+      "learning_rate": 4.36561198582313e-05,
+      "loss": 1.1212,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5050672563110374,
+      "grad_norm": 1.9304753598663904,
+      "learning_rate": 4.364293817399052e-05,
+      "loss": 1.2989,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5065413672378847,
+      "grad_norm": 1.8308767491666493,
+      "learning_rate": 4.362974480361862e-05,
+      "loss": 1.1554,
+      "step": 1022
+    },
+    {
+      "epoch": 1.508015478164732,
+      "grad_norm": 1.964352980613792,
+      "learning_rate": 4.361653975538572e-05,
+      "loss": 1.3509,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5094895890915792,
+      "grad_norm": 2.082338693331692,
+      "learning_rate": 4.3603323037569265e-05,
+      "loss": 1.279,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5109637000184264,
+      "grad_norm": 1.8059765724276355,
+      "learning_rate": 4.359009465845402e-05,
+      "loss": 1.0633,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5124378109452736,
+      "grad_norm": 1.895122129183627,
+      "learning_rate": 4.3576854626332055e-05,
+      "loss": 1.1434,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5139119218721209,
+      "grad_norm": 1.837701801974174,
+      "learning_rate": 4.356360294950275e-05,
+      "loss": 1.2943,
+      "step": 1027
+    },
+    {
+      "epoch": 1.515386032798968,
+      "grad_norm": 2.002332519309243,
+      "learning_rate": 4.3550339636272775e-05,
+      "loss": 1.3521,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5168601437258153,
+      "grad_norm": 1.7799747653844067,
+      "learning_rate": 4.35370646949561e-05,
+      "loss": 1.2239,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5183342546526626,
+      "grad_norm": 1.8763600820718749,
+      "learning_rate": 4.352377813387398e-05,
+      "loss": 1.2009,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5198083655795098,
+      "grad_norm": 1.873681558551299,
+      "learning_rate": 4.3510479961354964e-05,
+      "loss": 1.2526,
+      "step": 1031
+    },
+    {
+      "epoch": 1.521282476506357,
+      "grad_norm": 1.6504819946277676,
+      "learning_rate": 4.349717018573487e-05,
+      "loss": 1.1681,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5227565874332043,
+      "grad_norm": 1.856955805727659,
+      "learning_rate": 4.348384881535679e-05,
+      "loss": 1.2538,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5242306983600515,
+      "grad_norm": 1.8931735807616588,
+      "learning_rate": 4.347051585857109e-05,
+      "loss": 1.3238,
+      "step": 1034
+    },
+    {
+      "epoch": 1.525704809286899,
+      "grad_norm": 1.830682123555617,
+      "learning_rate": 4.34571713237354e-05,
+      "loss": 1.1462,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5271789202137462,
+      "grad_norm": 1.9411429336017156,
+      "learning_rate": 4.344381521921458e-05,
+      "loss": 1.2485,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5286530311405935,
+      "grad_norm": 1.7587618828367326,
+      "learning_rate": 4.3430447553380785e-05,
+      "loss": 1.074,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5301271420674407,
+      "grad_norm": 1.9504948721678683,
+      "learning_rate": 4.34170683346134e-05,
+      "loss": 1.207,
+      "step": 1038
+    },
+    {
+      "epoch": 1.531601252994288,
+      "grad_norm": 1.8973504867251907,
+      "learning_rate": 4.3403677571299026e-05,
+      "loss": 1.4281,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5330753639211352,
+      "grad_norm": 1.9216689332726449,
+      "learning_rate": 4.339027527183154e-05,
+      "loss": 1.1606,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5345494748479824,
+      "grad_norm": 1.757229788941443,
+      "learning_rate": 4.337686144461204e-05,
+      "loss": 1.1878,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5360235857748297,
+      "grad_norm": 1.802586641385353,
+      "learning_rate": 4.3363436098048825e-05,
+      "loss": 1.3005,
+      "step": 1042
+    },
+    {
+      "epoch": 1.537497696701677,
+      "grad_norm": 1.9067754802909465,
+      "learning_rate": 4.3349999240557446e-05,
+      "loss": 1.1313,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5389718076285241,
+      "grad_norm": 2.126324633455197,
+      "learning_rate": 4.333655088056065e-05,
+      "loss": 1.4061,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5404459185553714,
+      "grad_norm": 1.9076362947139993,
+      "learning_rate": 4.332309102648841e-05,
+      "loss": 1.3356,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5419200294822186,
+      "grad_norm": 1.771936461982236,
+      "learning_rate": 4.330961968677788e-05,
+      "loss": 1.1578,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5433941404090659,
+      "grad_norm": 1.9262885301859323,
+      "learning_rate": 4.329613686987344e-05,
+      "loss": 1.3629,
+      "step": 1047
+    },
+    {
+      "epoch": 1.544868251335913,
+      "grad_norm": 1.8219757994416996,
+      "learning_rate": 4.328264258422665e-05,
+      "loss": 1.2104,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5463423622627603,
+      "grad_norm": 1.9569555941822794,
+      "learning_rate": 4.3269136838296264e-05,
+      "loss": 1.345,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5478164731896076,
+      "grad_norm": 1.819906230334432,
+      "learning_rate": 4.325561964054822e-05,
+      "loss": 1.4101,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5492905841164548,
+      "grad_norm": 2.082157896960155,
+      "learning_rate": 4.324209099945563e-05,
+      "loss": 1.1818,
+      "step": 1051
+    },
+    {
+      "epoch": 1.550764695043302,
+      "grad_norm": 1.86058710164103,
+      "learning_rate": 4.322855092349878e-05,
+      "loss": 1.3853,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5522388059701493,
+      "grad_norm": 1.836146111506626,
+      "learning_rate": 4.321499942116511e-05,
+      "loss": 1.1401,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5537129168969965,
+      "grad_norm": 1.8128654506179072,
+      "learning_rate": 4.320143650094927e-05,
+      "loss": 1.2487,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5551870278238438,
+      "grad_norm": 1.8970593131952311,
+      "learning_rate": 4.318786217135301e-05,
+      "loss": 1.3857,
+      "step": 1055
+    },
+    {
+      "epoch": 1.556661138750691,
+      "grad_norm": 1.9615967998632555,
+      "learning_rate": 4.3174276440885276e-05,
+      "loss": 1.185,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5581352496775382,
+      "grad_norm": 1.7684784508908382,
+      "learning_rate": 4.316067931806212e-05,
+      "loss": 1.0542,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5596093606043855,
+      "grad_norm": 1.9063932646399908,
+      "learning_rate": 4.3147070811406765e-05,
+      "loss": 1.2336,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5610834715312327,
+      "grad_norm": 1.863391417119542,
+      "learning_rate": 4.313345092944957e-05,
+      "loss": 1.2063,
+      "step": 1059
+    },
+    {
+      "epoch": 1.56255758245808,
+      "grad_norm": 1.8144653451300685,
+      "learning_rate": 4.3119819680728e-05,
+      "loss": 1.3092,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5640316933849272,
+      "grad_norm": 1.905964939503467,
+      "learning_rate": 4.310617707378668e-05,
+      "loss": 1.2017,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5655058043117744,
+      "grad_norm": 1.9321008332675937,
+      "learning_rate": 4.309252311717732e-05,
+      "loss": 1.2136,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5669799152386217,
+      "grad_norm": 2.0171368274189496,
+      "learning_rate": 4.307885781945876e-05,
+      "loss": 1.2324,
+      "step": 1063
+    },
+    {
+      "epoch": 1.568454026165469,
+      "grad_norm": 1.8587233739919908,
+      "learning_rate": 4.3065181189196956e-05,
+      "loss": 1.3017,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5699281370923162,
+      "grad_norm": 1.9124230662311437,
+      "learning_rate": 4.305149323496497e-05,
+      "loss": 1.3507,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5714022480191634,
+      "grad_norm": 1.8777700942570117,
+      "learning_rate": 4.303779396534293e-05,
+      "loss": 1.2357,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5728763589460106,
+      "grad_norm": 1.7811982665998625,
+      "learning_rate": 4.30240833889181e-05,
+      "loss": 1.2835,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5743504698728579,
+      "grad_norm": 1.9165017291986042,
+      "learning_rate": 4.30103615142848e-05,
+      "loss": 1.258,
+      "step": 1068
+    },
+    {
+      "epoch": 1.575824580799705,
+      "grad_norm": 1.7652063464180527,
+      "learning_rate": 4.2996628350044454e-05,
+      "loss": 1.2886,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5772986917265523,
+      "grad_norm": 1.8500412755949607,
+      "learning_rate": 4.298288390480554e-05,
+      "loss": 1.1388,
+      "step": 1070
+    },
+    {
+      "epoch": 1.5787728026533996,
+      "grad_norm": 1.9184986011936476,
+      "learning_rate": 4.296912818718363e-05,
+      "loss": 1.1086,
+      "step": 1071
+    },
+    {
+      "epoch": 1.5802469135802468,
+      "grad_norm": 1.915202645916812,
+      "learning_rate": 4.295536120580135e-05,
+      "loss": 1.3577,
+      "step": 1072
+    },
+    {
+      "epoch": 1.581721024507094,
+      "grad_norm": 1.985155687928824,
+      "learning_rate": 4.2941582969288384e-05,
+      "loss": 1.3213,
+      "step": 1073
+    },
+    {
+      "epoch": 1.5831951354339413,
+      "grad_norm": 1.9104833030380386,
+      "learning_rate": 4.292779348628148e-05,
+      "loss": 1.3851,
+      "step": 1074
+    },
+    {
+      "epoch": 1.5846692463607885,
+      "grad_norm": 1.91145109197785,
+      "learning_rate": 4.2913992765424434e-05,
+      "loss": 1.2434,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5861433572876358,
+      "grad_norm": 1.730167405733206,
+      "learning_rate": 4.2900180815368076e-05,
+      "loss": 1.2735,
+      "step": 1076
+    },
+    {
+      "epoch": 1.587617468214483,
+      "grad_norm": 1.8422407665106706,
+      "learning_rate": 4.2886357644770294e-05,
+      "loss": 1.213,
+      "step": 1077
+    },
+    {
+      "epoch": 1.5890915791413303,
+      "grad_norm": 2.0168250008631006,
+      "learning_rate": 4.287252326229598e-05,
+      "loss": 1.1869,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5905656900681775,
+      "grad_norm": 1.8473379494054016,
+      "learning_rate": 4.285867767661709e-05,
+      "loss": 1.219,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5920398009950247,
+      "grad_norm": 1.7956929838679612,
+      "learning_rate": 4.284482089641257e-05,
+      "loss": 1.1474,
+      "step": 1080
+    },
+    {
+      "epoch": 1.593513911921872,
+      "grad_norm": 1.8855936870054948,
+      "learning_rate": 4.283095293036842e-05,
+      "loss": 1.1987,
+      "step": 1081
+    },
+    {
+      "epoch": 1.5949880228487192,
+      "grad_norm": 1.7377318467712326,
+      "learning_rate": 4.281707378717761e-05,
+      "loss": 1.2977,
+      "step": 1082
+    },
+    {
+      "epoch": 1.5964621337755664,
+      "grad_norm": 1.941549253665688,
+      "learning_rate": 4.280318347554013e-05,
+      "loss": 1.3362,
+      "step": 1083
+    },
+    {
+      "epoch": 1.597936244702414,
+      "grad_norm": 1.79061648914385,
+      "learning_rate": 4.2789282004163e-05,
+      "loss": 1.2801,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5994103556292611,
+      "grad_norm": 2.1294220185862045,
+      "learning_rate": 4.27753693817602e-05,
+      "loss": 1.1791,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6008844665561084,
+      "grad_norm": 1.789910261366166,
+      "learning_rate": 4.276144561705271e-05,
+      "loss": 1.0298,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6023585774829556,
+      "grad_norm": 1.83935987584918,
+      "learning_rate": 4.27475107187685e-05,
+      "loss": 1.1299,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6038326884098029,
+      "grad_norm": 1.8073147232681004,
+      "learning_rate": 4.273356469564251e-05,
+      "loss": 1.2355,
+      "step": 1088
+    },
+    {
+      "epoch": 1.60530679933665,
+      "grad_norm": 1.9020461333342862,
+      "learning_rate": 4.271960755641668e-05,
+      "loss": 1.14,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6067809102634973,
+      "grad_norm": 1.90104953990682,
+      "learning_rate": 4.270563930983986e-05,
+      "loss": 1.2043,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6082550211903446,
+      "grad_norm": 1.8690935565439237,
+      "learning_rate": 4.269165996466793e-05,
+      "loss": 1.3084,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6097291321171918,
+      "grad_norm": 1.8732310533714696,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 1.2108,
+      "step": 1092
+    },
+    {
+      "epoch": 1.611203243044039,
+      "grad_norm": 1.8326481347884551,
+      "learning_rate": 4.266366801359689e-05,
+      "loss": 1.2946,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6126773539708863,
+      "grad_norm": 1.7119660791218507,
+      "learning_rate": 4.264965542524424e-05,
+      "loss": 0.9716,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6141514648977335,
+      "grad_norm": 2.006191564166435,
+      "learning_rate": 4.263563177338938e-05,
+      "loss": 1.2558,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6156255758245808,
+      "grad_norm": 1.756098271667975,
+      "learning_rate": 4.262159706682291e-05,
+      "loss": 1.2415,
+      "step": 1096
+    },
+    {
+      "epoch": 1.617099686751428,
+      "grad_norm": 1.8145735488201515,
+      "learning_rate": 4.2607551314342297e-05,
+      "loss": 1.196,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6185737976782753,
+      "grad_norm": 1.9683976236492993,
+      "learning_rate": 4.259349452475202e-05,
+      "loss": 1.1848,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6200479086051225,
+      "grad_norm": 1.9825803195417955,
+      "learning_rate": 4.25794267068634e-05,
+      "loss": 1.2926,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6215220195319697,
+      "grad_norm": 2.0098460545586327,
+      "learning_rate": 4.256534786949472e-05,
+      "loss": 1.3416,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6229961304588172,
+      "grad_norm": 2.1685209653552024,
+      "learning_rate": 4.255125802147114e-05,
+      "loss": 1.3728,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6244702413856644,
+      "grad_norm": 1.938307686086913,
+      "learning_rate": 4.253715717162474e-05,
+      "loss": 1.1128,
+      "step": 1102
+    },
+    {
+      "epoch": 1.6259443523125117,
+      "grad_norm": 1.8195898889287117,
+      "learning_rate": 4.252304532879449e-05,
+      "loss": 1.2526,
+      "step": 1103
+    },
+    {
+      "epoch": 1.627418463239359,
+      "grad_norm": 2.042627017905458,
+      "learning_rate": 4.2508922501826244e-05,
+      "loss": 1.2557,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6288925741662061,
+      "grad_norm": 1.9459736018571157,
+      "learning_rate": 4.249478869957276e-05,
+      "loss": 1.2338,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6303666850930534,
+      "grad_norm": 2.0336472670217516,
+      "learning_rate": 4.248064393089366e-05,
+      "loss": 1.2837,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6318407960199006,
+      "grad_norm": 1.8672951383968779,
+      "learning_rate": 4.246648820465544e-05,
+      "loss": 1.2686,
+      "step": 1107
+    },
+    {
+      "epoch": 1.6333149069467479,
+      "grad_norm": 1.9262759673926062,
+      "learning_rate": 4.2452321529731475e-05,
+      "loss": 1.1191,
+      "step": 1108
+    },
+    {
+      "epoch": 1.634789017873595,
+      "grad_norm": 1.9928605681710756,
+      "learning_rate": 4.2438143915002e-05,
+      "loss": 1.2318,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6362631288004423,
+      "grad_norm": 1.7655129186113852,
+      "learning_rate": 4.242395536935409e-05,
+      "loss": 1.387,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6377372397272896,
+      "grad_norm": 2.0112439234080486,
+      "learning_rate": 4.2409755901681716e-05,
+      "loss": 1.3361,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6392113506541368,
+      "grad_norm": 1.787524299189659,
+      "learning_rate": 4.239554552088563e-05,
+      "loss": 1.2788,
+      "step": 1112
+    },
+    {
+      "epoch": 1.640685461580984,
+      "grad_norm": 1.8300088810314366,
+      "learning_rate": 4.238132423587349e-05,
+      "loss": 1.3415,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6421595725078313,
+      "grad_norm": 1.9643185711874585,
+      "learning_rate": 4.236709205555973e-05,
+      "loss": 1.2452,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6436336834346785,
+      "grad_norm": 1.8802687053198865,
+      "learning_rate": 4.235284898886568e-05,
+      "loss": 1.4005,
+      "step": 1115
+    },
+    {
+      "epoch": 1.6451077943615258,
+      "grad_norm": 1.8477788323996884,
+      "learning_rate": 4.233859504471943e-05,
+      "loss": 1.1429,
+      "step": 1116
+    },
+    {
+      "epoch": 1.646581905288373,
+      "grad_norm": 1.964663548482063,
+      "learning_rate": 4.2324330232055924e-05,
+      "loss": 1.1998,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6480560162152202,
+      "grad_norm": 1.9047670067707487,
+      "learning_rate": 4.231005455981692e-05,
+      "loss": 1.1231,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6495301271420675,
+      "grad_norm": 1.845622905622793,
+      "learning_rate": 4.2295768036950953e-05,
+      "loss": 1.2158,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6510042380689147,
+      "grad_norm": 1.7227520213244676,
+      "learning_rate": 4.22814706724134e-05,
+      "loss": 1.226,
+      "step": 1120
+    },
+    {
+      "epoch": 1.652478348995762,
+      "grad_norm": 1.9877921291603484,
+      "learning_rate": 4.226716247516641e-05,
+      "loss": 1.2949,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6539524599226092,
+      "grad_norm": 1.810466678164202,
+      "learning_rate": 4.2252843454178925e-05,
+      "loss": 1.3307,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6554265708494564,
+      "grad_norm": 1.7627726107315673,
+      "learning_rate": 4.223851361842668e-05,
+      "loss": 1.1535,
+      "step": 1123
+    },
+    {
+      "epoch": 1.6569006817763037,
+      "grad_norm": 1.8907676832095228,
+      "learning_rate": 4.222417297689217e-05,
+      "loss": 1.2935,
+      "step": 1124
+    },
+    {
+      "epoch": 1.658374792703151,
+      "grad_norm": 1.7970703018863439,
+      "learning_rate": 4.2209821538564684e-05,
+      "loss": 1.2032,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6598489036299982,
+      "grad_norm": 1.8964828664064546,
+      "learning_rate": 4.219545931244027e-05,
+      "loss": 1.3649,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6613230145568454,
+      "grad_norm": 1.824122685715049,
+      "learning_rate": 4.218108630752174e-05,
+      "loss": 1.4419,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6627971254836926,
+      "grad_norm": 1.912200894371983,
+      "learning_rate": 4.2166702532818665e-05,
+      "loss": 1.297,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6642712364105399,
+      "grad_norm": 1.9770491532431862,
+      "learning_rate": 4.2152307997347365e-05,
+      "loss": 1.2846,
+      "step": 1129
+    },
+    {
+      "epoch": 1.665745347337387,
+      "grad_norm": 1.9258810330861753,
+      "learning_rate": 4.213790271013089e-05,
+      "loss": 1.2353,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6672194582642343,
+      "grad_norm": 2.0050289585752474,
+      "learning_rate": 4.212348668019906e-05,
+      "loss": 1.195,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6686935691910816,
+      "grad_norm": 1.830339313380367,
+      "learning_rate": 4.2109059916588414e-05,
+      "loss": 1.0695,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6701676801179288,
+      "grad_norm": 1.833505001122671,
+      "learning_rate": 4.20946224283422e-05,
+      "loss": 1.0125,
+      "step": 1133
+    },
+    {
+      "epoch": 1.671641791044776,
+      "grad_norm": 1.9804112051705174,
+      "learning_rate": 4.2080174224510426e-05,
+      "loss": 1.324,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6731159019716233,
+      "grad_norm": 1.710219628914054,
+      "learning_rate": 4.2065715314149775e-05,
+      "loss": 1.0566,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6745900128984705,
+      "grad_norm": 1.64057282864407,
+      "learning_rate": 4.2051245706323696e-05,
+      "loss": 1.1222,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6760641238253178,
+      "grad_norm": 1.9637349149147647,
+      "learning_rate": 4.2036765410102285e-05,
+      "loss": 1.0906,
+      "step": 1137
+    },
+    {
+      "epoch": 1.677538234752165,
+      "grad_norm": 1.8240049837320405,
+      "learning_rate": 4.202227443456238e-05,
+      "loss": 1.2323,
+      "step": 1138
+    },
+    {
+      "epoch": 1.6790123456790123,
+      "grad_norm": 1.9480210410593393,
+      "learning_rate": 4.200777278878749e-05,
+      "loss": 1.2079,
+      "step": 1139
+    },
+    {
+      "epoch": 1.6804864566058595,
+      "grad_norm": 1.9074388142743126,
+      "learning_rate": 4.199326048186782e-05,
+      "loss": 1.2942,
+      "step": 1140
+    },
+    {
+      "epoch": 1.6819605675327067,
+      "grad_norm": 1.8919630753904124,
+      "learning_rate": 4.197873752290027e-05,
+      "loss": 1.2149,
+      "step": 1141
+    },
+    {
+      "epoch": 1.683434678459554,
+      "grad_norm": 1.7600493715687344,
+      "learning_rate": 4.1964203920988385e-05,
+      "loss": 1.0952,
+      "step": 1142
+    },
+    {
+      "epoch": 1.6849087893864012,
+      "grad_norm": 1.9261659930388288,
+      "learning_rate": 4.19496596852424e-05,
+      "loss": 1.3109,
+      "step": 1143
+    },
+    {
+      "epoch": 1.6863829003132484,
+      "grad_norm": 1.8768810373489273,
+      "learning_rate": 4.1935104824779246e-05,
+      "loss": 1.294,
+      "step": 1144
+    },
+    {
+      "epoch": 1.6878570112400957,
+      "grad_norm": 1.9677935502228958,
+      "learning_rate": 4.192053934872247e-05,
+      "loss": 1.2749,
+      "step": 1145
+    },
+    {
+      "epoch": 1.689331122166943,
+      "grad_norm": 2.037120142439185,
+      "learning_rate": 4.1905963266202276e-05,
+      "loss": 1.4639,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6908052330937902,
+      "grad_norm": 1.861661642962001,
+      "learning_rate": 4.189137658635555e-05,
+      "loss": 1.3886,
+      "step": 1147
+    },
+    {
+      "epoch": 1.6922793440206374,
+      "grad_norm": 2.3855936798206425,
+      "learning_rate": 4.187677931832578e-05,
+      "loss": 1.3504,
+      "step": 1148
+    },
+    {
+      "epoch": 1.6937534549474846,
+      "grad_norm": 2.1097919193374524,
+      "learning_rate": 4.1862171471263126e-05,
+      "loss": 1.3451,
+      "step": 1149
+    },
+    {
+      "epoch": 1.695227565874332,
+      "grad_norm": 1.9216781143195996,
+      "learning_rate": 4.184755305432436e-05,
+      "loss": 1.3488,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6967016768011793,
+      "grad_norm": 1.9479156602935095,
+      "learning_rate": 4.1832924076672876e-05,
+      "loss": 1.2722,
+      "step": 1151
+    },
+    {
+      "epoch": 1.6981757877280266,
+      "grad_norm": 1.8703568506678965,
+      "learning_rate": 4.181828454747872e-05,
+      "loss": 1.2535,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6996498986548738,
+      "grad_norm": 1.873169386817977,
+      "learning_rate": 4.180363447591849e-05,
+      "loss": 1.4875,
+      "step": 1153
+    },
+    {
+      "epoch": 1.701124009581721,
+      "grad_norm": 1.9670314857239748,
+      "learning_rate": 4.178897387117546e-05,
+      "loss": 1.4447,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7025981205085683,
+      "grad_norm": 1.727028831232438,
+      "learning_rate": 4.177430274243947e-05,
+      "loss": 1.2714,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7040722314354155,
+      "grad_norm": 1.7399758297787102,
+      "learning_rate": 4.175962109890696e-05,
+      "loss": 1.2887,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7055463423622628,
+      "grad_norm": 1.9946159969364994,
+      "learning_rate": 4.1744928949780975e-05,
+      "loss": 1.332,
+      "step": 1157
+    },
+    {
+      "epoch": 1.70702045328911,
+      "grad_norm": 1.8748412382940545,
+      "learning_rate": 4.173022630427113e-05,
+      "loss": 1.1747,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7084945642159572,
+      "grad_norm": 1.801780307701713,
+      "learning_rate": 4.1715513171593614e-05,
+      "loss": 1.3079,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7099686751428045,
+      "grad_norm": 1.8248668230141263,
+      "learning_rate": 4.170078956097121e-05,
+      "loss": 1.3446,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7114427860696517,
+      "grad_norm": 1.8900217402181687,
+      "learning_rate": 4.168605548163326e-05,
+      "loss": 1.2939,
+      "step": 1161
+    },
+    {
+      "epoch": 1.712916896996499,
+      "grad_norm": 1.7064373458461526,
+      "learning_rate": 4.167131094281565e-05,
+      "loss": 1.0821,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7143910079233462,
+      "grad_norm": 1.7611034189905972,
+      "learning_rate": 4.165655595376088e-05,
+      "loss": 1.3065,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7158651188501934,
+      "grad_norm": 1.805938353266847,
+      "learning_rate": 4.1641790523717935e-05,
+      "loss": 1.1311,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7173392297770407,
+      "grad_norm": 1.9556661935175772,
+      "learning_rate": 4.162701466194237e-05,
+      "loss": 1.3561,
+      "step": 1165
+    },
+    {
+      "epoch": 1.718813340703888,
+      "grad_norm": 1.7340302683202522,
+      "learning_rate": 4.161222837769627e-05,
+      "loss": 1.0482,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7202874516307354,
+      "grad_norm": 1.8453258956435885,
+      "learning_rate": 4.159743168024829e-05,
+      "loss": 1.337,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7217615625575826,
+      "grad_norm": 1.8831672809056383,
+      "learning_rate": 4.158262457887356e-05,
+      "loss": 1.1831,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7232356734844299,
+      "grad_norm": 1.9917243568758662,
+      "learning_rate": 4.156780708285378e-05,
+      "loss": 1.2676,
+      "step": 1169
+    },
+    {
+      "epoch": 1.724709784411277,
+      "grad_norm": 1.8132331286236758,
+      "learning_rate": 4.155297920147713e-05,
+      "loss": 1.1476,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7261838953381243,
+      "grad_norm": 1.8651559543601395,
+      "learning_rate": 4.153814094403831e-05,
+      "loss": 1.2898,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7276580062649716,
+      "grad_norm": 1.9299369372802475,
+      "learning_rate": 4.1523292319838524e-05,
+      "loss": 1.0423,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7291321171918188,
+      "grad_norm": 2.0821933933697365,
+      "learning_rate": 4.150843333818549e-05,
+      "loss": 1.3333,
+      "step": 1173
+    },
+    {
+      "epoch": 1.730606228118666,
+      "grad_norm": 1.7358736316595578,
+      "learning_rate": 4.149356400839339e-05,
+      "loss": 1.047,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7320803390455133,
+      "grad_norm": 2.31799046088717,
+      "learning_rate": 4.1478684339782926e-05,
+      "loss": 1.3923,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7335544499723605,
+      "grad_norm": 1.8034789427227818,
+      "learning_rate": 4.1463794341681244e-05,
+      "loss": 1.2442,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7350285608992078,
+      "grad_norm": 1.9622704584282626,
+      "learning_rate": 4.1448894023422005e-05,
+      "loss": 1.2256,
+      "step": 1177
+    },
+    {
+      "epoch": 1.736502671826055,
+      "grad_norm": 2.0212032988843904,
+      "learning_rate": 4.143398339434529e-05,
+      "loss": 1.3743,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7379767827529022,
+      "grad_norm": 1.9532401089126716,
+      "learning_rate": 4.1419062463797695e-05,
+      "loss": 1.2189,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7394508936797495,
+      "grad_norm": 1.9884353308191478,
+      "learning_rate": 4.140413124113225e-05,
+      "loss": 1.4263,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7409250046065967,
+      "grad_norm": 2.0396914626524727,
+      "learning_rate": 4.138918973570842e-05,
+      "loss": 1.1713,
+      "step": 1181
+    },
+    {
+      "epoch": 1.742399115533444,
+      "grad_norm": 1.89297005735597,
+      "learning_rate": 4.1374237956892133e-05,
+      "loss": 1.2778,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7438732264602912,
+      "grad_norm": 2.0251309293640483,
+      "learning_rate": 4.135927591405577e-05,
+      "loss": 1.295,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7453473373871384,
+      "grad_norm": 2.0096657360228924,
+      "learning_rate": 4.134430361657813e-05,
+      "loss": 1.1209,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7468214483139857,
+      "grad_norm": 1.9954076733814325,
+      "learning_rate": 4.1329321073844415e-05,
+      "loss": 1.201,
+      "step": 1185
+    },
+    {
+      "epoch": 1.748295559240833,
+      "grad_norm": 2.01562937654715,
+      "learning_rate": 4.131432829524631e-05,
+      "loss": 1.1608,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7497696701676801,
+      "grad_norm": 2.01413476546496,
+      "learning_rate": 4.129932529018187e-05,
+      "loss": 1.2642,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7512437810945274,
+      "grad_norm": 2.1457401613420806,
+      "learning_rate": 4.128431206805557e-05,
+      "loss": 1.433,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7527178920213746,
+      "grad_norm": 1.8965854480340019,
+      "learning_rate": 4.126928863827827e-05,
+      "loss": 1.2827,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7541920029482219,
+      "grad_norm": 1.752547861661921,
+      "learning_rate": 4.1254255010267285e-05,
+      "loss": 1.284,
+      "step": 1190
+    },
+    {
+      "epoch": 1.755666113875069,
+      "grad_norm": 1.8389301988175208,
+      "learning_rate": 4.123921119344627e-05,
+      "loss": 1.2685,
+      "step": 1191
+    },
+    {
+      "epoch": 1.7571402248019163,
+      "grad_norm": 1.8339557097025971,
+      "learning_rate": 4.122415719724528e-05,
+      "loss": 1.4762,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7586143357287636,
+      "grad_norm": 1.865088715782675,
+      "learning_rate": 4.120909303110078e-05,
+      "loss": 1.2256,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7600884466556108,
+      "grad_norm": 1.6230941012749778,
+      "learning_rate": 4.119401870445555e-05,
+      "loss": 1.1583,
+      "step": 1194
+    },
+    {
+      "epoch": 1.761562557582458,
+      "grad_norm": 1.9337570583853347,
+      "learning_rate": 4.1178934226758803e-05,
+      "loss": 1.1924,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7630366685093053,
+      "grad_norm": 1.8186642629858298,
+      "learning_rate": 4.1163839607466084e-05,
+      "loss": 1.1415,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7645107794361525,
+      "grad_norm": 1.7903219395963126,
+      "learning_rate": 4.114873485603927e-05,
+      "loss": 1.2286,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7659848903629998,
+      "grad_norm": 1.6935997674622014,
+      "learning_rate": 4.113361998194665e-05,
+      "loss": 1.1085,
+      "step": 1198
+    },
+    {
+      "epoch": 1.767459001289847,
+      "grad_norm": 1.9573374982593792,
+      "learning_rate": 4.111849499466281e-05,
+      "loss": 1.4939,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7689331122166942,
+      "grad_norm": 1.7747274404452258,
+      "learning_rate": 4.110335990366868e-05,
+      "loss": 1.1663,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7704072231435415,
+      "grad_norm": 1.739624190469356,
+      "learning_rate": 4.108821471845155e-05,
+      "loss": 1.2981,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7718813340703887,
+      "grad_norm": 1.8854769772601188,
+      "learning_rate": 4.107305944850502e-05,
+      "loss": 1.3968,
+      "step": 1202
+    },
+    {
+      "epoch": 1.773355444997236,
+      "grad_norm": 1.7286527934848495,
+      "learning_rate": 4.105789410332901e-05,
+      "loss": 1.0797,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7748295559240832,
+      "grad_norm": 1.7856795798062597,
+      "learning_rate": 4.104271869242975e-05,
+      "loss": 1.1614,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7763036668509304,
+      "grad_norm": 1.7031263998883295,
+      "learning_rate": 4.10275332253198e-05,
+      "loss": 1.1268,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 1.7717712910221877,
+      "learning_rate": 4.1012337711518e-05,
+      "loss": 1.2461,
+      "step": 1206
+    },
+    {
+      "epoch": 1.779251888704625,
+      "grad_norm": 1.9373306692638488,
+      "learning_rate": 4.099713216054952e-05,
+      "loss": 1.3029,
+      "step": 1207
+    },
+    {
+      "epoch": 1.7807259996314722,
+      "grad_norm": 1.8833718636321377,
+      "learning_rate": 4.098191658194578e-05,
+      "loss": 1.3794,
+      "step": 1208
+    },
+    {
+      "epoch": 1.7822001105583194,
+      "grad_norm": 1.921731214649604,
+      "learning_rate": 4.096669098524451e-05,
+      "loss": 1.2974,
+      "step": 1209
+    },
+    {
+      "epoch": 1.7836742214851666,
+      "grad_norm": 1.6931901295000011,
+      "learning_rate": 4.095145537998972e-05,
+      "loss": 1.0482,
+      "step": 1210
+    },
+    {
+      "epoch": 1.7851483324120139,
+      "grad_norm": 1.7316180474900313,
+      "learning_rate": 4.0936209775731686e-05,
+      "loss": 1.1968,
+      "step": 1211
+    },
+    {
+      "epoch": 1.786622443338861,
+      "grad_norm": 2.0403582084538323,
+      "learning_rate": 4.0920954182026965e-05,
+      "loss": 1.3086,
+      "step": 1212
+    },
+    {
+      "epoch": 1.7880965542657083,
+      "grad_norm": 1.9066816372828312,
+      "learning_rate": 4.090568860843836e-05,
+      "loss": 1.1558,
+      "step": 1213
+    },
+    {
+      "epoch": 1.7895706651925556,
+      "grad_norm": 2.127502426522616,
+      "learning_rate": 4.089041306453494e-05,
+      "loss": 1.1972,
+      "step": 1214
+    },
+    {
+      "epoch": 1.7910447761194028,
+      "grad_norm": 2.0013660295769196,
+      "learning_rate": 4.0875127559892015e-05,
+      "loss": 1.2047,
+      "step": 1215
+    },
+    {
+      "epoch": 1.79251888704625,
+      "grad_norm": 1.9904459203537852,
+      "learning_rate": 4.085983210409114e-05,
+      "loss": 1.2446,
+      "step": 1216
+    },
+    {
+      "epoch": 1.7939929979730975,
+      "grad_norm": 1.814407791540448,
+      "learning_rate": 4.084452670672012e-05,
+      "loss": 1.3015,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7954671088999448,
+      "grad_norm": 1.8119198100722682,
+      "learning_rate": 4.082921137737299e-05,
+      "loss": 1.3315,
+      "step": 1218
+    },
+    {
+      "epoch": 1.796941219826792,
+      "grad_norm": 2.0153242966254457,
+      "learning_rate": 4.081388612564999e-05,
+      "loss": 1.2139,
+      "step": 1219
+    },
+    {
+      "epoch": 1.7984153307536392,
+      "grad_norm": 1.837036193648923,
+      "learning_rate": 4.07985509611576e-05,
+      "loss": 1.2622,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7998894416804865,
+      "grad_norm": 1.8516436208989888,
+      "learning_rate": 4.078320589350851e-05,
+      "loss": 1.306,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8013635526073337,
+      "grad_norm": 1.7578483747424278,
+      "learning_rate": 4.076785093232162e-05,
+      "loss": 1.2373,
+      "step": 1222
+    },
+    {
+      "epoch": 1.802837663534181,
+      "grad_norm": 1.8275956259573485,
+      "learning_rate": 4.0752486087222006e-05,
+      "loss": 1.2986,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8043117744610282,
+      "grad_norm": 2.0905801491708997,
+      "learning_rate": 4.073711136784099e-05,
+      "loss": 1.2647,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8057858853878754,
+      "grad_norm": 2.008867157496913,
+      "learning_rate": 4.072172678381603e-05,
+      "loss": 1.3697,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8072599963147227,
+      "grad_norm": 1.840144560048702,
+      "learning_rate": 4.07063323447908e-05,
+      "loss": 1.1796,
+      "step": 1226
+    },
+    {
+      "epoch": 1.80873410724157,
+      "grad_norm": 1.7366482382740513,
+      "learning_rate": 4.0690928060415144e-05,
+      "loss": 1.1902,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8102082181684171,
+      "grad_norm": 1.951229121348806,
+      "learning_rate": 4.067551394034508e-05,
+      "loss": 1.2459,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8116823290952644,
+      "grad_norm": 1.9515362190327847,
+      "learning_rate": 4.066008999424279e-05,
+      "loss": 1.4361,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8131564400221116,
+      "grad_norm": 1.804719404458753,
+      "learning_rate": 4.06446562317766e-05,
+      "loss": 1.0962,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8146305509489589,
+      "grad_norm": 1.9655295780957824,
+      "learning_rate": 4.062921266262102e-05,
+      "loss": 1.3538,
+      "step": 1231
+    },
+    {
+      "epoch": 1.816104661875806,
+      "grad_norm": 1.8144401820843206,
+      "learning_rate": 4.0613759296456675e-05,
+      "loss": 1.2675,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8175787728026536,
+      "grad_norm": 1.837893803235994,
+      "learning_rate": 4.059829614297036e-05,
+      "loss": 1.2578,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8190528837295008,
+      "grad_norm": 1.8601099693440113,
+      "learning_rate": 4.058282321185498e-05,
+      "loss": 1.2072,
+      "step": 1234
+    },
+    {
+      "epoch": 1.820526994656348,
+      "grad_norm": 1.9126803643440953,
+      "learning_rate": 4.0567340512809586e-05,
+      "loss": 1.3911,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8220011055831953,
+      "grad_norm": 1.836488941245533,
+      "learning_rate": 4.0551848055539345e-05,
+      "loss": 1.1442,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8234752165100425,
+      "grad_norm": 2.200785592006832,
+      "learning_rate": 4.0536345849755545e-05,
+      "loss": 1.4227,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8249493274368898,
+      "grad_norm": 2.0864498352192595,
+      "learning_rate": 4.0520833905175576e-05,
+      "loss": 1.3014,
+      "step": 1238
+    },
+    {
+      "epoch": 1.826423438363737,
+      "grad_norm": 2.0893770926850452,
+      "learning_rate": 4.0505312231522944e-05,
+      "loss": 1.2053,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8278975492905842,
+      "grad_norm": 1.9714247683770754,
+      "learning_rate": 4.048978083852724e-05,
+      "loss": 1.3151,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8293716602174315,
+      "grad_norm": 1.887759788268554,
+      "learning_rate": 4.0474239735924166e-05,
+      "loss": 1.1569,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8308457711442787,
+      "grad_norm": 1.8476009562245232,
+      "learning_rate": 4.045868893345549e-05,
+      "loss": 1.2439,
+      "step": 1242
+    },
+    {
+      "epoch": 1.832319882071126,
+      "grad_norm": 2.0706931861821705,
+      "learning_rate": 4.0443128440869084e-05,
+      "loss": 1.3339,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8337939929979732,
+      "grad_norm": 1.7172358692635747,
+      "learning_rate": 4.042755826791886e-05,
+      "loss": 1.0632,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8352681039248204,
+      "grad_norm": 1.9403614325918679,
+      "learning_rate": 4.041197842436484e-05,
+      "loss": 1.2979,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8367422148516677,
+      "grad_norm": 1.9239620258961594,
+      "learning_rate": 4.0396388919973074e-05,
+      "loss": 1.1112,
+      "step": 1246
+    },
+    {
+      "epoch": 1.838216325778515,
+      "grad_norm": 1.864200900904161,
+      "learning_rate": 4.038078976451567e-05,
+      "loss": 1.2435,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8396904367053621,
+      "grad_norm": 1.7706432764124702,
+      "learning_rate": 4.036518096777082e-05,
+      "loss": 1.1653,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8411645476322094,
+      "grad_norm": 1.8064599519033935,
+      "learning_rate": 4.0349562539522725e-05,
+      "loss": 1.2137,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8426386585590566,
+      "grad_norm": 1.8003111782244594,
+      "learning_rate": 4.033393448956162e-05,
+      "loss": 1.1869,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8426386585590566,
+      "eval_bleu": 0.048702265004150425,
+      "eval_bleu_1gram": 0.34895182098107314,
+      "eval_bleu_2gram": 0.12077243209842681,
+      "eval_bleu_3gram": 0.04874536480753008,
+      "eval_bleu_4gram": 0.022646375855692985,
+      "eval_rag_val_loss": 1.5452378831243003,
+      "eval_rouge1": 0.3332082488018953,
+      "eval_rouge2": 0.11316398251026843,
+      "eval_rougeL": 0.3126958399847377,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8441127694859039,
+      "grad_norm": 1.948282081750063,
+      "learning_rate": 4.03182968276838e-05,
+      "loss": 1.1966,
+      "step": 1251
+    },
+    {
+      "epoch": 1.845586880412751,
+      "grad_norm": 1.8194222954460737,
+      "learning_rate": 4.030264956369157e-05,
+      "loss": 1.1786,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8470609913395983,
+      "grad_norm": 1.7283532017835233,
+      "learning_rate": 4.028699270739326e-05,
+      "loss": 1.1096,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8485351022664456,
+      "grad_norm": 1.8168961059268236,
+      "learning_rate": 4.027132626860318e-05,
+      "loss": 1.1748,
+      "step": 1254
+    },
+    {
+      "epoch": 1.8500092131932928,
+      "grad_norm": 1.9992881938743918,
+      "learning_rate": 4.02556502571417e-05,
+      "loss": 1.3077,
+      "step": 1255
+    },
+    {
+      "epoch": 1.85148332412014,
+      "grad_norm": 2.0348938170527933,
+      "learning_rate": 4.023996468283515e-05,
+      "loss": 1.1406,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8529574350469873,
+      "grad_norm": 1.8747836941881846,
+      "learning_rate": 4.022426955551588e-05,
+      "loss": 1.1618,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8544315459738345,
+      "grad_norm": 1.7492383934509335,
+      "learning_rate": 4.020856488502221e-05,
+      "loss": 1.1624,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8559056569006818,
+      "grad_norm": 1.8520217519104791,
+      "learning_rate": 4.019285068119845e-05,
+      "loss": 1.3805,
+      "step": 1259
+    },
+    {
+      "epoch": 1.857379767827529,
+      "grad_norm": 2.0601375806105042,
+      "learning_rate": 4.017712695389487e-05,
+      "loss": 1.2948,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8588538787543762,
+      "grad_norm": 1.7210131224154699,
+      "learning_rate": 4.0161393712967756e-05,
+      "loss": 1.1627,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8603279896812235,
+      "grad_norm": 1.803741206982409,
+      "learning_rate": 4.01456509682793e-05,
+      "loss": 1.2001,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8618021006080707,
+      "grad_norm": 1.871970526752597,
+      "learning_rate": 4.012989872969768e-05,
+      "loss": 1.2769,
+      "step": 1263
+    },
+    {
+      "epoch": 1.863276211534918,
+      "grad_norm": 1.8615618115741301,
+      "learning_rate": 4.011413700709703e-05,
+      "loss": 1.1997,
+      "step": 1264
+    },
+    {
+      "epoch": 1.8647503224617652,
+      "grad_norm": 1.801478054585274,
+      "learning_rate": 4.009836581035742e-05,
+      "loss": 1.1766,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8662244333886124,
+      "grad_norm": 1.940315416001795,
+      "learning_rate": 4.008258514936486e-05,
+      "loss": 1.2656,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8676985443154597,
+      "grad_norm": 1.84795430462972,
+      "learning_rate": 4.006679503401129e-05,
+      "loss": 1.2304,
+      "step": 1267
+    },
+    {
+      "epoch": 1.869172655242307,
+      "grad_norm": 2.0210821516355213,
+      "learning_rate": 4.0050995474194576e-05,
+      "loss": 1.387,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8706467661691542,
+      "grad_norm": 1.8351643983961288,
+      "learning_rate": 4.003518647981852e-05,
+      "loss": 1.3058,
+      "step": 1269
+    },
+    {
+      "epoch": 1.8721208770960014,
+      "grad_norm": 1.8960338807841222,
+      "learning_rate": 4.0019368060792806e-05,
+      "loss": 1.0698,
+      "step": 1270
+    },
+    {
+      "epoch": 1.8735949880228486,
+      "grad_norm": 1.7714793952779375,
+      "learning_rate": 4.000354022703306e-05,
+      "loss": 1.2642,
+      "step": 1271
+    },
+    {
+      "epoch": 1.8750690989496959,
+      "grad_norm": 1.9631888547347802,
+      "learning_rate": 3.998770298846079e-05,
+      "loss": 1.1407,
+      "step": 1272
+    },
+    {
+      "epoch": 1.876543209876543,
+      "grad_norm": 1.8885628669784618,
+      "learning_rate": 3.9971856355003396e-05,
+      "loss": 1.2952,
+      "step": 1273
+    },
+    {
+      "epoch": 1.8780173208033903,
+      "grad_norm": 1.9533328746799457,
+      "learning_rate": 3.9956000336594185e-05,
+      "loss": 1.2955,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8794914317302376,
+      "grad_norm": 1.9145949265464923,
+      "learning_rate": 3.994013494317233e-05,
+      "loss": 1.3606,
+      "step": 1275
+    },
+    {
+      "epoch": 1.8809655426570848,
+      "grad_norm": 2.015132873546277,
+      "learning_rate": 3.9924260184682894e-05,
+      "loss": 1.282,
+      "step": 1276
+    },
+    {
+      "epoch": 1.882439653583932,
+      "grad_norm": 2.19542131799087,
+      "learning_rate": 3.9908376071076805e-05,
+      "loss": 1.2211,
+      "step": 1277
+    },
+    {
+      "epoch": 1.8839137645107793,
+      "grad_norm": 1.7760036236966599,
+      "learning_rate": 3.9892482612310836e-05,
+      "loss": 1.0788,
+      "step": 1278
+    },
+    {
+      "epoch": 1.8853878754376265,
+      "grad_norm": 1.8199160557555802,
+      "learning_rate": 3.9876579818347654e-05,
+      "loss": 1.1686,
+      "step": 1279
+    },
+    {
+      "epoch": 1.8868619863644738,
+      "grad_norm": 2.018995438999959,
+      "learning_rate": 3.986066769915575e-05,
+      "loss": 1.2312,
+      "step": 1280
+    },
+    {
+      "epoch": 1.888336097291321,
+      "grad_norm": 1.8573979357150452,
+      "learning_rate": 3.984474626470948e-05,
+      "loss": 1.3486,
+      "step": 1281
+    },
+    {
+      "epoch": 1.8898102082181683,
+      "grad_norm": 2.038570183847055,
+      "learning_rate": 3.982881552498902e-05,
+      "loss": 1.2178,
+      "step": 1282
+    },
+    {
+      "epoch": 1.8912843191450157,
+      "grad_norm": 1.7797308349692056,
+      "learning_rate": 3.981287548998039e-05,
+      "loss": 1.3454,
+      "step": 1283
+    },
+    {
+      "epoch": 1.892758430071863,
+      "grad_norm": 1.8065937098004767,
+      "learning_rate": 3.979692616967543e-05,
+      "loss": 1.1505,
+      "step": 1284
+    },
+    {
+      "epoch": 1.8942325409987102,
+      "grad_norm": 1.9516186818327408,
+      "learning_rate": 3.978096757407182e-05,
+      "loss": 1.3331,
+      "step": 1285
+    },
+    {
+      "epoch": 1.8957066519255574,
+      "grad_norm": 2.1568726456086833,
+      "learning_rate": 3.976499971317302e-05,
+      "loss": 1.2342,
+      "step": 1286
+    },
+    {
+      "epoch": 1.8971807628524047,
+      "grad_norm": 1.8644989321340673,
+      "learning_rate": 3.974902259698833e-05,
+      "loss": 1.4244,
+      "step": 1287
+    },
+    {
+      "epoch": 1.898654873779252,
+      "grad_norm": 1.8607895543298092,
+      "learning_rate": 3.973303623553283e-05,
+      "loss": 1.231,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9001289847060991,
+      "grad_norm": 1.7732760990319096,
+      "learning_rate": 3.9717040638827406e-05,
+      "loss": 1.212,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9016030956329464,
+      "grad_norm": 1.6088192174525329,
+      "learning_rate": 3.9701035816898734e-05,
+      "loss": 1.167,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9030772065597936,
+      "grad_norm": 1.783607328465992,
+      "learning_rate": 3.9685021779779264e-05,
+      "loss": 1.3402,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9045513174866409,
+      "grad_norm": 1.715458648899289,
+      "learning_rate": 3.966899853750724e-05,
+      "loss": 1.3654,
+      "step": 1292
+    },
+    {
+      "epoch": 1.906025428413488,
+      "grad_norm": 1.7972168970332127,
+      "learning_rate": 3.9652966100126655e-05,
+      "loss": 1.1601,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9074995393403353,
+      "grad_norm": 1.7842336060843975,
+      "learning_rate": 3.9636924477687265e-05,
+      "loss": 1.2925,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9089736502671826,
+      "grad_norm": 1.8458450827931445,
+      "learning_rate": 3.9620873680244616e-05,
+      "loss": 1.3226,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9104477611940298,
+      "grad_norm": 1.8390891437010344,
+      "learning_rate": 3.960481371785997e-05,
+      "loss": 1.1871,
+      "step": 1296
+    },
+    {
+      "epoch": 1.911921872120877,
+      "grad_norm": 1.7721933041959994,
+      "learning_rate": 3.958874460060035e-05,
+      "loss": 1.1808,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9133959830477243,
+      "grad_norm": 1.7477844382259737,
+      "learning_rate": 3.95726663385385e-05,
+      "loss": 1.1591,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9148700939745718,
+      "grad_norm": 1.9245669113067494,
+      "learning_rate": 3.955657894175293e-05,
+      "loss": 1.3242,
+      "step": 1299
+    },
+    {
+      "epoch": 1.916344204901419,
+      "grad_norm": 1.8511991506492163,
+      "learning_rate": 3.9540482420327845e-05,
+      "loss": 1.1912,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9178183158282662,
+      "grad_norm": 2.2430300309258033,
+      "learning_rate": 3.952437678435319e-05,
+      "loss": 1.2118,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9192924267551135,
+      "grad_norm": 1.9149109497480785,
+      "learning_rate": 3.950826204392461e-05,
+      "loss": 1.2133,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9207665376819607,
+      "grad_norm": 1.6859895516068424,
+      "learning_rate": 3.949213820914347e-05,
+      "loss": 1.2436,
+      "step": 1303
+    },
+    {
+      "epoch": 1.922240648608808,
+      "grad_norm": 2.03249063018103,
+      "learning_rate": 3.9476005290116814e-05,
+      "loss": 1.2645,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9237147595356552,
+      "grad_norm": 1.812264328466209,
+      "learning_rate": 3.94598632969574e-05,
+      "loss": 1.1948,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9251888704625024,
+      "grad_norm": 1.8221580089262344,
+      "learning_rate": 3.944371223978366e-05,
+      "loss": 1.1905,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9266629813893497,
+      "grad_norm": 2.0170351758394625,
+      "learning_rate": 3.942755212871973e-05,
+      "loss": 1.2373,
+      "step": 1307
+    },
+    {
+      "epoch": 1.928137092316197,
+      "grad_norm": 1.8732271714295254,
+      "learning_rate": 3.94113829738954e-05,
+      "loss": 1.1487,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9296112032430441,
+      "grad_norm": 1.818437519299654,
+      "learning_rate": 3.939520478544614e-05,
+      "loss": 1.3074,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9310853141698914,
+      "grad_norm": 1.803439811317709,
+      "learning_rate": 3.937901757351307e-05,
+      "loss": 1.3,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9325594250967386,
+      "grad_norm": 1.9745017908145668,
+      "learning_rate": 3.936282134824297e-05,
+      "loss": 1.1545,
+      "step": 1311
+    },
+    {
+      "epoch": 1.9340335360235859,
+      "grad_norm": 1.8393019346138344,
+      "learning_rate": 3.93466161197883e-05,
+      "loss": 1.3737,
+      "step": 1312
+    },
+    {
+      "epoch": 1.935507646950433,
+      "grad_norm": 1.7163009534858393,
+      "learning_rate": 3.933040189830711e-05,
+      "loss": 1.0954,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9369817578772803,
+      "grad_norm": 1.8541571227553122,
+      "learning_rate": 3.931417869396313e-05,
+      "loss": 1.2255,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9384558688041276,
+      "grad_norm": 1.7203077712933845,
+      "learning_rate": 3.929794651692571e-05,
+      "loss": 1.1313,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9399299797309748,
+      "grad_norm": 1.8044675882947914,
+      "learning_rate": 3.928170537736981e-05,
+      "loss": 1.1159,
+      "step": 1316
+    },
+    {
+      "epoch": 1.941404090657822,
+      "grad_norm": 1.6383659431857382,
+      "learning_rate": 3.9265455285476025e-05,
+      "loss": 1.1734,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9428782015846693,
+      "grad_norm": 1.7968349949903029,
+      "learning_rate": 3.9249196251430556e-05,
+      "loss": 1.3347,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9443523125115165,
+      "grad_norm": 1.9054955412172647,
+      "learning_rate": 3.92329282854252e-05,
+      "loss": 1.4535,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9458264234383638,
+      "grad_norm": 2.0137488334742364,
+      "learning_rate": 3.9216651397657364e-05,
+      "loss": 1.4461,
+      "step": 1320
+    },
+    {
+      "epoch": 1.947300534365211,
+      "grad_norm": 2.0266086076066325,
+      "learning_rate": 3.9200365598330056e-05,
+      "loss": 1.3215,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9487746452920582,
+      "grad_norm": 1.943006934023513,
+      "learning_rate": 3.9184070897651854e-05,
+      "loss": 1.383,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9502487562189055,
+      "grad_norm": 1.9190659502076364,
+      "learning_rate": 3.916776730583691e-05,
+      "loss": 1.1812,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9517228671457527,
+      "grad_norm": 2.111401072326197,
+      "learning_rate": 3.915145483310498e-05,
+      "loss": 1.322,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9531969780726,
+      "grad_norm": 1.9477670037476407,
+      "learning_rate": 3.9135133489681356e-05,
+      "loss": 1.2421,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9546710889994472,
+      "grad_norm": 1.8113719455271258,
+      "learning_rate": 3.91188032857969e-05,
+      "loss": 1.1733,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9561451999262944,
+      "grad_norm": 1.8227026968220015,
+      "learning_rate": 3.910246423168803e-05,
+      "loss": 1.2431,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9576193108531417,
+      "grad_norm": 2.092538853401969,
+      "learning_rate": 3.908611633759672e-05,
+      "loss": 1.3122,
+      "step": 1328
+    },
+    {
+      "epoch": 1.959093421779989,
+      "grad_norm": 1.8688696143612555,
+      "learning_rate": 3.906975961377046e-05,
+      "loss": 1.3971,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9605675327068361,
+      "grad_norm": 1.9890412262814996,
+      "learning_rate": 3.905339407046231e-05,
+      "loss": 1.4756,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9620416436336834,
+      "grad_norm": 2.012499019077607,
+      "learning_rate": 3.9037019717930826e-05,
+      "loss": 1.4202,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9635157545605306,
+      "grad_norm": 1.8199782821910526,
+      "learning_rate": 3.902063656644012e-05,
+      "loss": 1.2034,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9649898654873779,
+      "grad_norm": 1.7866698011090887,
+      "learning_rate": 3.900424462625977e-05,
+      "loss": 1.293,
+      "step": 1333
+    },
+    {
+      "epoch": 1.966463976414225,
+      "grad_norm": 1.7856965363846546,
+      "learning_rate": 3.898784390766491e-05,
+      "loss": 1.2813,
+      "step": 1334
+    },
+    {
+      "epoch": 1.9679380873410723,
+      "grad_norm": 1.7445537423897435,
+      "learning_rate": 3.897143442093616e-05,
+      "loss": 1.2391,
+      "step": 1335
+    },
+    {
+      "epoch": 1.9694121982679196,
+      "grad_norm": 1.7176120805886745,
+      "learning_rate": 3.895501617635964e-05,
+      "loss": 1.0774,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9708863091947668,
+      "grad_norm": 1.802678959781068,
+      "learning_rate": 3.893858918422693e-05,
+      "loss": 1.241,
+      "step": 1337
+    },
+    {
+      "epoch": 1.972360420121614,
+      "grad_norm": 1.7748554627577509,
+      "learning_rate": 3.892215345483515e-05,
+      "loss": 1.1883,
+      "step": 1338
+    },
+    {
+      "epoch": 1.9738345310484613,
+      "grad_norm": 1.9104180965882762,
+      "learning_rate": 3.890570899848685e-05,
+      "loss": 1.4005,
+      "step": 1339
+    },
+    {
+      "epoch": 1.9753086419753085,
+      "grad_norm": 1.7871510651942184,
+      "learning_rate": 3.888925582549006e-05,
+      "loss": 1.2861,
+      "step": 1340
+    },
+    {
+      "epoch": 1.9767827529021558,
+      "grad_norm": 1.7658735792330242,
+      "learning_rate": 3.887279394615829e-05,
+      "loss": 1.2319,
+      "step": 1341
+    },
+    {
+      "epoch": 1.978256863829003,
+      "grad_norm": 1.840935385266865,
+      "learning_rate": 3.885632337081049e-05,
+      "loss": 1.277,
+      "step": 1342
+    },
+    {
+      "epoch": 1.9797309747558502,
+      "grad_norm": 1.7053744658315741,
+      "learning_rate": 3.8839844109771086e-05,
+      "loss": 1.1554,
+      "step": 1343
+    },
+    {
+      "epoch": 1.9812050856826975,
+      "grad_norm": 1.8866519639730726,
+      "learning_rate": 3.8823356173369895e-05,
+      "loss": 1.403,
+      "step": 1344
+    },
+    {
+      "epoch": 1.9826791966095447,
+      "grad_norm": 1.626864610624901,
+      "learning_rate": 3.8806859571942244e-05,
+      "loss": 1.2483,
+      "step": 1345
+    },
+    {
+      "epoch": 1.984153307536392,
+      "grad_norm": 1.8965975789502902,
+      "learning_rate": 3.8790354315828846e-05,
+      "loss": 1.2473,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9856274184632392,
+      "grad_norm": 1.7827270139614713,
+      "learning_rate": 3.877384041537584e-05,
+      "loss": 1.4432,
+      "step": 1347
+    },
+    {
+      "epoch": 1.9871015293900864,
+      "grad_norm": 2.042235029348997,
+      "learning_rate": 3.8757317880934786e-05,
+      "loss": 1.3813,
+      "step": 1348
+    },
+    {
+      "epoch": 1.988575640316934,
+      "grad_norm": 1.7606918827091844,
+      "learning_rate": 3.8740786722862676e-05,
+      "loss": 1.1214,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9900497512437811,
+      "grad_norm": 1.8107840208854808,
+      "learning_rate": 3.872424695152189e-05,
+      "loss": 1.2968,
+      "step": 1350
+    },
+    {
+      "epoch": 1.9915238621706284,
+      "grad_norm": 1.9953451227136578,
+      "learning_rate": 3.870769857728022e-05,
+      "loss": 1.2203,
+      "step": 1351
+    },
+    {
+      "epoch": 1.9929979730974756,
+      "grad_norm": 1.9744930365124012,
+      "learning_rate": 3.869114161051082e-05,
+      "loss": 1.2328,
+      "step": 1352
+    },
+    {
+      "epoch": 1.9944720840243229,
+      "grad_norm": 1.8419812657276473,
+      "learning_rate": 3.867457606159226e-05,
+      "loss": 1.1942,
+      "step": 1353
+    },
+    {
+      "epoch": 1.99594619495117,
+      "grad_norm": 1.7611947305859108,
+      "learning_rate": 3.86580019409085e-05,
+      "loss": 1.1931,
+      "step": 1354
+    },
+    {
+      "epoch": 1.9974203058780173,
+      "grad_norm": 1.6344202330768596,
+      "learning_rate": 3.8641419258848835e-05,
+      "loss": 1.0649,
+      "step": 1355
+    },
+    {
+      "epoch": 1.9988944168048646,
+      "grad_norm": 1.9735311550892811,
+      "learning_rate": 3.862482802580795e-05,
+      "loss": 1.2363,
+      "step": 1356
+    },
+    {
+      "epoch": 2.000368527731712,
+      "grad_norm": 1.8442219275693046,
+      "learning_rate": 3.860822825218588e-05,
+      "loss": 1.1441,
+      "step": 1357
+    },
+    {
+      "epoch": 2.0018426386585593,
+      "grad_norm": 1.7857747748892803,
+      "learning_rate": 3.859161994838803e-05,
+      "loss": 1.0668,
+      "step": 1358
+    },
+    {
+      "epoch": 2.0033167495854065,
+      "grad_norm": 1.7010962626036672,
+      "learning_rate": 3.8575003124825135e-05,
+      "loss": 0.8501,
+      "step": 1359
+    },
+    {
+      "epoch": 2.0047908605122537,
+      "grad_norm": 1.7876735442953533,
+      "learning_rate": 3.855837779191329e-05,
+      "loss": 0.7953,
+      "step": 1360
+    },
+    {
+      "epoch": 2.006264971439101,
+      "grad_norm": 1.7058387624073088,
+      "learning_rate": 3.8541743960073893e-05,
+      "loss": 0.9207,
+      "step": 1361
+    },
+    {
+      "epoch": 2.0077390823659482,
+      "grad_norm": 1.6804007501823424,
+      "learning_rate": 3.8525101639733706e-05,
+      "loss": 0.9173,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0092131932927955,
+      "grad_norm": 1.8163975212953913,
+      "learning_rate": 3.850845084132478e-05,
+      "loss": 0.8751,
+      "step": 1363
+    },
+    {
+      "epoch": 2.0106873042196427,
+      "grad_norm": 1.5901131366568777,
+      "learning_rate": 3.84917915752845e-05,
+      "loss": 1.0104,
+      "step": 1364
+    },
+    {
+      "epoch": 2.01216141514649,
+      "grad_norm": 1.655024020778441,
+      "learning_rate": 3.847512385205556e-05,
+      "loss": 0.7875,
+      "step": 1365
+    },
+    {
+      "epoch": 2.013635526073337,
+      "grad_norm": 1.7146582709522675,
+      "learning_rate": 3.845844768208593e-05,
+      "loss": 0.9048,
+      "step": 1366
+    },
+    {
+      "epoch": 2.0151096370001844,
+      "grad_norm": 1.9761531843276894,
+      "learning_rate": 3.8441763075828904e-05,
+      "loss": 0.8797,
+      "step": 1367
+    },
+    {
+      "epoch": 2.0165837479270317,
+      "grad_norm": 1.972628693340942,
+      "learning_rate": 3.842507004374304e-05,
+      "loss": 0.8751,
+      "step": 1368
+    },
+    {
+      "epoch": 2.018057858853879,
+      "grad_norm": 1.9838013908782093,
+      "learning_rate": 3.8408368596292224e-05,
+      "loss": 0.9271,
+      "step": 1369
+    },
+    {
+      "epoch": 2.019531969780726,
+      "grad_norm": 2.0103643802570343,
+      "learning_rate": 3.839165874394555e-05,
+      "loss": 0.8759,
+      "step": 1370
+    },
+    {
+      "epoch": 2.0210060807075734,
+      "grad_norm": 2.082068924267785,
+      "learning_rate": 3.8374940497177434e-05,
+      "loss": 0.9728,
+      "step": 1371
+    },
+    {
+      "epoch": 2.0224801916344206,
+      "grad_norm": 2.1116217057104105,
+      "learning_rate": 3.835821386646753e-05,
+      "loss": 0.9845,
+      "step": 1372
+    },
+    {
+      "epoch": 2.023954302561268,
+      "grad_norm": 2.0508596060179722,
+      "learning_rate": 3.834147886230074e-05,
+      "loss": 0.9927,
+      "step": 1373
+    },
+    {
+      "epoch": 2.025428413488115,
+      "grad_norm": 1.997013842989832,
+      "learning_rate": 3.8324735495167246e-05,
+      "loss": 1.0003,
+      "step": 1374
+    },
+    {
+      "epoch": 2.0269025244149623,
+      "grad_norm": 2.1340776540754716,
+      "learning_rate": 3.8307983775562435e-05,
+      "loss": 0.891,
+      "step": 1375
+    },
+    {
+      "epoch": 2.0283766353418096,
+      "grad_norm": 2.1687888853781043,
+      "learning_rate": 3.8291223713986955e-05,
+      "loss": 0.8061,
+      "step": 1376
+    },
+    {
+      "epoch": 2.029850746268657,
+      "grad_norm": 2.18468446735995,
+      "learning_rate": 3.827445532094669e-05,
+      "loss": 0.9903,
+      "step": 1377
+    },
+    {
+      "epoch": 2.031324857195504,
+      "grad_norm": 2.0818416086827014,
+      "learning_rate": 3.8257678606952705e-05,
+      "loss": 0.8366,
+      "step": 1378
+    },
+    {
+      "epoch": 2.0327989681223513,
+      "grad_norm": 2.0400775122408774,
+      "learning_rate": 3.824089358252133e-05,
+      "loss": 0.9875,
+      "step": 1379
+    },
+    {
+      "epoch": 2.0342730790491985,
+      "grad_norm": 2.183074834920526,
+      "learning_rate": 3.822410025817406e-05,
+      "loss": 0.8794,
+      "step": 1380
+    },
+    {
+      "epoch": 2.0357471899760458,
+      "grad_norm": 2.0378284447053066,
+      "learning_rate": 3.820729864443764e-05,
+      "loss": 0.789,
+      "step": 1381
+    },
+    {
+      "epoch": 2.037221300902893,
+      "grad_norm": 1.9510525503701919,
+      "learning_rate": 3.819048875184398e-05,
+      "loss": 0.8298,
+      "step": 1382
+    },
+    {
+      "epoch": 2.0386954118297402,
+      "grad_norm": 2.022089328637764,
+      "learning_rate": 3.8173670590930165e-05,
+      "loss": 0.7973,
+      "step": 1383
+    },
+    {
+      "epoch": 2.0401695227565875,
+      "grad_norm": 2.1023363014749834,
+      "learning_rate": 3.815684417223851e-05,
+      "loss": 0.9196,
+      "step": 1384
+    },
+    {
+      "epoch": 2.0416436336834347,
+      "grad_norm": 2.0829569921718964,
+      "learning_rate": 3.814000950631647e-05,
+      "loss": 0.9855,
+      "step": 1385
+    },
+    {
+      "epoch": 2.043117744610282,
+      "grad_norm": 2.1110524607202072,
+      "learning_rate": 3.812316660371666e-05,
+      "loss": 0.856,
+      "step": 1386
+    },
+    {
+      "epoch": 2.044591855537129,
+      "grad_norm": 2.163798953097036,
+      "learning_rate": 3.810631547499692e-05,
+      "loss": 0.8027,
+      "step": 1387
+    },
+    {
+      "epoch": 2.0460659664639764,
+      "grad_norm": 1.881829349365738,
+      "learning_rate": 3.808945613072017e-05,
+      "loss": 0.7646,
+      "step": 1388
+    },
+    {
+      "epoch": 2.0475400773908237,
+      "grad_norm": 1.986875025535312,
+      "learning_rate": 3.807258858145453e-05,
+      "loss": 0.9371,
+      "step": 1389
+    },
+    {
+      "epoch": 2.049014188317671,
+      "grad_norm": 1.9492742801928584,
+      "learning_rate": 3.8055712837773225e-05,
+      "loss": 0.9211,
+      "step": 1390
+    },
+    {
+      "epoch": 2.050488299244518,
+      "grad_norm": 2.028289985705516,
+      "learning_rate": 3.803882891025466e-05,
+      "loss": 0.8203,
+      "step": 1391
+    },
+    {
+      "epoch": 2.0519624101713654,
+      "grad_norm": 1.9081633923863692,
+      "learning_rate": 3.802193680948236e-05,
+      "loss": 0.8979,
+      "step": 1392
+    },
+    {
+      "epoch": 2.0534365210982126,
+      "grad_norm": 2.1828756226333,
+      "learning_rate": 3.800503654604493e-05,
+      "loss": 0.82,
+      "step": 1393
+    },
+    {
+      "epoch": 2.05491063202506,
+      "grad_norm": 1.972326209531169,
+      "learning_rate": 3.798812813053615e-05,
+      "loss": 0.7363,
+      "step": 1394
+    },
+    {
+      "epoch": 2.056384742951907,
+      "grad_norm": 1.788951753745132,
+      "learning_rate": 3.7971211573554865e-05,
+      "loss": 0.9638,
+      "step": 1395
+    },
+    {
+      "epoch": 2.0578588538787543,
+      "grad_norm": 2.0160018213808826,
+      "learning_rate": 3.795428688570505e-05,
+      "loss": 0.8725,
+      "step": 1396
+    },
+    {
+      "epoch": 2.0593329648056016,
+      "grad_norm": 2.041376197030208,
+      "learning_rate": 3.793735407759577e-05,
+      "loss": 0.8207,
+      "step": 1397
+    },
+    {
+      "epoch": 2.060807075732449,
+      "grad_norm": 2.18431229617341,
+      "learning_rate": 3.792041315984118e-05,
+      "loss": 0.9525,
+      "step": 1398
+    },
+    {
+      "epoch": 2.062281186659296,
+      "grad_norm": 2.0891993019203627,
+      "learning_rate": 3.7903464143060506e-05,
+      "loss": 0.8858,
+      "step": 1399
+    },
+    {
+      "epoch": 2.0637552975861433,
+      "grad_norm": 2.1831610017085556,
+      "learning_rate": 3.788650703787808e-05,
+      "loss": 1.0065,
+      "step": 1400
+    },
+    {
+      "epoch": 2.0652294085129905,
+      "grad_norm": 1.8813815555642048,
+      "learning_rate": 3.7869541854923275e-05,
+      "loss": 0.8553,
+      "step": 1401
+    },
+    {
+      "epoch": 2.0667035194398378,
+      "grad_norm": 2.028008794761714,
+      "learning_rate": 3.785256860483054e-05,
+      "loss": 0.8109,
+      "step": 1402
+    },
+    {
+      "epoch": 2.068177630366685,
+      "grad_norm": 2.3541636734214815,
+      "learning_rate": 3.783558729823939e-05,
+      "loss": 1.0432,
+      "step": 1403
+    },
+    {
+      "epoch": 2.0696517412935322,
+      "grad_norm": 2.077943492792988,
+      "learning_rate": 3.781859794579436e-05,
+      "loss": 0.8413,
+      "step": 1404
+    },
+    {
+      "epoch": 2.0711258522203795,
+      "grad_norm": 2.1728780063348743,
+      "learning_rate": 3.780160055814507e-05,
+      "loss": 0.9425,
+      "step": 1405
+    },
+    {
+      "epoch": 2.0725999631472267,
+      "grad_norm": 2.134764954897539,
+      "learning_rate": 3.778459514594613e-05,
+      "loss": 0.8866,
+      "step": 1406
+    },
+    {
+      "epoch": 2.074074074074074,
+      "grad_norm": 2.074411616724855,
+      "learning_rate": 3.776758171985723e-05,
+      "loss": 0.781,
+      "step": 1407
+    },
+    {
+      "epoch": 2.075548185000921,
+      "grad_norm": 2.194043861986983,
+      "learning_rate": 3.775056029054304e-05,
+      "loss": 0.9234,
+      "step": 1408
+    },
+    {
+      "epoch": 2.0770222959277684,
+      "grad_norm": 2.157607604843163,
+      "learning_rate": 3.773353086867328e-05,
+      "loss": 0.8635,
+      "step": 1409
+    },
+    {
+      "epoch": 2.0784964068546157,
+      "grad_norm": 1.9892615279405041,
+      "learning_rate": 3.7716493464922654e-05,
+      "loss": 0.9011,
+      "step": 1410
+    },
+    {
+      "epoch": 2.079970517781463,
+      "grad_norm": 1.9791073907374646,
+      "learning_rate": 3.769944808997088e-05,
+      "loss": 0.8467,
+      "step": 1411
+    },
+    {
+      "epoch": 2.08144462870831,
+      "grad_norm": 2.168802187069107,
+      "learning_rate": 3.768239475450269e-05,
+      "loss": 0.915,
+      "step": 1412
+    },
+    {
+      "epoch": 2.0829187396351574,
+      "grad_norm": 2.2209268556218733,
+      "learning_rate": 3.7665333469207766e-05,
+      "loss": 0.8558,
+      "step": 1413
+    },
+    {
+      "epoch": 2.0843928505620046,
+      "grad_norm": 1.8506138942891583,
+      "learning_rate": 3.7648264244780804e-05,
+      "loss": 0.8275,
+      "step": 1414
+    },
+    {
+      "epoch": 2.085866961488852,
+      "grad_norm": 1.9264525197264786,
+      "learning_rate": 3.7631187091921483e-05,
+      "loss": 0.8782,
+      "step": 1415
+    },
+    {
+      "epoch": 2.087341072415699,
+      "grad_norm": 2.1009224409760647,
+      "learning_rate": 3.761410202133443e-05,
+      "loss": 0.9072,
+      "step": 1416
+    },
+    {
+      "epoch": 2.0888151833425463,
+      "grad_norm": 1.9073569413881535,
+      "learning_rate": 3.759700904372924e-05,
+      "loss": 0.8315,
+      "step": 1417
+    },
+    {
+      "epoch": 2.0902892942693936,
+      "grad_norm": 1.9217323312295775,
+      "learning_rate": 3.757990816982046e-05,
+      "loss": 0.8848,
+      "step": 1418
+    },
+    {
+      "epoch": 2.091763405196241,
+      "grad_norm": 2.090187115848783,
+      "learning_rate": 3.756279941032761e-05,
+      "loss": 0.9534,
+      "step": 1419
+    },
+    {
+      "epoch": 2.093237516123088,
+      "grad_norm": 2.096881858484732,
+      "learning_rate": 3.754568277597512e-05,
+      "loss": 0.9431,
+      "step": 1420
+    },
+    {
+      "epoch": 2.0947116270499353,
+      "grad_norm": 1.8975055432144436,
+      "learning_rate": 3.7528558277492395e-05,
+      "loss": 0.8695,
+      "step": 1421
+    },
+    {
+      "epoch": 2.096185737976783,
+      "grad_norm": 2.009660160126084,
+      "learning_rate": 3.751142592561373e-05,
+      "loss": 0.8803,
+      "step": 1422
+    },
+    {
+      "epoch": 2.09765984890363,
+      "grad_norm": 2.173873969431794,
+      "learning_rate": 3.749428573107837e-05,
+      "loss": 0.9142,
+      "step": 1423
+    },
+    {
+      "epoch": 2.0991339598304775,
+      "grad_norm": 2.20462688360011,
+      "learning_rate": 3.747713770463046e-05,
+      "loss": 0.9121,
+      "step": 1424
+    },
+    {
+      "epoch": 2.1006080707573247,
+      "grad_norm": 2.2194274083155934,
+      "learning_rate": 3.7459981857019064e-05,
+      "loss": 0.9857,
+      "step": 1425
+    },
+    {
+      "epoch": 2.102082181684172,
+      "grad_norm": 2.261097877155455,
+      "learning_rate": 3.7442818198998156e-05,
+      "loss": 0.9549,
+      "step": 1426
+    },
+    {
+      "epoch": 2.103556292611019,
+      "grad_norm": 2.0146018100063494,
+      "learning_rate": 3.7425646741326585e-05,
+      "loss": 0.937,
+      "step": 1427
+    },
+    {
+      "epoch": 2.1050304035378664,
+      "grad_norm": 2.0066025944024872,
+      "learning_rate": 3.74084674947681e-05,
+      "loss": 0.8741,
+      "step": 1428
+    },
+    {
+      "epoch": 2.1065045144647137,
+      "grad_norm": 2.0212309010393055,
+      "learning_rate": 3.739128047009134e-05,
+      "loss": 0.8549,
+      "step": 1429
+    },
+    {
+      "epoch": 2.107978625391561,
+      "grad_norm": 1.8842845559754433,
+      "learning_rate": 3.7374085678069794e-05,
+      "loss": 0.9084,
+      "step": 1430
+    },
+    {
+      "epoch": 2.109452736318408,
+      "grad_norm": 1.9773793097124184,
+      "learning_rate": 3.735688312948186e-05,
+      "loss": 0.9287,
+      "step": 1431
+    },
+    {
+      "epoch": 2.1109268472452554,
+      "grad_norm": 2.3539347801669837,
+      "learning_rate": 3.733967283511077e-05,
+      "loss": 1.05,
+      "step": 1432
+    },
+    {
+      "epoch": 2.1124009581721026,
+      "grad_norm": 2.3238116765365935,
+      "learning_rate": 3.7322454805744605e-05,
+      "loss": 1.0657,
+      "step": 1433
+    },
+    {
+      "epoch": 2.11387506909895,
+      "grad_norm": 1.961026132715875,
+      "learning_rate": 3.730522905217632e-05,
+      "loss": 0.8128,
+      "step": 1434
+    },
+    {
+      "epoch": 2.115349180025797,
+      "grad_norm": 2.0635096072987067,
+      "learning_rate": 3.728799558520369e-05,
+      "loss": 1.0626,
+      "step": 1435
+    },
+    {
+      "epoch": 2.1168232909526443,
+      "grad_norm": 2.263305529094511,
+      "learning_rate": 3.7270754415629346e-05,
+      "loss": 0.9281,
+      "step": 1436
+    },
+    {
+      "epoch": 2.1182974018794916,
+      "grad_norm": 2.2399624407548515,
+      "learning_rate": 3.725350555426072e-05,
+      "loss": 0.9108,
+      "step": 1437
+    },
+    {
+      "epoch": 2.119771512806339,
+      "grad_norm": 2.0148110335653224,
+      "learning_rate": 3.7236249011910085e-05,
+      "loss": 0.7875,
+      "step": 1438
+    },
+    {
+      "epoch": 2.121245623733186,
+      "grad_norm": 1.9612332307233578,
+      "learning_rate": 3.7218984799394534e-05,
+      "loss": 0.9921,
+      "step": 1439
+    },
+    {
+      "epoch": 2.1227197346600333,
+      "grad_norm": 2.170199179935063,
+      "learning_rate": 3.7201712927535954e-05,
+      "loss": 0.9712,
+      "step": 1440
+    },
+    {
+      "epoch": 2.1241938455868805,
+      "grad_norm": 1.964319056686202,
+      "learning_rate": 3.7184433407161026e-05,
+      "loss": 0.9123,
+      "step": 1441
+    },
+    {
+      "epoch": 2.1256679565137278,
+      "grad_norm": 2.232532728751973,
+      "learning_rate": 3.716714624910126e-05,
+      "loss": 0.8909,
+      "step": 1442
+    },
+    {
+      "epoch": 2.127142067440575,
+      "grad_norm": 2.162264741658338,
+      "learning_rate": 3.714985146419291e-05,
+      "loss": 0.8618,
+      "step": 1443
+    },
+    {
+      "epoch": 2.1286161783674222,
+      "grad_norm": 2.112909816740118,
+      "learning_rate": 3.713254906327703e-05,
+      "loss": 0.8012,
+      "step": 1444
+    },
+    {
+      "epoch": 2.1300902892942695,
+      "grad_norm": 2.01443209575205,
+      "learning_rate": 3.711523905719946e-05,
+      "loss": 0.9044,
+      "step": 1445
+    },
+    {
+      "epoch": 2.1315644002211167,
+      "grad_norm": 1.9147663924156424,
+      "learning_rate": 3.70979214568108e-05,
+      "loss": 0.8598,
+      "step": 1446
+    },
+    {
+      "epoch": 2.133038511147964,
+      "grad_norm": 2.4083290531248007,
+      "learning_rate": 3.70805962729664e-05,
+      "loss": 0.9821,
+      "step": 1447
+    },
+    {
+      "epoch": 2.134512622074811,
+      "grad_norm": 2.0323565476810055,
+      "learning_rate": 3.706326351652636e-05,
+      "loss": 0.9298,
+      "step": 1448
+    },
+    {
+      "epoch": 2.1359867330016584,
+      "grad_norm": 2.289939634202132,
+      "learning_rate": 3.704592319835557e-05,
+      "loss": 0.7555,
+      "step": 1449
+    },
+    {
+      "epoch": 2.1374608439285057,
+      "grad_norm": 2.0825533614815175,
+      "learning_rate": 3.702857532932359e-05,
+      "loss": 0.9834,
+      "step": 1450
+    },
+    {
+      "epoch": 2.138934954855353,
+      "grad_norm": 2.1134694239523424,
+      "learning_rate": 3.7011219920304774e-05,
+      "loss": 0.9361,
+      "step": 1451
+    },
+    {
+      "epoch": 2.1404090657822,
+      "grad_norm": 2.081678750721065,
+      "learning_rate": 3.699385698217816e-05,
+      "loss": 1.0017,
+      "step": 1452
+    },
+    {
+      "epoch": 2.1418831767090474,
+      "grad_norm": 2.1914111458626793,
+      "learning_rate": 3.6976486525827546e-05,
+      "loss": 0.8345,
+      "step": 1453
+    },
+    {
+      "epoch": 2.1433572876358946,
+      "grad_norm": 1.9090090936411204,
+      "learning_rate": 3.695910856214141e-05,
+      "loss": 0.9289,
+      "step": 1454
+    },
+    {
+      "epoch": 2.144831398562742,
+      "grad_norm": 2.1081727169007447,
+      "learning_rate": 3.694172310201295e-05,
+      "loss": 1.0178,
+      "step": 1455
+    },
+    {
+      "epoch": 2.146305509489589,
+      "grad_norm": 1.8817820914957895,
+      "learning_rate": 3.692433015634005e-05,
+      "loss": 0.8786,
+      "step": 1456
+    },
+    {
+      "epoch": 2.1477796204164363,
+      "grad_norm": 2.0570426178533685,
+      "learning_rate": 3.690692973602532e-05,
+      "loss": 0.9622,
+      "step": 1457
+    },
+    {
+      "epoch": 2.1492537313432836,
+      "grad_norm": 2.24391644638823,
+      "learning_rate": 3.6889521851976005e-05,
+      "loss": 1.0092,
+      "step": 1458
+    },
+    {
+      "epoch": 2.150727842270131,
+      "grad_norm": 2.246333950301708,
+      "learning_rate": 3.6872106515104065e-05,
+      "loss": 0.846,
+      "step": 1459
+    },
+    {
+      "epoch": 2.152201953196978,
+      "grad_norm": 2.0396881897429697,
+      "learning_rate": 3.6854683736326125e-05,
+      "loss": 0.9784,
+      "step": 1460
+    },
+    {
+      "epoch": 2.1536760641238253,
+      "grad_norm": 2.2976278511479196,
+      "learning_rate": 3.683725352656348e-05,
+      "loss": 0.9658,
+      "step": 1461
+    },
+    {
+      "epoch": 2.1551501750506725,
+      "grad_norm": 1.977616522884941,
+      "learning_rate": 3.681981589674206e-05,
+      "loss": 0.8477,
+      "step": 1462
+    },
+    {
+      "epoch": 2.1566242859775198,
+      "grad_norm": 1.9743275427118583,
+      "learning_rate": 3.6802370857792464e-05,
+      "loss": 0.8927,
+      "step": 1463
+    },
+    {
+      "epoch": 2.158098396904367,
+      "grad_norm": 2.0488298705232673,
+      "learning_rate": 3.678491842064995e-05,
+      "loss": 0.9523,
+      "step": 1464
+    },
+    {
+      "epoch": 2.1595725078312142,
+      "grad_norm": 2.046360912264953,
+      "learning_rate": 3.6767458596254364e-05,
+      "loss": 0.8787,
+      "step": 1465
+    },
+    {
+      "epoch": 2.1610466187580615,
+      "grad_norm": 2.1274202812849783,
+      "learning_rate": 3.674999139555024e-05,
+      "loss": 0.8368,
+      "step": 1466
+    },
+    {
+      "epoch": 2.1625207296849087,
+      "grad_norm": 2.2073350005301484,
+      "learning_rate": 3.67325168294867e-05,
+      "loss": 0.8863,
+      "step": 1467
+    },
+    {
+      "epoch": 2.163994840611756,
+      "grad_norm": 1.8879472669979658,
+      "learning_rate": 3.67150349090175e-05,
+      "loss": 0.8869,
+      "step": 1468
+    },
+    {
+      "epoch": 2.165468951538603,
+      "grad_norm": 2.039205261143766,
+      "learning_rate": 3.669754564510099e-05,
+      "loss": 0.9612,
+      "step": 1469
+    },
+    {
+      "epoch": 2.1669430624654504,
+      "grad_norm": 2.2291650385865536,
+      "learning_rate": 3.668004904870014e-05,
+      "loss": 0.8797,
+      "step": 1470
+    },
+    {
+      "epoch": 2.1684171733922977,
+      "grad_norm": 2.11697510502034,
+      "learning_rate": 3.666254513078251e-05,
+      "loss": 0.8437,
+      "step": 1471
+    },
+    {
+      "epoch": 2.169891284319145,
+      "grad_norm": 2.0246922200214605,
+      "learning_rate": 3.664503390232024e-05,
+      "loss": 0.8228,
+      "step": 1472
+    },
+    {
+      "epoch": 2.171365395245992,
+      "grad_norm": 2.1023154345774095,
+      "learning_rate": 3.6627515374290065e-05,
+      "loss": 0.992,
+      "step": 1473
+    },
+    {
+      "epoch": 2.1728395061728394,
+      "grad_norm": 1.8848701092314866,
+      "learning_rate": 3.66099895576733e-05,
+      "loss": 0.7264,
+      "step": 1474
+    },
+    {
+      "epoch": 2.1743136170996866,
+      "grad_norm": 1.9998146209634018,
+      "learning_rate": 3.6592456463455804e-05,
+      "loss": 0.9129,
+      "step": 1475
+    },
+    {
+      "epoch": 2.175787728026534,
+      "grad_norm": 2.2937966822725357,
+      "learning_rate": 3.657491610262802e-05,
+      "loss": 0.9663,
+      "step": 1476
+    },
+    {
+      "epoch": 2.177261838953381,
+      "grad_norm": 2.331690573289138,
+      "learning_rate": 3.655736848618495e-05,
+      "loss": 0.9636,
+      "step": 1477
+    },
+    {
+      "epoch": 2.1787359498802283,
+      "grad_norm": 2.111940646182858,
+      "learning_rate": 3.653981362512612e-05,
+      "loss": 0.8608,
+      "step": 1478
+    },
+    {
+      "epoch": 2.1802100608070756,
+      "grad_norm": 2.032584236341984,
+      "learning_rate": 3.652225153045562e-05,
+      "loss": 0.8771,
+      "step": 1479
+    },
+    {
+      "epoch": 2.181684171733923,
+      "grad_norm": 1.8614791376551973,
+      "learning_rate": 3.650468221318206e-05,
+      "loss": 0.7011,
+      "step": 1480
+    },
+    {
+      "epoch": 2.18315828266077,
+      "grad_norm": 2.0351159499968645,
+      "learning_rate": 3.648710568431859e-05,
+      "loss": 0.8497,
+      "step": 1481
+    },
+    {
+      "epoch": 2.1846323935876173,
+      "grad_norm": 1.992118924026525,
+      "learning_rate": 3.6469521954882865e-05,
+      "loss": 0.8525,
+      "step": 1482
+    },
+    {
+      "epoch": 2.1861065045144645,
+      "grad_norm": 2.098903406022416,
+      "learning_rate": 3.645193103589707e-05,
+      "loss": 1.0211,
+      "step": 1483
+    },
+    {
+      "epoch": 2.1875806154413118,
+      "grad_norm": 2.184934146683098,
+      "learning_rate": 3.6434332938387875e-05,
+      "loss": 0.8939,
+      "step": 1484
+    },
+    {
+      "epoch": 2.189054726368159,
+      "grad_norm": 2.1576725786657422,
+      "learning_rate": 3.6416727673386484e-05,
+      "loss": 0.8568,
+      "step": 1485
+    },
+    {
+      "epoch": 2.1905288372950062,
+      "grad_norm": 2.1528091619794893,
+      "learning_rate": 3.639911525192857e-05,
+      "loss": 1.0871,
+      "step": 1486
+    },
+    {
+      "epoch": 2.1920029482218535,
+      "grad_norm": 2.1805461983233605,
+      "learning_rate": 3.638149568505428e-05,
+      "loss": 0.9945,
+      "step": 1487
+    },
+    {
+      "epoch": 2.1934770591487007,
+      "grad_norm": 2.1263172330548277,
+      "learning_rate": 3.636386898380827e-05,
+      "loss": 1.0502,
+      "step": 1488
+    },
+    {
+      "epoch": 2.194951170075548,
+      "grad_norm": 2.2224662712979213,
+      "learning_rate": 3.634623515923965e-05,
+      "loss": 0.9194,
+      "step": 1489
+    },
+    {
+      "epoch": 2.196425281002395,
+      "grad_norm": 2.141030154903602,
+      "learning_rate": 3.632859422240199e-05,
+      "loss": 0.9408,
+      "step": 1490
+    },
+    {
+      "epoch": 2.1978993919292424,
+      "grad_norm": 1.97506916794195,
+      "learning_rate": 3.631094618435334e-05,
+      "loss": 0.9132,
+      "step": 1491
+    },
+    {
+      "epoch": 2.19937350285609,
+      "grad_norm": 2.024728723887344,
+      "learning_rate": 3.629329105615617e-05,
+      "loss": 0.7939,
+      "step": 1492
+    },
+    {
+      "epoch": 2.2008476137829374,
+      "grad_norm": 2.006878706204851,
+      "learning_rate": 3.6275628848877445e-05,
+      "loss": 0.8811,
+      "step": 1493
+    },
+    {
+      "epoch": 2.2023217247097846,
+      "grad_norm": 2.128366832464544,
+      "learning_rate": 3.6257959573588505e-05,
+      "loss": 0.9491,
+      "step": 1494
+    },
+    {
+      "epoch": 2.203795835636632,
+      "grad_norm": 2.001973727980873,
+      "learning_rate": 3.624028324136517e-05,
+      "loss": 0.8953,
+      "step": 1495
+    },
+    {
+      "epoch": 2.205269946563479,
+      "grad_norm": 1.9872733148821282,
+      "learning_rate": 3.622259986328765e-05,
+      "loss": 0.7757,
+      "step": 1496
+    },
+    {
+      "epoch": 2.2067440574903263,
+      "grad_norm": 2.011522361280237,
+      "learning_rate": 3.620490945044059e-05,
+      "loss": 0.8904,
+      "step": 1497
+    },
+    {
+      "epoch": 2.2082181684171736,
+      "grad_norm": 2.0440218310151446,
+      "learning_rate": 3.618721201391304e-05,
+      "loss": 1.0982,
+      "step": 1498
+    },
+    {
+      "epoch": 2.209692279344021,
+      "grad_norm": 2.3159360360123054,
+      "learning_rate": 3.616950756479846e-05,
+      "loss": 0.9941,
+      "step": 1499
+    },
+    {
+      "epoch": 2.211166390270868,
+      "grad_norm": 1.991417290891538,
+      "learning_rate": 3.615179611419469e-05,
+      "loss": 0.8157,
+      "step": 1500
+    },
+    {
+      "epoch": 2.211166390270868,
+      "eval_bleu": 0.04873016621100759,
+      "eval_bleu_1gram": 0.34546108364969297,
+      "eval_bleu_2gram": 0.12150810336143084,
+      "eval_bleu_3gram": 0.04877699663074917,
+      "eval_bleu_4gram": 0.02300488317527046,
+      "eval_rag_val_loss": 1.6415073826748838,
+      "eval_rouge1": 0.33328230586389573,
+      "eval_rouge2": 0.11506413960285228,
+      "eval_rougeL": 0.313148189190874,
+      "step": 1500
+    },
+    {
+      "epoch": 2.2126405011977153,
+      "grad_norm": 2.105556280974188,
+      "learning_rate": 3.613407767320398e-05,
+      "loss": 0.7993,
+      "step": 1501
+    },
+    {
+      "epoch": 2.2141146121245625,
+      "grad_norm": 2.1702186251084985,
+      "learning_rate": 3.6116352252932936e-05,
+      "loss": 0.9826,
+      "step": 1502
+    },
+    {
+      "epoch": 2.2155887230514097,
+      "grad_norm": 2.110317104610807,
+      "learning_rate": 3.609861986449256e-05,
+      "loss": 0.7707,
+      "step": 1503
+    },
+    {
+      "epoch": 2.217062833978257,
+      "grad_norm": 2.236017543764845,
+      "learning_rate": 3.6080880518998216e-05,
+      "loss": 0.8911,
+      "step": 1504
+    },
+    {
+      "epoch": 2.2185369449051042,
+      "grad_norm": 2.155143675119247,
+      "learning_rate": 3.606313422756962e-05,
+      "loss": 1.0983,
+      "step": 1505
+    },
+    {
+      "epoch": 2.2200110558319515,
+      "grad_norm": 2.051700640267227,
+      "learning_rate": 3.604538100133086e-05,
+      "loss": 1.0635,
+      "step": 1506
+    },
+    {
+      "epoch": 2.2214851667587987,
+      "grad_norm": 1.9430346040260391,
+      "learning_rate": 3.602762085141035e-05,
+      "loss": 0.8652,
+      "step": 1507
+    },
+    {
+      "epoch": 2.222959277685646,
+      "grad_norm": 2.1888095750440337,
+      "learning_rate": 3.600985378894086e-05,
+      "loss": 0.9552,
+      "step": 1508
+    },
+    {
+      "epoch": 2.224433388612493,
+      "grad_norm": 2.1659819425779454,
+      "learning_rate": 3.599207982505949e-05,
+      "loss": 0.8315,
+      "step": 1509
+    },
+    {
+      "epoch": 2.2259074995393404,
+      "grad_norm": 2.3219815297800914,
+      "learning_rate": 3.597429897090765e-05,
+      "loss": 1.0268,
+      "step": 1510
+    },
+    {
+      "epoch": 2.2273816104661877,
+      "grad_norm": 2.4124955211854777,
+      "learning_rate": 3.5956511237631106e-05,
+      "loss": 0.8832,
+      "step": 1511
+    },
+    {
+      "epoch": 2.228855721393035,
+      "grad_norm": 2.1960766097152726,
+      "learning_rate": 3.59387166363799e-05,
+      "loss": 0.9568,
+      "step": 1512
+    },
+    {
+      "epoch": 2.230329832319882,
+      "grad_norm": 1.9597512718923582,
+      "learning_rate": 3.592091517830838e-05,
+      "loss": 0.8691,
+      "step": 1513
+    },
+    {
+      "epoch": 2.2318039432467294,
+      "grad_norm": 2.086257281419265,
+      "learning_rate": 3.5903106874575235e-05,
+      "loss": 0.9168,
+      "step": 1514
+    },
+    {
+      "epoch": 2.2332780541735766,
+      "grad_norm": 2.133130842771444,
+      "learning_rate": 3.5885291736343375e-05,
+      "loss": 0.9482,
+      "step": 1515
+    },
+    {
+      "epoch": 2.234752165100424,
+      "grad_norm": 2.0770376702291715,
+      "learning_rate": 3.586746977478006e-05,
+      "loss": 0.9055,
+      "step": 1516
+    },
+    {
+      "epoch": 2.236226276027271,
+      "grad_norm": 2.050196856839468,
+      "learning_rate": 3.58496410010568e-05,
+      "loss": 0.9004,
+      "step": 1517
+    },
+    {
+      "epoch": 2.2377003869541183,
+      "grad_norm": 2.2821588665345036,
+      "learning_rate": 3.583180542634937e-05,
+      "loss": 0.9628,
+      "step": 1518
+    },
+    {
+      "epoch": 2.2391744978809656,
+      "grad_norm": 2.1189697196708783,
+      "learning_rate": 3.5813963061837815e-05,
+      "loss": 1.1239,
+      "step": 1519
+    },
+    {
+      "epoch": 2.240648608807813,
+      "grad_norm": 2.0305813202250422,
+      "learning_rate": 3.5796113918706426e-05,
+      "loss": 1.1239,
+      "step": 1520
+    },
+    {
+      "epoch": 2.24212271973466,
+      "grad_norm": 2.0927562348000968,
+      "learning_rate": 3.577825800814376e-05,
+      "loss": 0.9268,
+      "step": 1521
+    },
+    {
+      "epoch": 2.2435968306615073,
+      "grad_norm": 2.0088263300960847,
+      "learning_rate": 3.576039534134262e-05,
+      "loss": 0.9577,
+      "step": 1522
+    },
+    {
+      "epoch": 2.2450709415883545,
+      "grad_norm": 2.0326487493570893,
+      "learning_rate": 3.57425259295e-05,
+      "loss": 0.8378,
+      "step": 1523
+    },
+    {
+      "epoch": 2.2465450525152018,
+      "grad_norm": 2.04329397264233,
+      "learning_rate": 3.5724649783817185e-05,
+      "loss": 0.9176,
+      "step": 1524
+    },
+    {
+      "epoch": 2.248019163442049,
+      "grad_norm": 2.0977054639900152,
+      "learning_rate": 3.5706766915499646e-05,
+      "loss": 0.9114,
+      "step": 1525
+    },
+    {
+      "epoch": 2.2494932743688962,
+      "grad_norm": 2.089667253973781,
+      "learning_rate": 3.568887733575706e-05,
+      "loss": 0.9543,
+      "step": 1526
+    },
+    {
+      "epoch": 2.2509673852957435,
+      "grad_norm": 2.130977525940725,
+      "learning_rate": 3.567098105580333e-05,
+      "loss": 0.8649,
+      "step": 1527
+    },
+    {
+      "epoch": 2.2524414962225907,
+      "grad_norm": 2.1621832554986837,
+      "learning_rate": 3.5653078086856546e-05,
+      "loss": 0.8767,
+      "step": 1528
+    },
+    {
+      "epoch": 2.253915607149438,
+      "grad_norm": 2.0102921547817134,
+      "learning_rate": 3.563516844013901e-05,
+      "loss": 0.9834,
+      "step": 1529
+    },
+    {
+      "epoch": 2.255389718076285,
+      "grad_norm": 1.9456040497119693,
+      "learning_rate": 3.561725212687718e-05,
+      "loss": 0.9435,
+      "step": 1530
+    },
+    {
+      "epoch": 2.2568638290031324,
+      "grad_norm": 2.13741044420168,
+      "learning_rate": 3.559932915830172e-05,
+      "loss": 0.9592,
+      "step": 1531
+    },
+    {
+      "epoch": 2.2583379399299797,
+      "grad_norm": 2.0989162418761205,
+      "learning_rate": 3.558139954564746e-05,
+      "loss": 0.8713,
+      "step": 1532
+    },
+    {
+      "epoch": 2.259812050856827,
+      "grad_norm": 1.9467204563774225,
+      "learning_rate": 3.556346330015338e-05,
+      "loss": 0.7914,
+      "step": 1533
+    },
+    {
+      "epoch": 2.261286161783674,
+      "grad_norm": 2.111455385958536,
+      "learning_rate": 3.554552043306264e-05,
+      "loss": 0.9559,
+      "step": 1534
+    },
+    {
+      "epoch": 2.2627602727105214,
+      "grad_norm": 2.314534400768346,
+      "learning_rate": 3.552757095562253e-05,
+      "loss": 1.0114,
+      "step": 1535
+    },
+    {
+      "epoch": 2.2642343836373686,
+      "grad_norm": 2.132745091976501,
+      "learning_rate": 3.55096148790845e-05,
+      "loss": 0.9328,
+      "step": 1536
+    },
+    {
+      "epoch": 2.265708494564216,
+      "grad_norm": 2.2102806814096696,
+      "learning_rate": 3.5491652214704115e-05,
+      "loss": 0.9129,
+      "step": 1537
+    },
+    {
+      "epoch": 2.267182605491063,
+      "grad_norm": 2.1735395461039544,
+      "learning_rate": 3.547368297374109e-05,
+      "loss": 0.9542,
+      "step": 1538
+    },
+    {
+      "epoch": 2.2686567164179103,
+      "grad_norm": 2.142869649578109,
+      "learning_rate": 3.545570716745927e-05,
+      "loss": 0.8873,
+      "step": 1539
+    },
+    {
+      "epoch": 2.2701308273447576,
+      "grad_norm": 2.31773402739152,
+      "learning_rate": 3.543772480712658e-05,
+      "loss": 1.0277,
+      "step": 1540
+    },
+    {
+      "epoch": 2.271604938271605,
+      "grad_norm": 2.048202200846928,
+      "learning_rate": 3.5419735904015095e-05,
+      "loss": 0.9348,
+      "step": 1541
+    },
+    {
+      "epoch": 2.273079049198452,
+      "grad_norm": 2.072265625,
+      "learning_rate": 3.540174046940096e-05,
+      "loss": 0.9552,
+      "step": 1542
+    },
+    {
+      "epoch": 2.2745531601252993,
+      "grad_norm": 2.1244810256020905,
+      "learning_rate": 3.538373851456442e-05,
+      "loss": 1.0018,
+      "step": 1543
+    },
+    {
+      "epoch": 2.2760272710521465,
+      "grad_norm": 2.412079820953815,
+      "learning_rate": 3.536573005078981e-05,
+      "loss": 0.972,
+      "step": 1544
+    },
+    {
+      "epoch": 2.2775013819789938,
+      "grad_norm": 2.118810953259364,
+      "learning_rate": 3.5347715089365576e-05,
+      "loss": 0.9555,
+      "step": 1545
+    },
+    {
+      "epoch": 2.278975492905841,
+      "grad_norm": 2.2370507284729766,
+      "learning_rate": 3.532969364158417e-05,
+      "loss": 0.915,
+      "step": 1546
+    },
+    {
+      "epoch": 2.2804496038326882,
+      "grad_norm": 2.1738247250155913,
+      "learning_rate": 3.5311665718742184e-05,
+      "loss": 1.0054,
+      "step": 1547
+    },
+    {
+      "epoch": 2.2819237147595355,
+      "grad_norm": 2.0754793004212426,
+      "learning_rate": 3.529363133214021e-05,
+      "loss": 0.8591,
+      "step": 1548
+    },
+    {
+      "epoch": 2.2833978256863827,
+      "grad_norm": 2.163074474433509,
+      "learning_rate": 3.527559049308291e-05,
+      "loss": 0.9694,
+      "step": 1549
+    },
+    {
+      "epoch": 2.28487193661323,
+      "grad_norm": 2.188200048419773,
+      "learning_rate": 3.525754321287902e-05,
+      "loss": 1.1303,
+      "step": 1550
+    },
+    {
+      "epoch": 2.286346047540077,
+      "grad_norm": 2.187646915406885,
+      "learning_rate": 3.523948950284127e-05,
+      "loss": 0.9452,
+      "step": 1551
+    },
+    {
+      "epoch": 2.287820158466925,
+      "grad_norm": 2.2667391438210025,
+      "learning_rate": 3.522142937428645e-05,
+      "loss": 0.8507,
+      "step": 1552
+    },
+    {
+      "epoch": 2.289294269393772,
+      "grad_norm": 2.059186527964843,
+      "learning_rate": 3.5203362838535355e-05,
+      "loss": 0.9941,
+      "step": 1553
+    },
+    {
+      "epoch": 2.2907683803206194,
+      "grad_norm": 2.1223300541394625,
+      "learning_rate": 3.518528990691281e-05,
+      "loss": 0.9625,
+      "step": 1554
+    },
+    {
+      "epoch": 2.2922424912474666,
+      "grad_norm": 2.0428141139716907,
+      "learning_rate": 3.516721059074764e-05,
+      "loss": 0.8442,
+      "step": 1555
+    },
+    {
+      "epoch": 2.293716602174314,
+      "grad_norm": 2.2479874829252657,
+      "learning_rate": 3.5149124901372677e-05,
+      "loss": 1.0568,
+      "step": 1556
+    },
+    {
+      "epoch": 2.295190713101161,
+      "grad_norm": 2.2184364204710407,
+      "learning_rate": 3.513103285012475e-05,
+      "loss": 1.0052,
+      "step": 1557
+    },
+    {
+      "epoch": 2.2966648240280083,
+      "grad_norm": 1.987701391417144,
+      "learning_rate": 3.511293444834466e-05,
+      "loss": 0.9989,
+      "step": 1558
+    },
+    {
+      "epoch": 2.2981389349548556,
+      "grad_norm": 2.0599799510757935,
+      "learning_rate": 3.509482970737722e-05,
+      "loss": 1.0301,
+      "step": 1559
+    },
+    {
+      "epoch": 2.299613045881703,
+      "grad_norm": 2.140687704908184,
+      "learning_rate": 3.5076718638571185e-05,
+      "loss": 0.8967,
+      "step": 1560
+    },
+    {
+      "epoch": 2.30108715680855,
+      "grad_norm": 2.3877592175752818,
+      "learning_rate": 3.505860125327928e-05,
+      "loss": 0.9361,
+      "step": 1561
+    },
+    {
+      "epoch": 2.3025612677353973,
+      "grad_norm": 2.125644137423384,
+      "learning_rate": 3.504047756285822e-05,
+      "loss": 0.8999,
+      "step": 1562
+    },
+    {
+      "epoch": 2.3040353786622445,
+      "grad_norm": 1.9433866105974673,
+      "learning_rate": 3.5022347578668644e-05,
+      "loss": 0.8091,
+      "step": 1563
+    },
+    {
+      "epoch": 2.3055094895890917,
+      "grad_norm": 2.0982249250982865,
+      "learning_rate": 3.5004211312075143e-05,
+      "loss": 0.9879,
+      "step": 1564
+    },
+    {
+      "epoch": 2.306983600515939,
+      "grad_norm": 2.43759438747665,
+      "learning_rate": 3.498606877444625e-05,
+      "loss": 0.9294,
+      "step": 1565
+    },
+    {
+      "epoch": 2.308457711442786,
+      "grad_norm": 2.1661650981873426,
+      "learning_rate": 3.4967919977154406e-05,
+      "loss": 0.9955,
+      "step": 1566
+    },
+    {
+      "epoch": 2.3099318223696335,
+      "grad_norm": 2.0028949290011826,
+      "learning_rate": 3.4949764931576014e-05,
+      "loss": 0.8,
+      "step": 1567
+    },
+    {
+      "epoch": 2.3114059332964807,
+      "grad_norm": 2.009714850700947,
+      "learning_rate": 3.4931603649091374e-05,
+      "loss": 0.8421,
+      "step": 1568
+    },
+    {
+      "epoch": 2.312880044223328,
+      "grad_norm": 2.417448465454824,
+      "learning_rate": 3.4913436141084676e-05,
+      "loss": 1.1608,
+      "step": 1569
+    },
+    {
+      "epoch": 2.314354155150175,
+      "grad_norm": 2.1768051286075494,
+      "learning_rate": 3.489526241894406e-05,
+      "loss": 0.8538,
+      "step": 1570
+    },
+    {
+      "epoch": 2.3158282660770224,
+      "grad_norm": 2.07055547476035,
+      "learning_rate": 3.487708249406153e-05,
+      "loss": 0.8612,
+      "step": 1571
+    },
+    {
+      "epoch": 2.3173023770038697,
+      "grad_norm": 2.353764311128565,
+      "learning_rate": 3.4858896377832966e-05,
+      "loss": 1.0192,
+      "step": 1572
+    },
+    {
+      "epoch": 2.318776487930717,
+      "grad_norm": 2.2784707894075784,
+      "learning_rate": 3.4840704081658155e-05,
+      "loss": 0.9219,
+      "step": 1573
+    },
+    {
+      "epoch": 2.320250598857564,
+      "grad_norm": 2.2050579059981468,
+      "learning_rate": 3.482250561694075e-05,
+      "loss": 0.9986,
+      "step": 1574
+    },
+    {
+      "epoch": 2.3217247097844114,
+      "grad_norm": 2.136927620618377,
+      "learning_rate": 3.4804300995088264e-05,
+      "loss": 0.9289,
+      "step": 1575
+    },
+    {
+      "epoch": 2.3231988207112586,
+      "grad_norm": 2.180358674002088,
+      "learning_rate": 3.478609022751207e-05,
+      "loss": 0.8347,
+      "step": 1576
+    },
+    {
+      "epoch": 2.324672931638106,
+      "grad_norm": 2.2248727912113884,
+      "learning_rate": 3.4767873325627406e-05,
+      "loss": 0.9333,
+      "step": 1577
+    },
+    {
+      "epoch": 2.326147042564953,
+      "grad_norm": 2.162521309279041,
+      "learning_rate": 3.4749650300853343e-05,
+      "loss": 0.949,
+      "step": 1578
+    },
+    {
+      "epoch": 2.3276211534918003,
+      "grad_norm": 2.082663148531871,
+      "learning_rate": 3.473142116461279e-05,
+      "loss": 0.9482,
+      "step": 1579
+    },
+    {
+      "epoch": 2.3290952644186476,
+      "grad_norm": 2.085001989313361,
+      "learning_rate": 3.47131859283325e-05,
+      "loss": 0.8233,
+      "step": 1580
+    },
+    {
+      "epoch": 2.330569375345495,
+      "grad_norm": 2.176165178761462,
+      "learning_rate": 3.469494460344304e-05,
+      "loss": 0.871,
+      "step": 1581
+    },
+    {
+      "epoch": 2.332043486272342,
+      "grad_norm": 2.0867394895104945,
+      "learning_rate": 3.467669720137879e-05,
+      "loss": 0.9044,
+      "step": 1582
+    },
+    {
+      "epoch": 2.3335175971991893,
+      "grad_norm": 2.152126848261575,
+      "learning_rate": 3.465844373357794e-05,
+      "loss": 0.9604,
+      "step": 1583
+    },
+    {
+      "epoch": 2.3349917081260365,
+      "grad_norm": 2.1952117136562106,
+      "learning_rate": 3.464018421148249e-05,
+      "loss": 0.9005,
+      "step": 1584
+    },
+    {
+      "epoch": 2.3364658190528838,
+      "grad_norm": 2.1958414438328324,
+      "learning_rate": 3.462191864653821e-05,
+      "loss": 0.9823,
+      "step": 1585
+    },
+    {
+      "epoch": 2.337939929979731,
+      "grad_norm": 2.0407326826159697,
+      "learning_rate": 3.460364705019472e-05,
+      "loss": 0.8598,
+      "step": 1586
+    },
+    {
+      "epoch": 2.3394140409065782,
+      "grad_norm": 2.0465947316023176,
+      "learning_rate": 3.458536943390536e-05,
+      "loss": 0.9172,
+      "step": 1587
+    },
+    {
+      "epoch": 2.3408881518334255,
+      "grad_norm": 2.39592658704524,
+      "learning_rate": 3.456708580912725e-05,
+      "loss": 0.9101,
+      "step": 1588
+    },
+    {
+      "epoch": 2.3423622627602727,
+      "grad_norm": 2.0897879619817177,
+      "learning_rate": 3.4548796187321295e-05,
+      "loss": 0.8849,
+      "step": 1589
+    },
+    {
+      "epoch": 2.34383637368712,
+      "grad_norm": 2.07649890121993,
+      "learning_rate": 3.453050057995217e-05,
+      "loss": 0.9358,
+      "step": 1590
+    },
+    {
+      "epoch": 2.345310484613967,
+      "grad_norm": 2.0713856344987676,
+      "learning_rate": 3.451219899848827e-05,
+      "loss": 0.9659,
+      "step": 1591
+    },
+    {
+      "epoch": 2.3467845955408144,
+      "grad_norm": 2.2172347791602873,
+      "learning_rate": 3.449389145440175e-05,
+      "loss": 0.9438,
+      "step": 1592
+    },
+    {
+      "epoch": 2.3482587064676617,
+      "grad_norm": 2.1269267827584213,
+      "learning_rate": 3.4475577959168505e-05,
+      "loss": 0.9422,
+      "step": 1593
+    },
+    {
+      "epoch": 2.349732817394509,
+      "grad_norm": 2.182007460800828,
+      "learning_rate": 3.445725852426817e-05,
+      "loss": 0.9615,
+      "step": 1594
+    },
+    {
+      "epoch": 2.351206928321356,
+      "grad_norm": 2.151156726435392,
+      "learning_rate": 3.443893316118407e-05,
+      "loss": 1.0295,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3526810392482034,
+      "grad_norm": 2.3361724429169253,
+      "learning_rate": 3.4420601881403284e-05,
+      "loss": 1.0373,
+      "step": 1596
+    },
+    {
+      "epoch": 2.3541551501750506,
+      "grad_norm": 2.106883067633596,
+      "learning_rate": 3.440226469641658e-05,
+      "loss": 0.8609,
+      "step": 1597
+    },
+    {
+      "epoch": 2.355629261101898,
+      "grad_norm": 1.9681753500648527,
+      "learning_rate": 3.4383921617718427e-05,
+      "loss": 0.7817,
+      "step": 1598
+    },
+    {
+      "epoch": 2.357103372028745,
+      "grad_norm": 2.025628041962785,
+      "learning_rate": 3.4365572656807e-05,
+      "loss": 0.9318,
+      "step": 1599
+    },
+    {
+      "epoch": 2.3585774829555923,
+      "grad_norm": 1.9314713977707498,
+      "learning_rate": 3.4347217825184134e-05,
+      "loss": 0.7432,
+      "step": 1600
+    },
+    {
+      "epoch": 2.3600515938824396,
+      "grad_norm": 2.006976120408059,
+      "learning_rate": 3.432885713435539e-05,
+      "loss": 0.7835,
+      "step": 1601
+    },
+    {
+      "epoch": 2.361525704809287,
+      "grad_norm": 1.988069354384554,
+      "learning_rate": 3.431049059582996e-05,
+      "loss": 1.1013,
+      "step": 1602
+    },
+    {
+      "epoch": 2.362999815736134,
+      "grad_norm": 1.8315693287184354,
+      "learning_rate": 3.4292118221120715e-05,
+      "loss": 0.8786,
+      "step": 1603
+    },
+    {
+      "epoch": 2.3644739266629813,
+      "grad_norm": 1.9643463051108345,
+      "learning_rate": 3.42737400217442e-05,
+      "loss": 0.966,
+      "step": 1604
+    },
+    {
+      "epoch": 2.3659480375898285,
+      "grad_norm": 2.0706976766178204,
+      "learning_rate": 3.425535600922059e-05,
+      "loss": 0.9188,
+      "step": 1605
+    },
+    {
+      "epoch": 2.3674221485166758,
+      "grad_norm": 2.2718845582364806,
+      "learning_rate": 3.423696619507369e-05,
+      "loss": 1.0676,
+      "step": 1606
+    },
+    {
+      "epoch": 2.368896259443523,
+      "grad_norm": 2.1186866098109087,
+      "learning_rate": 3.4218570590831e-05,
+      "loss": 0.8887,
+      "step": 1607
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 2.15284748032915,
+      "learning_rate": 3.4200169208023594e-05,
+      "loss": 0.9728,
+      "step": 1608
+    },
+    {
+      "epoch": 2.3718444812972175,
+      "grad_norm": 2.0876352780304197,
+      "learning_rate": 3.418176205818618e-05,
+      "loss": 0.9943,
+      "step": 1609
+    },
+    {
+      "epoch": 2.3733185922240647,
+      "grad_norm": 2.09037280649979,
+      "learning_rate": 3.4163349152857096e-05,
+      "loss": 0.9014,
+      "step": 1610
+    },
+    {
+      "epoch": 2.374792703150912,
+      "grad_norm": 2.0884968918513978,
+      "learning_rate": 3.4144930503578286e-05,
+      "loss": 0.9424,
+      "step": 1611
+    },
+    {
+      "epoch": 2.376266814077759,
+      "grad_norm": 2.1697687048634884,
+      "learning_rate": 3.412650612189528e-05,
+      "loss": 0.8936,
+      "step": 1612
+    },
+    {
+      "epoch": 2.3777409250046064,
+      "grad_norm": 2.573344381717162,
+      "learning_rate": 3.4108076019357204e-05,
+      "loss": 0.9592,
+      "step": 1613
+    },
+    {
+      "epoch": 2.3792150359314537,
+      "grad_norm": 2.2670968372139333,
+      "learning_rate": 3.4089640207516786e-05,
+      "loss": 0.9989,
+      "step": 1614
+    },
+    {
+      "epoch": 2.380689146858301,
+      "grad_norm": 2.131349725711338,
+      "learning_rate": 3.4071198697930315e-05,
+      "loss": 0.9266,
+      "step": 1615
+    },
+    {
+      "epoch": 2.382163257785148,
+      "grad_norm": 2.427159898939151,
+      "learning_rate": 3.405275150215766e-05,
+      "loss": 0.9548,
+      "step": 1616
+    },
+    {
+      "epoch": 2.3836373687119954,
+      "grad_norm": 2.084556207972814,
+      "learning_rate": 3.403429863176226e-05,
+      "loss": 0.9502,
+      "step": 1617
+    },
+    {
+      "epoch": 2.3851114796388426,
+      "grad_norm": 2.096098991327774,
+      "learning_rate": 3.40158400983111e-05,
+      "loss": 0.7499,
+      "step": 1618
+    },
+    {
+      "epoch": 2.38658559056569,
+      "grad_norm": 2.0713654917240403,
+      "learning_rate": 3.399737591337471e-05,
+      "loss": 1.0064,
+      "step": 1619
+    },
+    {
+      "epoch": 2.388059701492537,
+      "grad_norm": 2.069673835562989,
+      "learning_rate": 3.397890608852718e-05,
+      "loss": 0.9333,
+      "step": 1620
+    },
+    {
+      "epoch": 2.3895338124193843,
+      "grad_norm": 1.9691526818582668,
+      "learning_rate": 3.396043063534613e-05,
+      "loss": 0.8549,
+      "step": 1621
+    },
+    {
+      "epoch": 2.3910079233462316,
+      "grad_norm": 2.3543610506629316,
+      "learning_rate": 3.39419495654127e-05,
+      "loss": 0.9269,
+      "step": 1622
+    },
+    {
+      "epoch": 2.392482034273079,
+      "grad_norm": 2.1781065755363147,
+      "learning_rate": 3.3923462890311544e-05,
+      "loss": 0.9849,
+      "step": 1623
+    },
+    {
+      "epoch": 2.393956145199926,
+      "grad_norm": 2.092803855172961,
+      "learning_rate": 3.3904970621630866e-05,
+      "loss": 0.8728,
+      "step": 1624
+    },
+    {
+      "epoch": 2.3954302561267733,
+      "grad_norm": 2.165155678629464,
+      "learning_rate": 3.388647277096234e-05,
+      "loss": 0.9673,
+      "step": 1625
+    },
+    {
+      "epoch": 2.396904367053621,
+      "grad_norm": 2.1029362496051953,
+      "learning_rate": 3.386796934990115e-05,
+      "loss": 1.0627,
+      "step": 1626
+    },
+    {
+      "epoch": 2.398378477980468,
+      "grad_norm": 1.859216763471053,
+      "learning_rate": 3.3849460370045966e-05,
+      "loss": 0.7436,
+      "step": 1627
+    },
+    {
+      "epoch": 2.3998525889073155,
+      "grad_norm": 2.0326469899383826,
+      "learning_rate": 3.3830945842998954e-05,
+      "loss": 0.9289,
+      "step": 1628
+    },
+    {
+      "epoch": 2.4013266998341627,
+      "grad_norm": 2.190326064061408,
+      "learning_rate": 3.381242578036576e-05,
+      "loss": 0.9337,
+      "step": 1629
+    },
+    {
+      "epoch": 2.40280081076101,
+      "grad_norm": 2.3612744094128177,
+      "learning_rate": 3.379390019375548e-05,
+      "loss": 0.9605,
+      "step": 1630
+    },
+    {
+      "epoch": 2.404274921687857,
+      "grad_norm": 2.2711833252512696,
+      "learning_rate": 3.377536909478069e-05,
+      "loss": 0.9584,
+      "step": 1631
+    },
+    {
+      "epoch": 2.4057490326147044,
+      "grad_norm": 1.9657601557176625,
+      "learning_rate": 3.3756832495057414e-05,
+      "loss": 0.8141,
+      "step": 1632
+    },
+    {
+      "epoch": 2.4072231435415516,
+      "grad_norm": 2.380512565908135,
+      "learning_rate": 3.373829040620513e-05,
+      "loss": 0.8898,
+      "step": 1633
+    },
+    {
+      "epoch": 2.408697254468399,
+      "grad_norm": 2.1856932672011813,
+      "learning_rate": 3.3719742839846743e-05,
+      "loss": 0.8777,
+      "step": 1634
+    },
+    {
+      "epoch": 2.410171365395246,
+      "grad_norm": 1.946063896204426,
+      "learning_rate": 3.370118980760861e-05,
+      "loss": 0.8107,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4116454763220934,
+      "grad_norm": 2.0184666910170606,
+      "learning_rate": 3.3682631321120504e-05,
+      "loss": 0.8594,
+      "step": 1636
+    },
+    {
+      "epoch": 2.4131195872489406,
+      "grad_norm": 2.3139880779490065,
+      "learning_rate": 3.366406739201562e-05,
+      "loss": 1.0516,
+      "step": 1637
+    },
+    {
+      "epoch": 2.414593698175788,
+      "grad_norm": 2.170370335422805,
+      "learning_rate": 3.364549803193057e-05,
+      "loss": 0.9245,
+      "step": 1638
+    },
+    {
+      "epoch": 2.416067809102635,
+      "grad_norm": 2.193439376880252,
+      "learning_rate": 3.362692325250534e-05,
+      "loss": 0.909,
+      "step": 1639
+    },
+    {
+      "epoch": 2.4175419200294823,
+      "grad_norm": 2.27545747454107,
+      "learning_rate": 3.360834306538336e-05,
+      "loss": 1.0398,
+      "step": 1640
+    },
+    {
+      "epoch": 2.4190160309563296,
+      "grad_norm": 2.1886718880719593,
+      "learning_rate": 3.3589757482211416e-05,
+      "loss": 1.0136,
+      "step": 1641
+    },
+    {
+      "epoch": 2.420490141883177,
+      "grad_norm": 2.114314871848405,
+      "learning_rate": 3.3571166514639684e-05,
+      "loss": 0.8983,
+      "step": 1642
+    },
+    {
+      "epoch": 2.421964252810024,
+      "grad_norm": 2.212360476819845,
+      "learning_rate": 3.3552570174321724e-05,
+      "loss": 0.9351,
+      "step": 1643
+    },
+    {
+      "epoch": 2.4234383637368713,
+      "grad_norm": 2.1295423764707793,
+      "learning_rate": 3.353396847291446e-05,
+      "loss": 0.8284,
+      "step": 1644
+    },
+    {
+      "epoch": 2.4249124746637185,
+      "grad_norm": 2.221608107535075,
+      "learning_rate": 3.3515361422078165e-05,
+      "loss": 1.0301,
+      "step": 1645
+    },
+    {
+      "epoch": 2.4263865855905657,
+      "grad_norm": 2.2171231605275685,
+      "learning_rate": 3.3496749033476485e-05,
+      "loss": 1.0551,
+      "step": 1646
+    },
+    {
+      "epoch": 2.427860696517413,
+      "grad_norm": 1.8549922291060805,
+      "learning_rate": 3.347813131877638e-05,
+      "loss": 0.7652,
+      "step": 1647
+    },
+    {
+      "epoch": 2.4293348074442602,
+      "grad_norm": 2.36213694209873,
+      "learning_rate": 3.34595082896482e-05,
+      "loss": 1.0077,
+      "step": 1648
+    },
+    {
+      "epoch": 2.4308089183711075,
+      "grad_norm": 1.9535966227459283,
+      "learning_rate": 3.344087995776558e-05,
+      "loss": 0.8663,
+      "step": 1649
+    },
+    {
+      "epoch": 2.4322830292979547,
+      "grad_norm": 2.3428508814119944,
+      "learning_rate": 3.34222463348055e-05,
+      "loss": 1.135,
+      "step": 1650
+    },
+    {
+      "epoch": 2.433757140224802,
+      "grad_norm": 2.383183084655947,
+      "learning_rate": 3.340360743244825e-05,
+      "loss": 0.7825,
+      "step": 1651
+    },
+    {
+      "epoch": 2.435231251151649,
+      "grad_norm": 2.3485923081551503,
+      "learning_rate": 3.338496326237743e-05,
+      "loss": 0.9195,
+      "step": 1652
+    },
+    {
+      "epoch": 2.4367053620784964,
+      "grad_norm": 2.155528956945027,
+      "learning_rate": 3.336631383627995e-05,
+      "loss": 1.0636,
+      "step": 1653
+    },
+    {
+      "epoch": 2.4381794730053437,
+      "grad_norm": 2.195288824313503,
+      "learning_rate": 3.334765916584599e-05,
+      "loss": 0.8705,
+      "step": 1654
+    },
+    {
+      "epoch": 2.439653583932191,
+      "grad_norm": 2.3171733338739053,
+      "learning_rate": 3.332899926276905e-05,
+      "loss": 0.9485,
+      "step": 1655
+    },
+    {
+      "epoch": 2.441127694859038,
+      "grad_norm": 2.1115926882042584,
+      "learning_rate": 3.33103341387459e-05,
+      "loss": 0.8916,
+      "step": 1656
+    },
+    {
+      "epoch": 2.4426018057858854,
+      "grad_norm": 1.9914602111473398,
+      "learning_rate": 3.3291663805476566e-05,
+      "loss": 1.026,
+      "step": 1657
+    },
+    {
+      "epoch": 2.4440759167127326,
+      "grad_norm": 2.040109416794873,
+      "learning_rate": 3.3272988274664364e-05,
+      "loss": 0.9017,
+      "step": 1658
+    },
+    {
+      "epoch": 2.44555002763958,
+      "grad_norm": 1.9436320818002248,
+      "learning_rate": 3.325430755801584e-05,
+      "loss": 0.876,
+      "step": 1659
+    },
+    {
+      "epoch": 2.447024138566427,
+      "grad_norm": 2.1072596362915634,
+      "learning_rate": 3.323562166724082e-05,
+      "loss": 0.961,
+      "step": 1660
+    },
+    {
+      "epoch": 2.4484982494932743,
+      "grad_norm": 2.115524708837307,
+      "learning_rate": 3.321693061405235e-05,
+      "loss": 1.0675,
+      "step": 1661
+    },
+    {
+      "epoch": 2.4499723604201216,
+      "grad_norm": 2.1230358695523166,
+      "learning_rate": 3.319823441016673e-05,
+      "loss": 0.9437,
+      "step": 1662
+    },
+    {
+      "epoch": 2.451446471346969,
+      "grad_norm": 2.119031152666345,
+      "learning_rate": 3.317953306730347e-05,
+      "loss": 0.9348,
+      "step": 1663
+    },
+    {
+      "epoch": 2.452920582273816,
+      "grad_norm": 2.071947251292825,
+      "learning_rate": 3.316082659718532e-05,
+      "loss": 0.9308,
+      "step": 1664
+    },
+    {
+      "epoch": 2.4543946932006633,
+      "grad_norm": 2.2747578670216,
+      "learning_rate": 3.314211501153823e-05,
+      "loss": 0.8905,
+      "step": 1665
+    },
+    {
+      "epoch": 2.4558688041275105,
+      "grad_norm": 2.1415525779566638,
+      "learning_rate": 3.312339832209137e-05,
+      "loss": 0.8951,
+      "step": 1666
+    },
+    {
+      "epoch": 2.4573429150543578,
+      "grad_norm": 2.0920363288263304,
+      "learning_rate": 3.3104676540577094e-05,
+      "loss": 0.9556,
+      "step": 1667
+    },
+    {
+      "epoch": 2.458817025981205,
+      "grad_norm": 2.2609223065396336,
+      "learning_rate": 3.308594967873095e-05,
+      "loss": 1.0335,
+      "step": 1668
+    },
+    {
+      "epoch": 2.4602911369080522,
+      "grad_norm": 2.191359249161743,
+      "learning_rate": 3.3067217748291695e-05,
+      "loss": 1.0301,
+      "step": 1669
+    },
+    {
+      "epoch": 2.4617652478348995,
+      "grad_norm": 2.0595600097167033,
+      "learning_rate": 3.304848076100122e-05,
+      "loss": 1.074,
+      "step": 1670
+    },
+    {
+      "epoch": 2.4632393587617467,
+      "grad_norm": 2.1670555597144157,
+      "learning_rate": 3.302973872860463e-05,
+      "loss": 0.8694,
+      "step": 1671
+    },
+    {
+      "epoch": 2.464713469688594,
+      "grad_norm": 2.1512601308899946,
+      "learning_rate": 3.301099166285017e-05,
+      "loss": 0.9075,
+      "step": 1672
+    },
+    {
+      "epoch": 2.466187580615441,
+      "grad_norm": 2.138781791196569,
+      "learning_rate": 3.299223957548923e-05,
+      "loss": 0.9028,
+      "step": 1673
+    },
+    {
+      "epoch": 2.4676616915422884,
+      "grad_norm": 2.040666672659351,
+      "learning_rate": 3.2973482478276364e-05,
+      "loss": 1.003,
+      "step": 1674
+    },
+    {
+      "epoch": 2.4691358024691357,
+      "grad_norm": 2.184293522222493,
+      "learning_rate": 3.2954720382969263e-05,
+      "loss": 0.9832,
+      "step": 1675
+    },
+    {
+      "epoch": 2.470609913395983,
+      "grad_norm": 2.21155272641892,
+      "learning_rate": 3.293595330132876e-05,
+      "loss": 0.951,
+      "step": 1676
+    },
+    {
+      "epoch": 2.47208402432283,
+      "grad_norm": 2.103558014820677,
+      "learning_rate": 3.291718124511879e-05,
+      "loss": 0.9952,
+      "step": 1677
+    },
+    {
+      "epoch": 2.4735581352496774,
+      "grad_norm": 2.3208753115169474,
+      "learning_rate": 3.289840422610643e-05,
+      "loss": 1.0309,
+      "step": 1678
+    },
+    {
+      "epoch": 2.4750322461765246,
+      "grad_norm": 1.9706674956322985,
+      "learning_rate": 3.287962225606185e-05,
+      "loss": 0.9542,
+      "step": 1679
+    },
+    {
+      "epoch": 2.476506357103372,
+      "grad_norm": 2.089495193077924,
+      "learning_rate": 3.286083534675835e-05,
+      "loss": 1.0185,
+      "step": 1680
+    },
+    {
+      "epoch": 2.477980468030219,
+      "grad_norm": 2.145761383726894,
+      "learning_rate": 3.284204350997229e-05,
+      "loss": 0.9694,
+      "step": 1681
+    },
+    {
+      "epoch": 2.4794545789570663,
+      "grad_norm": 2.1918397634440576,
+      "learning_rate": 3.282324675748314e-05,
+      "loss": 0.8156,
+      "step": 1682
+    },
+    {
+      "epoch": 2.4809286898839136,
+      "grad_norm": 2.0903991531612696,
+      "learning_rate": 3.280444510107346e-05,
+      "loss": 0.9545,
+      "step": 1683
+    },
+    {
+      "epoch": 2.4824028008107613,
+      "grad_norm": 2.218925254240221,
+      "learning_rate": 3.278563855252885e-05,
+      "loss": 0.9508,
+      "step": 1684
+    },
+    {
+      "epoch": 2.4838769117376085,
+      "grad_norm": 2.1670144119945287,
+      "learning_rate": 3.276682712363801e-05,
+      "loss": 1.2227,
+      "step": 1685
+    },
+    {
+      "epoch": 2.4853510226644557,
+      "grad_norm": 2.1599083425360295,
+      "learning_rate": 3.274801082619269e-05,
+      "loss": 0.9885,
+      "step": 1686
+    },
+    {
+      "epoch": 2.486825133591303,
+      "grad_norm": 2.1722997894920595,
+      "learning_rate": 3.2729189671987695e-05,
+      "loss": 0.8845,
+      "step": 1687
+    },
+    {
+      "epoch": 2.48829924451815,
+      "grad_norm": 2.1762462509447262,
+      "learning_rate": 3.271036367282085e-05,
+      "loss": 0.9866,
+      "step": 1688
+    },
+    {
+      "epoch": 2.4897733554449974,
+      "grad_norm": 2.1546321825834034,
+      "learning_rate": 3.269153284049306e-05,
+      "loss": 0.8365,
+      "step": 1689
+    },
+    {
+      "epoch": 2.4912474663718447,
+      "grad_norm": 2.1551317272949717,
+      "learning_rate": 3.267269718680822e-05,
+      "loss": 0.9565,
+      "step": 1690
+    },
+    {
+      "epoch": 2.492721577298692,
+      "grad_norm": 2.079049958627438,
+      "learning_rate": 3.265385672357327e-05,
+      "loss": 1.0143,
+      "step": 1691
+    },
+    {
+      "epoch": 2.494195688225539,
+      "grad_norm": 2.0610591595121157,
+      "learning_rate": 3.2635011462598145e-05,
+      "loss": 0.9417,
+      "step": 1692
+    },
+    {
+      "epoch": 2.4956697991523864,
+      "grad_norm": 1.949729409150316,
+      "learning_rate": 3.261616141569581e-05,
+      "loss": 0.9508,
+      "step": 1693
+    },
+    {
+      "epoch": 2.4971439100792336,
+      "grad_norm": 2.2995523431754945,
+      "learning_rate": 3.2597306594682225e-05,
+      "loss": 0.9785,
+      "step": 1694
+    },
+    {
+      "epoch": 2.498618021006081,
+      "grad_norm": 2.102322806059451,
+      "learning_rate": 3.257844701137633e-05,
+      "loss": 1.1082,
+      "step": 1695
+    },
+    {
+      "epoch": 2.500092131932928,
+      "grad_norm": 2.1459871502408916,
+      "learning_rate": 3.255958267760006e-05,
+      "loss": 1.0246,
+      "step": 1696
+    },
+    {
+      "epoch": 2.5015662428597754,
+      "grad_norm": 2.095364417927057,
+      "learning_rate": 3.254071360517833e-05,
+      "loss": 1.1134,
+      "step": 1697
+    },
+    {
+      "epoch": 2.5030403537866226,
+      "grad_norm": 1.9574809823506967,
+      "learning_rate": 3.252183980593901e-05,
+      "loss": 0.9581,
+      "step": 1698
+    },
+    {
+      "epoch": 2.50451446471347,
+      "grad_norm": 2.0274580553939545,
+      "learning_rate": 3.250296129171295e-05,
+      "loss": 0.9124,
+      "step": 1699
+    },
+    {
+      "epoch": 2.505988575640317,
+      "grad_norm": 2.241454320254508,
+      "learning_rate": 3.2484078074333954e-05,
+      "loss": 0.8277,
+      "step": 1700
+    },
+    {
+      "epoch": 2.5074626865671643,
+      "grad_norm": 1.9206174131939369,
+      "learning_rate": 3.246519016563876e-05,
+      "loss": 0.9121,
+      "step": 1701
+    },
+    {
+      "epoch": 2.5089367974940116,
+      "grad_norm": 2.098198790333586,
+      "learning_rate": 3.244629757746706e-05,
+      "loss": 0.8616,
+      "step": 1702
+    },
+    {
+      "epoch": 2.510410908420859,
+      "grad_norm": 1.9350718387802128,
+      "learning_rate": 3.242740032166149e-05,
+      "loss": 0.979,
+      "step": 1703
+    },
+    {
+      "epoch": 2.511885019347706,
+      "grad_norm": 2.2151144625880947,
+      "learning_rate": 3.240849841006758e-05,
+      "loss": 0.9639,
+      "step": 1704
+    },
+    {
+      "epoch": 2.5133591302745533,
+      "grad_norm": 2.1544258030677392,
+      "learning_rate": 3.23895918545338e-05,
+      "loss": 0.851,
+      "step": 1705
+    },
+    {
+      "epoch": 2.5148332412014005,
+      "grad_norm": 2.156089223866205,
+      "learning_rate": 3.237068066691152e-05,
+      "loss": 1.1099,
+      "step": 1706
+    },
+    {
+      "epoch": 2.5163073521282477,
+      "grad_norm": 2.150213554776883,
+      "learning_rate": 3.2351764859055034e-05,
+      "loss": 1.1039,
+      "step": 1707
+    },
+    {
+      "epoch": 2.517781463055095,
+      "grad_norm": 1.9318008896364856,
+      "learning_rate": 3.233284444282152e-05,
+      "loss": 0.8274,
+      "step": 1708
+    },
+    {
+      "epoch": 2.519255573981942,
+      "grad_norm": 2.0140289374870353,
+      "learning_rate": 3.2313919430071026e-05,
+      "loss": 0.9672,
+      "step": 1709
+    },
+    {
+      "epoch": 2.5207296849087895,
+      "grad_norm": 1.9871690558561252,
+      "learning_rate": 3.2294989832666514e-05,
+      "loss": 0.9749,
+      "step": 1710
+    },
+    {
+      "epoch": 2.5222037958356367,
+      "grad_norm": 1.900264786791097,
+      "learning_rate": 3.22760556624738e-05,
+      "loss": 0.9303,
+      "step": 1711
+    },
+    {
+      "epoch": 2.523677906762484,
+      "grad_norm": 2.2311870534015847,
+      "learning_rate": 3.225711693136156e-05,
+      "loss": 1.0793,
+      "step": 1712
+    },
+    {
+      "epoch": 2.525152017689331,
+      "grad_norm": 2.1005053502527504,
+      "learning_rate": 3.223817365120136e-05,
+      "loss": 0.8998,
+      "step": 1713
+    },
+    {
+      "epoch": 2.5266261286161784,
+      "grad_norm": 2.121638837433541,
+      "learning_rate": 3.221922583386758e-05,
+      "loss": 0.9425,
+      "step": 1714
+    },
+    {
+      "epoch": 2.5281002395430257,
+      "grad_norm": 2.2763327289527804,
+      "learning_rate": 3.220027349123748e-05,
+      "loss": 1.0774,
+      "step": 1715
+    },
+    {
+      "epoch": 2.529574350469873,
+      "grad_norm": 2.189693441142029,
+      "learning_rate": 3.2181316635191125e-05,
+      "loss": 0.9742,
+      "step": 1716
+    },
+    {
+      "epoch": 2.53104846139672,
+      "grad_norm": 2.0344893228546495,
+      "learning_rate": 3.2162355277611416e-05,
+      "loss": 0.8612,
+      "step": 1717
+    },
+    {
+      "epoch": 2.5325225723235674,
+      "grad_norm": 1.99004281481974,
+      "learning_rate": 3.214338943038409e-05,
+      "loss": 0.9278,
+      "step": 1718
+    },
+    {
+      "epoch": 2.5339966832504146,
+      "grad_norm": 1.8806389256599139,
+      "learning_rate": 3.21244191053977e-05,
+      "loss": 0.8617,
+      "step": 1719
+    },
+    {
+      "epoch": 2.535470794177262,
+      "grad_norm": 2.126287743324814,
+      "learning_rate": 3.2105444314543584e-05,
+      "loss": 0.9358,
+      "step": 1720
+    },
+    {
+      "epoch": 2.536944905104109,
+      "grad_norm": 2.4017904993658874,
+      "learning_rate": 3.208646506971589e-05,
+      "loss": 1.2069,
+      "step": 1721
+    },
+    {
+      "epoch": 2.5384190160309563,
+      "grad_norm": 2.0375225931798906,
+      "learning_rate": 3.206748138281157e-05,
+      "loss": 1.0304,
+      "step": 1722
+    },
+    {
+      "epoch": 2.5398931269578036,
+      "grad_norm": 2.0652366746399387,
+      "learning_rate": 3.204849326573034e-05,
+      "loss": 0.9532,
+      "step": 1723
+    },
+    {
+      "epoch": 2.541367237884651,
+      "grad_norm": 2.0550017025979965,
+      "learning_rate": 3.20295007303747e-05,
+      "loss": 0.856,
+      "step": 1724
+    },
+    {
+      "epoch": 2.542841348811498,
+      "grad_norm": 2.160482040588414,
+      "learning_rate": 3.201050378864994e-05,
+      "loss": 0.9894,
+      "step": 1725
+    },
+    {
+      "epoch": 2.5443154597383453,
+      "grad_norm": 2.0283493458667747,
+      "learning_rate": 3.1991502452464074e-05,
+      "loss": 0.87,
+      "step": 1726
+    },
+    {
+      "epoch": 2.5457895706651925,
+      "grad_norm": 2.017427451756768,
+      "learning_rate": 3.1972496733727906e-05,
+      "loss": 1.1002,
+      "step": 1727
+    },
+    {
+      "epoch": 2.5472636815920398,
+      "grad_norm": 2.1190578180825916,
+      "learning_rate": 3.195348664435497e-05,
+      "loss": 0.7762,
+      "step": 1728
+    },
+    {
+      "epoch": 2.548737792518887,
+      "grad_norm": 2.109400883268571,
+      "learning_rate": 3.193447219626153e-05,
+      "loss": 0.9088,
+      "step": 1729
+    },
+    {
+      "epoch": 2.5502119034457342,
+      "grad_norm": 1.9671511742932197,
+      "learning_rate": 3.191545340136661e-05,
+      "loss": 0.8853,
+      "step": 1730
+    },
+    {
+      "epoch": 2.5516860143725815,
+      "grad_norm": 2.164148211933585,
+      "learning_rate": 3.1896430271591937e-05,
+      "loss": 1.076,
+      "step": 1731
+    },
+    {
+      "epoch": 2.5531601252994287,
+      "grad_norm": 2.2774694810449305,
+      "learning_rate": 3.187740281886195e-05,
+      "loss": 0.9779,
+      "step": 1732
+    },
+    {
+      "epoch": 2.554634236226276,
+      "grad_norm": 2.188016449045282,
+      "learning_rate": 3.185837105510383e-05,
+      "loss": 0.8944,
+      "step": 1733
+    },
+    {
+      "epoch": 2.556108347153123,
+      "grad_norm": 2.145532037514011,
+      "learning_rate": 3.183933499224743e-05,
+      "loss": 1.0104,
+      "step": 1734
+    },
+    {
+      "epoch": 2.5575824580799704,
+      "grad_norm": 2.2024180042189125,
+      "learning_rate": 3.18202946422253e-05,
+      "loss": 0.9971,
+      "step": 1735
+    },
+    {
+      "epoch": 2.5590565690068177,
+      "grad_norm": 2.019033935231454,
+      "learning_rate": 3.18012500169727e-05,
+      "loss": 0.9574,
+      "step": 1736
+    },
+    {
+      "epoch": 2.560530679933665,
+      "grad_norm": 2.1676034980535923,
+      "learning_rate": 3.178220112842753e-05,
+      "loss": 0.9253,
+      "step": 1737
+    },
+    {
+      "epoch": 2.562004790860512,
+      "grad_norm": 2.108396288970941,
+      "learning_rate": 3.176314798853042e-05,
+      "loss": 0.8269,
+      "step": 1738
+    },
+    {
+      "epoch": 2.5634789017873594,
+      "grad_norm": 2.118128606095879,
+      "learning_rate": 3.17440906092246e-05,
+      "loss": 0.9522,
+      "step": 1739
+    },
+    {
+      "epoch": 2.5649530127142066,
+      "grad_norm": 2.2306281198194924,
+      "learning_rate": 3.1725029002456e-05,
+      "loss": 0.8739,
+      "step": 1740
+    },
+    {
+      "epoch": 2.566427123641054,
+      "grad_norm": 2.1995455836112026,
+      "learning_rate": 3.17059631801732e-05,
+      "loss": 0.9796,
+      "step": 1741
+    },
+    {
+      "epoch": 2.567901234567901,
+      "grad_norm": 2.0279094516847094,
+      "learning_rate": 3.168689315432741e-05,
+      "loss": 0.7954,
+      "step": 1742
+    },
+    {
+      "epoch": 2.5693753454947483,
+      "grad_norm": 2.15323483390432,
+      "learning_rate": 3.1667818936872465e-05,
+      "loss": 0.9948,
+      "step": 1743
+    },
+    {
+      "epoch": 2.5708494564215956,
+      "grad_norm": 2.214436811574459,
+      "learning_rate": 3.1648740539764844e-05,
+      "loss": 0.9673,
+      "step": 1744
+    },
+    {
+      "epoch": 2.572323567348443,
+      "grad_norm": 2.2543091306483807,
+      "learning_rate": 3.162965797496364e-05,
+      "loss": 1.0563,
+      "step": 1745
+    },
+    {
+      "epoch": 2.57379767827529,
+      "grad_norm": 1.859509631499959,
+      "learning_rate": 3.161057125443056e-05,
+      "loss": 0.7734,
+      "step": 1746
+    },
+    {
+      "epoch": 2.5752717892021373,
+      "grad_norm": 2.0305983451768137,
+      "learning_rate": 3.1591480390129914e-05,
+      "loss": 0.7847,
+      "step": 1747
+    },
+    {
+      "epoch": 2.5767459001289845,
+      "grad_norm": 2.242334553751078,
+      "learning_rate": 3.157238539402862e-05,
+      "loss": 0.9543,
+      "step": 1748
+    },
+    {
+      "epoch": 2.5782200110558318,
+      "grad_norm": 2.32862920229735,
+      "learning_rate": 3.155328627809617e-05,
+      "loss": 1.0283,
+      "step": 1749
+    },
+    {
+      "epoch": 2.579694121982679,
+      "grad_norm": 2.031294602124497,
+      "learning_rate": 3.1534183054304645e-05,
+      "loss": 1.0185,
+      "step": 1750
+    },
+    {
+      "epoch": 2.579694121982679,
+      "eval_bleu": 0.055573090242479375,
+      "eval_bleu_1gram": 0.34188218281027344,
+      "eval_bleu_2gram": 0.12114871340479708,
+      "eval_bleu_3gram": 0.0520342290224777,
+      "eval_bleu_4gram": 0.026826966821431385,
+      "eval_rag_val_loss": 1.643004084466606,
+      "eval_rouge1": 0.33405043556066294,
+      "eval_rouge2": 0.11513784981928654,
+      "eval_rougeL": 0.3135612357194884,
+      "step": 1750
+    },
+    {
+      "epoch": 2.5811682329095262,
+      "grad_norm": 2.1984838202937182,
+      "learning_rate": 3.1515075734628705e-05,
+      "loss": 0.8624,
+      "step": 1751
+    },
+    {
+      "epoch": 2.5826423438363735,
+      "grad_norm": 2.2407784636865675,
+      "learning_rate": 3.149596433104556e-05,
+      "loss": 1.0364,
+      "step": 1752
+    },
+    {
+      "epoch": 2.5841164547632207,
+      "grad_norm": 2.2441090713553575,
+      "learning_rate": 3.147684885553502e-05,
+      "loss": 0.9345,
+      "step": 1753
+    },
+    {
+      "epoch": 2.585590565690068,
+      "grad_norm": 2.416179728504946,
+      "learning_rate": 3.145772932007939e-05,
+      "loss": 0.9889,
+      "step": 1754
+    },
+    {
+      "epoch": 2.587064676616915,
+      "grad_norm": 2.243337834577121,
+      "learning_rate": 3.143860573666357e-05,
+      "loss": 0.7987,
+      "step": 1755
+    },
+    {
+      "epoch": 2.5885387875437624,
+      "grad_norm": 2.083553022881279,
+      "learning_rate": 3.1419478117274984e-05,
+      "loss": 1.007,
+      "step": 1756
+    },
+    {
+      "epoch": 2.5900128984706097,
+      "grad_norm": 2.3068097187438785,
+      "learning_rate": 3.140034647390357e-05,
+      "loss": 1.1439,
+      "step": 1757
+    },
+    {
+      "epoch": 2.591487009397457,
+      "grad_norm": 2.0400043519291553,
+      "learning_rate": 3.13812108185418e-05,
+      "loss": 0.9338,
+      "step": 1758
+    },
+    {
+      "epoch": 2.592961120324304,
+      "grad_norm": 2.1313970431169587,
+      "learning_rate": 3.136207116318466e-05,
+      "loss": 0.9051,
+      "step": 1759
+    },
+    {
+      "epoch": 2.5944352312511514,
+      "grad_norm": 2.5250123014244843,
+      "learning_rate": 3.1342927519829644e-05,
+      "loss": 1.1853,
+      "step": 1760
+    },
+    {
+      "epoch": 2.5959093421779986,
+      "grad_norm": 1.9879246599938454,
+      "learning_rate": 3.1323779900476744e-05,
+      "loss": 0.9466,
+      "step": 1761
+    },
+    {
+      "epoch": 2.597383453104846,
+      "grad_norm": 2.1238132977945967,
+      "learning_rate": 3.1304628317128446e-05,
+      "loss": 0.9648,
+      "step": 1762
+    },
+    {
+      "epoch": 2.5988575640316935,
+      "grad_norm": 2.0564555984255772,
+      "learning_rate": 3.128547278178972e-05,
+      "loss": 1.0392,
+      "step": 1763
+    },
+    {
+      "epoch": 2.600331674958541,
+      "grad_norm": 2.1716716863301433,
+      "learning_rate": 3.126631330646802e-05,
+      "loss": 0.9093,
+      "step": 1764
+    },
+    {
+      "epoch": 2.601805785885388,
+      "grad_norm": 2.081155617896485,
+      "learning_rate": 3.124714990317324e-05,
+      "loss": 1.0173,
+      "step": 1765
+    },
+    {
+      "epoch": 2.6032798968122353,
+      "grad_norm": 1.979675435042255,
+      "learning_rate": 3.122798258391779e-05,
+      "loss": 1.1098,
+      "step": 1766
+    },
+    {
+      "epoch": 2.6047540077390825,
+      "grad_norm": 1.9700741629026115,
+      "learning_rate": 3.120881136071649e-05,
+      "loss": 0.9434,
+      "step": 1767
+    },
+    {
+      "epoch": 2.6062281186659297,
+      "grad_norm": 1.9446365836746766,
+      "learning_rate": 3.118963624558662e-05,
+      "loss": 0.8653,
+      "step": 1768
+    },
+    {
+      "epoch": 2.607702229592777,
+      "grad_norm": 2.196249330940644,
+      "learning_rate": 3.11704572505479e-05,
+      "loss": 0.9166,
+      "step": 1769
+    },
+    {
+      "epoch": 2.609176340519624,
+      "grad_norm": 2.1288570853339617,
+      "learning_rate": 3.115127438762247e-05,
+      "loss": 1.0654,
+      "step": 1770
+    },
+    {
+      "epoch": 2.6106504514464715,
+      "grad_norm": 2.022968134698401,
+      "learning_rate": 3.113208766883494e-05,
+      "loss": 0.9494,
+      "step": 1771
+    },
+    {
+      "epoch": 2.6121245623733187,
+      "grad_norm": 1.8923497175365587,
+      "learning_rate": 3.111289710621228e-05,
+      "loss": 0.9244,
+      "step": 1772
+    },
+    {
+      "epoch": 2.613598673300166,
+      "grad_norm": 2.191640368788528,
+      "learning_rate": 3.109370271178389e-05,
+      "loss": 1.067,
+      "step": 1773
+    },
+    {
+      "epoch": 2.615072784227013,
+      "grad_norm": 2.0718999570606376,
+      "learning_rate": 3.10745044975816e-05,
+      "loss": 0.8979,
+      "step": 1774
+    },
+    {
+      "epoch": 2.6165468951538604,
+      "grad_norm": 2.192556368408781,
+      "learning_rate": 3.1055302475639594e-05,
+      "loss": 1.0057,
+      "step": 1775
+    },
+    {
+      "epoch": 2.6180210060807076,
+      "grad_norm": 2.0596661606431126,
+      "learning_rate": 3.103609665799445e-05,
+      "loss": 0.9229,
+      "step": 1776
+    },
+    {
+      "epoch": 2.619495117007555,
+      "grad_norm": 2.084100721806418,
+      "learning_rate": 3.1016887056685155e-05,
+      "loss": 0.9164,
+      "step": 1777
+    },
+    {
+      "epoch": 2.620969227934402,
+      "grad_norm": 2.1277173563537644,
+      "learning_rate": 3.0997673683753024e-05,
+      "loss": 0.9373,
+      "step": 1778
+    },
+    {
+      "epoch": 2.6224433388612494,
+      "grad_norm": 2.196438755007362,
+      "learning_rate": 3.0978456551241786e-05,
+      "loss": 0.868,
+      "step": 1779
+    },
+    {
+      "epoch": 2.6239174497880966,
+      "grad_norm": 2.0005884496942516,
+      "learning_rate": 3.095923567119748e-05,
+      "loss": 0.8726,
+      "step": 1780
+    },
+    {
+      "epoch": 2.625391560714944,
+      "grad_norm": 2.08627545194374,
+      "learning_rate": 3.094001105566852e-05,
+      "loss": 0.9243,
+      "step": 1781
+    },
+    {
+      "epoch": 2.626865671641791,
+      "grad_norm": 2.1493808963891095,
+      "learning_rate": 3.0920782716705654e-05,
+      "loss": 0.8746,
+      "step": 1782
+    },
+    {
+      "epoch": 2.6283397825686383,
+      "grad_norm": 2.2030218654868663,
+      "learning_rate": 3.0901550666361964e-05,
+      "loss": 1.0587,
+      "step": 1783
+    },
+    {
+      "epoch": 2.6298138934954856,
+      "grad_norm": 2.2500929283449036,
+      "learning_rate": 3.088231491669287e-05,
+      "loss": 1.1282,
+      "step": 1784
+    },
+    {
+      "epoch": 2.631288004422333,
+      "grad_norm": 2.0809582205028807,
+      "learning_rate": 3.0863075479756084e-05,
+      "loss": 0.9671,
+      "step": 1785
+    },
+    {
+      "epoch": 2.63276211534918,
+      "grad_norm": 2.061466304795124,
+      "learning_rate": 3.084383236761166e-05,
+      "loss": 0.9586,
+      "step": 1786
+    },
+    {
+      "epoch": 2.6342362262760273,
+      "grad_norm": 2.4975333442791374,
+      "learning_rate": 3.0824585592321936e-05,
+      "loss": 1.0932,
+      "step": 1787
+    },
+    {
+      "epoch": 2.6357103372028745,
+      "grad_norm": 2.258729952885257,
+      "learning_rate": 3.080533516595155e-05,
+      "loss": 1.0924,
+      "step": 1788
+    },
+    {
+      "epoch": 2.6371844481297217,
+      "grad_norm": 2.154876271388944,
+      "learning_rate": 3.078608110056745e-05,
+      "loss": 1.0354,
+      "step": 1789
+    },
+    {
+      "epoch": 2.638658559056569,
+      "grad_norm": 2.024379672886872,
+      "learning_rate": 3.076682340823882e-05,
+      "loss": 1.0128,
+      "step": 1790
+    },
+    {
+      "epoch": 2.6401326699834162,
+      "grad_norm": 2.168539436078495,
+      "learning_rate": 3.074756210103715e-05,
+      "loss": 0.9384,
+      "step": 1791
+    },
+    {
+      "epoch": 2.6416067809102635,
+      "grad_norm": 2.062797640507273,
+      "learning_rate": 3.072829719103619e-05,
+      "loss": 0.9524,
+      "step": 1792
+    },
+    {
+      "epoch": 2.6430808918371107,
+      "grad_norm": 2.2362241762481614,
+      "learning_rate": 3.070902869031196e-05,
+      "loss": 0.8175,
+      "step": 1793
+    },
+    {
+      "epoch": 2.644555002763958,
+      "grad_norm": 1.9729043634575534,
+      "learning_rate": 3.0689756610942705e-05,
+      "loss": 0.8934,
+      "step": 1794
+    },
+    {
+      "epoch": 2.646029113690805,
+      "grad_norm": 2.049923323150976,
+      "learning_rate": 3.067048096500893e-05,
+      "loss": 0.829,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6475032246176524,
+      "grad_norm": 2.059977520570044,
+      "learning_rate": 3.065120176459338e-05,
+      "loss": 0.9228,
+      "step": 1796
+    },
+    {
+      "epoch": 2.6489773355444997,
+      "grad_norm": 2.12026022591524,
+      "learning_rate": 3.0631919021781e-05,
+      "loss": 0.985,
+      "step": 1797
+    },
+    {
+      "epoch": 2.650451446471347,
+      "grad_norm": 2.1315320541473604,
+      "learning_rate": 3.0612632748659e-05,
+      "loss": 0.9291,
+      "step": 1798
+    },
+    {
+      "epoch": 2.651925557398194,
+      "grad_norm": 2.2521885718201107,
+      "learning_rate": 3.0593342957316765e-05,
+      "loss": 0.9138,
+      "step": 1799
+    },
+    {
+      "epoch": 2.6533996683250414,
+      "grad_norm": 2.2148548374874046,
+      "learning_rate": 3.05740496598459e-05,
+      "loss": 0.9376,
+      "step": 1800
+    },
+    {
+      "epoch": 2.6548737792518886,
+      "grad_norm": 2.041306353583633,
+      "learning_rate": 3.055475286834021e-05,
+      "loss": 0.9404,
+      "step": 1801
+    },
+    {
+      "epoch": 2.656347890178736,
+      "grad_norm": 2.1500605330706204,
+      "learning_rate": 3.053545259489569e-05,
+      "loss": 0.924,
+      "step": 1802
+    },
+    {
+      "epoch": 2.657822001105583,
+      "grad_norm": 2.1694027668619222,
+      "learning_rate": 3.051614885161051e-05,
+      "loss": 0.888,
+      "step": 1803
+    },
+    {
+      "epoch": 2.6592961120324303,
+      "grad_norm": 2.186691025602148,
+      "learning_rate": 3.0496841650585022e-05,
+      "loss": 0.9828,
+      "step": 1804
+    },
+    {
+      "epoch": 2.6607702229592776,
+      "grad_norm": 2.3808591747205616,
+      "learning_rate": 3.0477531003921745e-05,
+      "loss": 1.0999,
+      "step": 1805
+    },
+    {
+      "epoch": 2.662244333886125,
+      "grad_norm": 2.0975944183158957,
+      "learning_rate": 3.0458216923725356e-05,
+      "loss": 0.9419,
+      "step": 1806
+    },
+    {
+      "epoch": 2.663718444812972,
+      "grad_norm": 2.0947238308548624,
+      "learning_rate": 3.043889942210268e-05,
+      "loss": 0.8704,
+      "step": 1807
+    },
+    {
+      "epoch": 2.6651925557398193,
+      "grad_norm": 1.972787018020252,
+      "learning_rate": 3.0419578511162695e-05,
+      "loss": 0.9647,
+      "step": 1808
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 2.068045241984393,
+      "learning_rate": 3.0400254203016503e-05,
+      "loss": 0.9381,
+      "step": 1809
+    },
+    {
+      "epoch": 2.6681407775935138,
+      "grad_norm": 2.1525107871038958,
+      "learning_rate": 3.0380926509777364e-05,
+      "loss": 1.1002,
+      "step": 1810
+    },
+    {
+      "epoch": 2.669614888520361,
+      "grad_norm": 1.995996401926899,
+      "learning_rate": 3.0361595443560624e-05,
+      "loss": 1.0872,
+      "step": 1811
+    },
+    {
+      "epoch": 2.6710889994472087,
+      "grad_norm": 2.041661385769405,
+      "learning_rate": 3.034226101648377e-05,
+      "loss": 1.0054,
+      "step": 1812
+    },
+    {
+      "epoch": 2.672563110374056,
+      "grad_norm": 2.2057666497646826,
+      "learning_rate": 3.0322923240666377e-05,
+      "loss": 0.9795,
+      "step": 1813
+    },
+    {
+      "epoch": 2.674037221300903,
+      "grad_norm": 2.050299538849172,
+      "learning_rate": 3.030358212823014e-05,
+      "loss": 0.9672,
+      "step": 1814
+    },
+    {
+      "epoch": 2.6755113322277504,
+      "grad_norm": 2.051742589967626,
+      "learning_rate": 3.0284237691298823e-05,
+      "loss": 0.9873,
+      "step": 1815
+    },
+    {
+      "epoch": 2.6769854431545976,
+      "grad_norm": 1.9295027795016872,
+      "learning_rate": 3.0264889941998285e-05,
+      "loss": 0.8804,
+      "step": 1816
+    },
+    {
+      "epoch": 2.678459554081445,
+      "grad_norm": 2.180346755001575,
+      "learning_rate": 3.0245538892456455e-05,
+      "loss": 1.0665,
+      "step": 1817
+    },
+    {
+      "epoch": 2.679933665008292,
+      "grad_norm": 2.0464484083288443,
+      "learning_rate": 3.0226184554803357e-05,
+      "loss": 0.9575,
+      "step": 1818
+    },
+    {
+      "epoch": 2.6814077759351393,
+      "grad_norm": 2.08952953792962,
+      "learning_rate": 3.0206826941171035e-05,
+      "loss": 0.9013,
+      "step": 1819
+    },
+    {
+      "epoch": 2.6828818868619866,
+      "grad_norm": 2.3457320923807248,
+      "learning_rate": 3.0187466063693614e-05,
+      "loss": 0.9603,
+      "step": 1820
+    },
+    {
+      "epoch": 2.684355997788834,
+      "grad_norm": 2.047790439113738,
+      "learning_rate": 3.0168101934507266e-05,
+      "loss": 1.0001,
+      "step": 1821
+    },
+    {
+      "epoch": 2.685830108715681,
+      "grad_norm": 2.1518726975746922,
+      "learning_rate": 3.0148734565750176e-05,
+      "loss": 0.9236,
+      "step": 1822
+    },
+    {
+      "epoch": 2.6873042196425283,
+      "grad_norm": 2.2367997251103877,
+      "learning_rate": 3.012936396956259e-05,
+      "loss": 0.9916,
+      "step": 1823
+    },
+    {
+      "epoch": 2.6887783305693755,
+      "grad_norm": 2.3769347692453224,
+      "learning_rate": 3.0109990158086764e-05,
+      "loss": 0.8328,
+      "step": 1824
+    },
+    {
+      "epoch": 2.690252441496223,
+      "grad_norm": 2.0610036334691046,
+      "learning_rate": 3.0090613143466956e-05,
+      "loss": 0.9659,
+      "step": 1825
+    },
+    {
+      "epoch": 2.69172655242307,
+      "grad_norm": 2.233069405214683,
+      "learning_rate": 3.0071232937849457e-05,
+      "loss": 0.9709,
+      "step": 1826
+    },
+    {
+      "epoch": 2.6932006633499173,
+      "grad_norm": 2.154831903721188,
+      "learning_rate": 3.0051849553382555e-05,
+      "loss": 0.9363,
+      "step": 1827
+    },
+    {
+      "epoch": 2.6946747742767645,
+      "grad_norm": 2.0612165908465574,
+      "learning_rate": 3.0032463002216505e-05,
+      "loss": 0.925,
+      "step": 1828
+    },
+    {
+      "epoch": 2.6961488852036117,
+      "grad_norm": 2.233269905024685,
+      "learning_rate": 3.001307329650357e-05,
+      "loss": 0.9821,
+      "step": 1829
+    },
+    {
+      "epoch": 2.697622996130459,
+      "grad_norm": 2.1985717689055693,
+      "learning_rate": 2.9993680448397988e-05,
+      "loss": 1.0333,
+      "step": 1830
+    },
+    {
+      "epoch": 2.699097107057306,
+      "grad_norm": 2.065792230482113,
+      "learning_rate": 2.997428447005596e-05,
+      "loss": 0.9076,
+      "step": 1831
+    },
+    {
+      "epoch": 2.7005712179841534,
+      "grad_norm": 2.164533763485579,
+      "learning_rate": 2.9954885373635655e-05,
+      "loss": 1.1866,
+      "step": 1832
+    },
+    {
+      "epoch": 2.7020453289110007,
+      "grad_norm": 2.258621546054044,
+      "learning_rate": 2.9935483171297186e-05,
+      "loss": 1.0448,
+      "step": 1833
+    },
+    {
+      "epoch": 2.703519439837848,
+      "grad_norm": 2.198199237419151,
+      "learning_rate": 2.991607787520263e-05,
+      "loss": 0.9786,
+      "step": 1834
+    },
+    {
+      "epoch": 2.704993550764695,
+      "grad_norm": 1.947070082294161,
+      "learning_rate": 2.989666949751599e-05,
+      "loss": 0.98,
+      "step": 1835
+    },
+    {
+      "epoch": 2.7064676616915424,
+      "grad_norm": 2.1596233131299214,
+      "learning_rate": 2.9877258050403212e-05,
+      "loss": 0.9382,
+      "step": 1836
+    },
+    {
+      "epoch": 2.7079417726183896,
+      "grad_norm": 1.9660084724864615,
+      "learning_rate": 2.985784354603215e-05,
+      "loss": 0.8798,
+      "step": 1837
+    },
+    {
+      "epoch": 2.709415883545237,
+      "grad_norm": 2.124654854064859,
+      "learning_rate": 2.9838425996572583e-05,
+      "loss": 0.9097,
+      "step": 1838
+    },
+    {
+      "epoch": 2.710889994472084,
+      "grad_norm": 2.0373399262063634,
+      "learning_rate": 2.981900541419621e-05,
+      "loss": 1.073,
+      "step": 1839
+    },
+    {
+      "epoch": 2.7123641053989314,
+      "grad_norm": 2.366214262566019,
+      "learning_rate": 2.9799581811076605e-05,
+      "loss": 1.0424,
+      "step": 1840
+    },
+    {
+      "epoch": 2.7138382163257786,
+      "grad_norm": 2.2303612143653795,
+      "learning_rate": 2.978015519938926e-05,
+      "loss": 0.9905,
+      "step": 1841
+    },
+    {
+      "epoch": 2.715312327252626,
+      "grad_norm": 1.9686715852168089,
+      "learning_rate": 2.9760725591311545e-05,
+      "loss": 0.8199,
+      "step": 1842
+    },
+    {
+      "epoch": 2.716786438179473,
+      "grad_norm": 2.122109859991592,
+      "learning_rate": 2.9741292999022707e-05,
+      "loss": 0.9643,
+      "step": 1843
+    },
+    {
+      "epoch": 2.7182605491063203,
+      "grad_norm": 2.139339534410948,
+      "learning_rate": 2.9721857434703858e-05,
+      "loss": 0.8441,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7197346600331676,
+      "grad_norm": 2.169012914546158,
+      "learning_rate": 2.9702418910537983e-05,
+      "loss": 0.9663,
+      "step": 1845
+    },
+    {
+      "epoch": 2.721208770960015,
+      "grad_norm": 2.1349947877056454,
+      "learning_rate": 2.9682977438709914e-05,
+      "loss": 0.9757,
+      "step": 1846
+    },
+    {
+      "epoch": 2.722682881886862,
+      "grad_norm": 2.2382260726864955,
+      "learning_rate": 2.9663533031406344e-05,
+      "loss": 1.0964,
+      "step": 1847
+    },
+    {
+      "epoch": 2.7241569928137093,
+      "grad_norm": 2.1530888923972586,
+      "learning_rate": 2.9644085700815777e-05,
+      "loss": 0.9852,
+      "step": 1848
+    },
+    {
+      "epoch": 2.7256311037405565,
+      "grad_norm": 2.013907595434308,
+      "learning_rate": 2.9624635459128585e-05,
+      "loss": 0.7825,
+      "step": 1849
+    },
+    {
+      "epoch": 2.7271052146674037,
+      "grad_norm": 2.343103039776737,
+      "learning_rate": 2.960518231853695e-05,
+      "loss": 0.9298,
+      "step": 1850
+    },
+    {
+      "epoch": 2.728579325594251,
+      "grad_norm": 2.180973999575077,
+      "learning_rate": 2.9585726291234872e-05,
+      "loss": 0.9289,
+      "step": 1851
+    },
+    {
+      "epoch": 2.730053436521098,
+      "grad_norm": 2.0294113053486362,
+      "learning_rate": 2.9566267389418144e-05,
+      "loss": 1.0403,
+      "step": 1852
+    },
+    {
+      "epoch": 2.7315275474479455,
+      "grad_norm": 1.9998146805735708,
+      "learning_rate": 2.9546805625284384e-05,
+      "loss": 0.9106,
+      "step": 1853
+    },
+    {
+      "epoch": 2.7330016583747927,
+      "grad_norm": 2.129740644255068,
+      "learning_rate": 2.9527341011033e-05,
+      "loss": 1.0296,
+      "step": 1854
+    },
+    {
+      "epoch": 2.73447576930164,
+      "grad_norm": 2.1004496183902512,
+      "learning_rate": 2.9507873558865175e-05,
+      "loss": 1.0393,
+      "step": 1855
+    },
+    {
+      "epoch": 2.735949880228487,
+      "grad_norm": 2.2503740211725822,
+      "learning_rate": 2.9488403280983873e-05,
+      "loss": 1.0947,
+      "step": 1856
+    },
+    {
+      "epoch": 2.7374239911553344,
+      "grad_norm": 2.1085446454125645,
+      "learning_rate": 2.9468930189593845e-05,
+      "loss": 0.9323,
+      "step": 1857
+    },
+    {
+      "epoch": 2.7388981020821817,
+      "grad_norm": 2.296346383021392,
+      "learning_rate": 2.9449454296901603e-05,
+      "loss": 0.8923,
+      "step": 1858
+    },
+    {
+      "epoch": 2.740372213009029,
+      "grad_norm": 2.020647754865544,
+      "learning_rate": 2.9429975615115383e-05,
+      "loss": 1.1155,
+      "step": 1859
+    },
+    {
+      "epoch": 2.741846323935876,
+      "grad_norm": 2.0743597812444876,
+      "learning_rate": 2.9410494156445216e-05,
+      "loss": 1.0538,
+      "step": 1860
+    },
+    {
+      "epoch": 2.7433204348627234,
+      "grad_norm": 2.1427226682884144,
+      "learning_rate": 2.9391009933102836e-05,
+      "loss": 0.9989,
+      "step": 1861
+    },
+    {
+      "epoch": 2.7447945457895706,
+      "grad_norm": 1.99638361608485,
+      "learning_rate": 2.9371522957301734e-05,
+      "loss": 0.9557,
+      "step": 1862
+    },
+    {
+      "epoch": 2.746268656716418,
+      "grad_norm": 2.0440211311636567,
+      "learning_rate": 2.935203324125711e-05,
+      "loss": 0.9788,
+      "step": 1863
+    },
+    {
+      "epoch": 2.747742767643265,
+      "grad_norm": 2.3555224967632062,
+      "learning_rate": 2.9332540797185892e-05,
+      "loss": 1.141,
+      "step": 1864
+    },
+    {
+      "epoch": 2.7492168785701123,
+      "grad_norm": 2.146658643012522,
+      "learning_rate": 2.9313045637306714e-05,
+      "loss": 0.9468,
+      "step": 1865
+    },
+    {
+      "epoch": 2.7506909894969596,
+      "grad_norm": 2.0289027810755624,
+      "learning_rate": 2.9293547773839917e-05,
+      "loss": 1.1054,
+      "step": 1866
+    },
+    {
+      "epoch": 2.752165100423807,
+      "grad_norm": 2.0019487185546017,
+      "learning_rate": 2.9274047219007534e-05,
+      "loss": 0.7559,
+      "step": 1867
+    },
+    {
+      "epoch": 2.753639211350654,
+      "grad_norm": 2.1149406193903797,
+      "learning_rate": 2.925454398503328e-05,
+      "loss": 0.9985,
+      "step": 1868
+    },
+    {
+      "epoch": 2.7551133222775013,
+      "grad_norm": 1.9260872680921393,
+      "learning_rate": 2.9235038084142557e-05,
+      "loss": 0.8982,
+      "step": 1869
+    },
+    {
+      "epoch": 2.7565874332043485,
+      "grad_norm": 2.2864158354150113,
+      "learning_rate": 2.921552952856243e-05,
+      "loss": 1.016,
+      "step": 1870
+    },
+    {
+      "epoch": 2.7580615441311958,
+      "grad_norm": 1.9241097336126731,
+      "learning_rate": 2.919601833052163e-05,
+      "loss": 0.925,
+      "step": 1871
+    },
+    {
+      "epoch": 2.759535655058043,
+      "grad_norm": 2.2882905771697715,
+      "learning_rate": 2.9176504502250563e-05,
+      "loss": 0.9278,
+      "step": 1872
+    },
+    {
+      "epoch": 2.7610097659848902,
+      "grad_norm": 2.051962782776982,
+      "learning_rate": 2.9156988055981254e-05,
+      "loss": 0.9639,
+      "step": 1873
+    },
+    {
+      "epoch": 2.7624838769117375,
+      "grad_norm": 2.086700185723885,
+      "learning_rate": 2.9137469003947392e-05,
+      "loss": 0.9043,
+      "step": 1874
+    },
+    {
+      "epoch": 2.7639579878385847,
+      "grad_norm": 2.15568745227044,
+      "learning_rate": 2.9117947358384288e-05,
+      "loss": 0.9606,
+      "step": 1875
+    },
+    {
+      "epoch": 2.765432098765432,
+      "grad_norm": 2.1569483635053412,
+      "learning_rate": 2.909842313152888e-05,
+      "loss": 0.9616,
+      "step": 1876
+    },
+    {
+      "epoch": 2.766906209692279,
+      "grad_norm": 2.0924481572117615,
+      "learning_rate": 2.9078896335619732e-05,
+      "loss": 1.0455,
+      "step": 1877
+    },
+    {
+      "epoch": 2.7683803206191264,
+      "grad_norm": 2.1992132210534963,
+      "learning_rate": 2.9059366982897007e-05,
+      "loss": 0.9652,
+      "step": 1878
+    },
+    {
+      "epoch": 2.7698544315459737,
+      "grad_norm": 2.3053370708010936,
+      "learning_rate": 2.9039835085602473e-05,
+      "loss": 1.0063,
+      "step": 1879
+    },
+    {
+      "epoch": 2.771328542472821,
+      "grad_norm": 2.198036865515296,
+      "learning_rate": 2.9020300655979503e-05,
+      "loss": 1.1366,
+      "step": 1880
+    },
+    {
+      "epoch": 2.772802653399668,
+      "grad_norm": 2.2998020957721597,
+      "learning_rate": 2.9000763706273036e-05,
+      "loss": 1.0675,
+      "step": 1881
+    },
+    {
+      "epoch": 2.7742767643265154,
+      "grad_norm": 2.180870254820166,
+      "learning_rate": 2.8981224248729628e-05,
+      "loss": 1.113,
+      "step": 1882
+    },
+    {
+      "epoch": 2.7757508752533626,
+      "grad_norm": 1.93734082214149,
+      "learning_rate": 2.896168229559737e-05,
+      "loss": 0.8801,
+      "step": 1883
+    },
+    {
+      "epoch": 2.77722498618021,
+      "grad_norm": 2.0103423215201,
+      "learning_rate": 2.8942137859125928e-05,
+      "loss": 1.0129,
+      "step": 1884
+    },
+    {
+      "epoch": 2.778699097107057,
+      "grad_norm": 2.185063340165934,
+      "learning_rate": 2.8922590951566536e-05,
+      "loss": 0.9029,
+      "step": 1885
+    },
+    {
+      "epoch": 2.7801732080339043,
+      "grad_norm": 2.2095920526269057,
+      "learning_rate": 2.8903041585171963e-05,
+      "loss": 0.8379,
+      "step": 1886
+    },
+    {
+      "epoch": 2.7816473189607516,
+      "grad_norm": 2.1660106717354886,
+      "learning_rate": 2.8883489772196525e-05,
+      "loss": 1.0209,
+      "step": 1887
+    },
+    {
+      "epoch": 2.783121429887599,
+      "grad_norm": 2.167633635593321,
+      "learning_rate": 2.886393552489608e-05,
+      "loss": 1.1411,
+      "step": 1888
+    },
+    {
+      "epoch": 2.784595540814446,
+      "grad_norm": 2.3025102061357146,
+      "learning_rate": 2.8844378855527998e-05,
+      "loss": 0.9594,
+      "step": 1889
+    },
+    {
+      "epoch": 2.7860696517412933,
+      "grad_norm": 2.1065194475465554,
+      "learning_rate": 2.8824819776351176e-05,
+      "loss": 1.0102,
+      "step": 1890
+    },
+    {
+      "epoch": 2.7875437626681405,
+      "grad_norm": 1.9284450840310214,
+      "learning_rate": 2.8805258299626015e-05,
+      "loss": 0.9375,
+      "step": 1891
+    },
+    {
+      "epoch": 2.7890178735949878,
+      "grad_norm": 1.8859508523413062,
+      "learning_rate": 2.878569443761442e-05,
+      "loss": 0.8936,
+      "step": 1892
+    },
+    {
+      "epoch": 2.790491984521835,
+      "grad_norm": 2.1042186869117554,
+      "learning_rate": 2.8766128202579797e-05,
+      "loss": 0.9601,
+      "step": 1893
+    },
+    {
+      "epoch": 2.7919660954486822,
+      "grad_norm": 2.088913983449061,
+      "learning_rate": 2.874655960678704e-05,
+      "loss": 1.0908,
+      "step": 1894
+    },
+    {
+      "epoch": 2.79344020637553,
+      "grad_norm": 2.132596183119328,
+      "learning_rate": 2.87269886625025e-05,
+      "loss": 0.8279,
+      "step": 1895
+    },
+    {
+      "epoch": 2.794914317302377,
+      "grad_norm": 2.0647120317930434,
+      "learning_rate": 2.870741538199405e-05,
+      "loss": 0.9965,
+      "step": 1896
+    },
+    {
+      "epoch": 2.7963884282292244,
+      "grad_norm": 2.2337437518288055,
+      "learning_rate": 2.8687839777530977e-05,
+      "loss": 0.946,
+      "step": 1897
+    },
+    {
+      "epoch": 2.7978625391560716,
+      "grad_norm": 2.1167733888929954,
+      "learning_rate": 2.8668261861384045e-05,
+      "loss": 1.0438,
+      "step": 1898
+    },
+    {
+      "epoch": 2.799336650082919,
+      "grad_norm": 2.0151452964857115,
+      "learning_rate": 2.8648681645825472e-05,
+      "loss": 0.9852,
+      "step": 1899
+    },
+    {
+      "epoch": 2.800810761009766,
+      "grad_norm": 2.059927868182048,
+      "learning_rate": 2.8629099143128907e-05,
+      "loss": 1.0796,
+      "step": 1900
+    },
+    {
+      "epoch": 2.8022848719366134,
+      "grad_norm": 2.203255182843403,
+      "learning_rate": 2.860951436556944e-05,
+      "loss": 0.8487,
+      "step": 1901
+    },
+    {
+      "epoch": 2.8037589828634606,
+      "grad_norm": 2.261651072700207,
+      "learning_rate": 2.8589927325423576e-05,
+      "loss": 0.9975,
+      "step": 1902
+    },
+    {
+      "epoch": 2.805233093790308,
+      "grad_norm": 2.1972534721886143,
+      "learning_rate": 2.8570338034969264e-05,
+      "loss": 0.8591,
+      "step": 1903
+    },
+    {
+      "epoch": 2.806707204717155,
+      "grad_norm": 2.0294720424503625,
+      "learning_rate": 2.855074650648583e-05,
+      "loss": 0.9723,
+      "step": 1904
+    },
+    {
+      "epoch": 2.8081813156440023,
+      "grad_norm": 2.012433741645712,
+      "learning_rate": 2.853115275225403e-05,
+      "loss": 1.0298,
+      "step": 1905
+    },
+    {
+      "epoch": 2.8096554265708495,
+      "grad_norm": 2.0880314179099306,
+      "learning_rate": 2.8511556784556e-05,
+      "loss": 0.9103,
+      "step": 1906
+    },
+    {
+      "epoch": 2.811129537497697,
+      "grad_norm": 2.161099605589604,
+      "learning_rate": 2.8491958615675262e-05,
+      "loss": 0.8923,
+      "step": 1907
+    },
+    {
+      "epoch": 2.812603648424544,
+      "grad_norm": 2.071033855940649,
+      "learning_rate": 2.8472358257896732e-05,
+      "loss": 0.989,
+      "step": 1908
+    },
+    {
+      "epoch": 2.8140777593513913,
+      "grad_norm": 2.0227639981092995,
+      "learning_rate": 2.8452755723506687e-05,
+      "loss": 1.0256,
+      "step": 1909
+    },
+    {
+      "epoch": 2.8155518702782385,
+      "grad_norm": 1.95844058155714,
+      "learning_rate": 2.843315102479276e-05,
+      "loss": 0.9416,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8170259812050857,
+      "grad_norm": 2.118830532467706,
+      "learning_rate": 2.841354417404397e-05,
+      "loss": 0.8657,
+      "step": 1911
+    },
+    {
+      "epoch": 2.818500092131933,
+      "grad_norm": 2.1501538997837173,
+      "learning_rate": 2.8393935183550662e-05,
+      "loss": 1.0455,
+      "step": 1912
+    },
+    {
+      "epoch": 2.81997420305878,
+      "grad_norm": 2.123405082404658,
+      "learning_rate": 2.8374324065604517e-05,
+      "loss": 0.9056,
+      "step": 1913
+    },
+    {
+      "epoch": 2.8214483139856275,
+      "grad_norm": 2.0136367099181594,
+      "learning_rate": 2.8354710832498576e-05,
+      "loss": 1.0075,
+      "step": 1914
+    },
+    {
+      "epoch": 2.8229224249124747,
+      "grad_norm": 2.067461692519748,
+      "learning_rate": 2.833509549652717e-05,
+      "loss": 1.0024,
+      "step": 1915
+    },
+    {
+      "epoch": 2.824396535839322,
+      "grad_norm": 2.123742123971472,
+      "learning_rate": 2.831547806998598e-05,
+      "loss": 0.9694,
+      "step": 1916
+    },
+    {
+      "epoch": 2.825870646766169,
+      "grad_norm": 2.2401391219464584,
+      "learning_rate": 2.8295858565171983e-05,
+      "loss": 1.0497,
+      "step": 1917
+    },
+    {
+      "epoch": 2.8273447576930164,
+      "grad_norm": 1.9842866292184558,
+      "learning_rate": 2.8276236994383453e-05,
+      "loss": 0.904,
+      "step": 1918
+    },
+    {
+      "epoch": 2.8288188686198636,
+      "grad_norm": 1.9829123806277538,
+      "learning_rate": 2.825661336991998e-05,
+      "loss": 0.936,
+      "step": 1919
+    },
+    {
+      "epoch": 2.830292979546711,
+      "grad_norm": 2.0137752354731964,
+      "learning_rate": 2.8236987704082417e-05,
+      "loss": 0.9222,
+      "step": 1920
+    },
+    {
+      "epoch": 2.831767090473558,
+      "grad_norm": 1.9893906052265287,
+      "learning_rate": 2.8217360009172922e-05,
+      "loss": 0.8904,
+      "step": 1921
+    },
+    {
+      "epoch": 2.8332412014004054,
+      "grad_norm": 2.0695718844072566,
+      "learning_rate": 2.8197730297494896e-05,
+      "loss": 0.9024,
+      "step": 1922
+    },
+    {
+      "epoch": 2.8347153123272526,
+      "grad_norm": 2.0782482139371394,
+      "learning_rate": 2.8178098581353018e-05,
+      "loss": 0.8734,
+      "step": 1923
+    },
+    {
+      "epoch": 2.8361894232541,
+      "grad_norm": 2.074784196409333,
+      "learning_rate": 2.8158464873053237e-05,
+      "loss": 0.9932,
+      "step": 1924
+    },
+    {
+      "epoch": 2.837663534180947,
+      "grad_norm": 1.9607349306320534,
+      "learning_rate": 2.8138829184902727e-05,
+      "loss": 0.9758,
+      "step": 1925
+    },
+    {
+      "epoch": 2.8391376451077943,
+      "grad_norm": 2.1966187198752976,
+      "learning_rate": 2.811919152920991e-05,
+      "loss": 0.9256,
+      "step": 1926
+    },
+    {
+      "epoch": 2.8406117560346416,
+      "grad_norm": 2.149744142788989,
+      "learning_rate": 2.8099551918284468e-05,
+      "loss": 0.9008,
+      "step": 1927
+    },
+    {
+      "epoch": 2.842085866961489,
+      "grad_norm": 2.1642825018262246,
+      "learning_rate": 2.8079910364437263e-05,
+      "loss": 1.0446,
+      "step": 1928
+    },
+    {
+      "epoch": 2.843559977888336,
+      "grad_norm": 2.1750647085802366,
+      "learning_rate": 2.8060266879980408e-05,
+      "loss": 1.0718,
+      "step": 1929
+    },
+    {
+      "epoch": 2.8450340888151833,
+      "grad_norm": 2.0072262869301407,
+      "learning_rate": 2.8040621477227214e-05,
+      "loss": 1.0321,
+      "step": 1930
+    },
+    {
+      "epoch": 2.8465081997420305,
+      "grad_norm": 1.954013530801488,
+      "learning_rate": 2.8020974168492197e-05,
+      "loss": 0.9447,
+      "step": 1931
+    },
+    {
+      "epoch": 2.8479823106688777,
+      "grad_norm": 1.9962274614852062,
+      "learning_rate": 2.8001324966091076e-05,
+      "loss": 0.7935,
+      "step": 1932
+    },
+    {
+      "epoch": 2.849456421595725,
+      "grad_norm": 2.3085786718918557,
+      "learning_rate": 2.7981673882340726e-05,
+      "loss": 1.0391,
+      "step": 1933
+    },
+    {
+      "epoch": 2.8509305325225722,
+      "grad_norm": 2.093838590199608,
+      "learning_rate": 2.796202092955924e-05,
+      "loss": 0.8771,
+      "step": 1934
+    },
+    {
+      "epoch": 2.8524046434494195,
+      "grad_norm": 2.226016007666598,
+      "learning_rate": 2.7942366120065872e-05,
+      "loss": 0.8799,
+      "step": 1935
+    },
+    {
+      "epoch": 2.8538787543762667,
+      "grad_norm": 2.0024517052550013,
+      "learning_rate": 2.792270946618102e-05,
+      "loss": 0.8739,
+      "step": 1936
+    },
+    {
+      "epoch": 2.855352865303114,
+      "grad_norm": 2.22901358628443,
+      "learning_rate": 2.790305098022626e-05,
+      "loss": 0.997,
+      "step": 1937
+    },
+    {
+      "epoch": 2.856826976229961,
+      "grad_norm": 2.372418757696739,
+      "learning_rate": 2.78833906745243e-05,
+      "loss": 0.9276,
+      "step": 1938
+    },
+    {
+      "epoch": 2.8583010871568084,
+      "grad_norm": 2.2132512040392966,
+      "learning_rate": 2.7863728561399016e-05,
+      "loss": 0.9028,
+      "step": 1939
+    },
+    {
+      "epoch": 2.8597751980836557,
+      "grad_norm": 2.0593539432265318,
+      "learning_rate": 2.7844064653175378e-05,
+      "loss": 0.8346,
+      "step": 1940
+    },
+    {
+      "epoch": 2.861249309010503,
+      "grad_norm": 2.1577616382435725,
+      "learning_rate": 2.7824398962179503e-05,
+      "loss": 1.0189,
+      "step": 1941
+    },
+    {
+      "epoch": 2.86272341993735,
+      "grad_norm": 2.0930959334680383,
+      "learning_rate": 2.780473150073864e-05,
+      "loss": 0.9341,
+      "step": 1942
+    },
+    {
+      "epoch": 2.8641975308641974,
+      "grad_norm": 2.1796446218340813,
+      "learning_rate": 2.7785062281181124e-05,
+      "loss": 1.0079,
+      "step": 1943
+    },
+    {
+      "epoch": 2.8656716417910446,
+      "grad_norm": 2.2649931980393023,
+      "learning_rate": 2.7765391315836396e-05,
+      "loss": 1.0087,
+      "step": 1944
+    },
+    {
+      "epoch": 2.8671457527178923,
+      "grad_norm": 2.2403913474524915,
+      "learning_rate": 2.7745718617034998e-05,
+      "loss": 0.8887,
+      "step": 1945
+    },
+    {
+      "epoch": 2.8686198636447395,
+      "grad_norm": 2.066933577232033,
+      "learning_rate": 2.7726044197108557e-05,
+      "loss": 1.0993,
+      "step": 1946
+    },
+    {
+      "epoch": 2.8700939745715868,
+      "grad_norm": 2.176907095703998,
+      "learning_rate": 2.7706368068389778e-05,
+      "loss": 0.7714,
+      "step": 1947
+    },
+    {
+      "epoch": 2.871568085498434,
+      "grad_norm": 2.324074312738905,
+      "learning_rate": 2.7686690243212432e-05,
+      "loss": 1.1721,
+      "step": 1948
+    },
+    {
+      "epoch": 2.8730421964252812,
+      "grad_norm": 2.0571385837692198,
+      "learning_rate": 2.7667010733911354e-05,
+      "loss": 0.965,
+      "step": 1949
+    },
+    {
+      "epoch": 2.8745163073521285,
+      "grad_norm": 2.135276628570073,
+      "learning_rate": 2.7647329552822455e-05,
+      "loss": 0.9041,
+      "step": 1950
+    },
+    {
+      "epoch": 2.8759904182789757,
+      "grad_norm": 2.164665936780751,
+      "learning_rate": 2.762764671228267e-05,
+      "loss": 0.8907,
+      "step": 1951
+    },
+    {
+      "epoch": 2.877464529205823,
+      "grad_norm": 2.239392815108347,
+      "learning_rate": 2.760796222462998e-05,
+      "loss": 0.9501,
+      "step": 1952
+    },
+    {
+      "epoch": 2.87893864013267,
+      "grad_norm": 2.0317727223296007,
+      "learning_rate": 2.7588276102203398e-05,
+      "loss": 0.8825,
+      "step": 1953
+    },
+    {
+      "epoch": 2.8804127510595174,
+      "grad_norm": 2.1517928122791745,
+      "learning_rate": 2.7568588357342973e-05,
+      "loss": 0.9281,
+      "step": 1954
+    },
+    {
+      "epoch": 2.8818868619863647,
+      "grad_norm": 2.173488319620076,
+      "learning_rate": 2.754889900238975e-05,
+      "loss": 0.9885,
+      "step": 1955
+    },
+    {
+      "epoch": 2.883360972913212,
+      "grad_norm": 2.3318064212248997,
+      "learning_rate": 2.7529208049685807e-05,
+      "loss": 1.049,
+      "step": 1956
+    },
+    {
+      "epoch": 2.884835083840059,
+      "grad_norm": 2.189165843454745,
+      "learning_rate": 2.7509515511574208e-05,
+      "loss": 0.9454,
+      "step": 1957
+    },
+    {
+      "epoch": 2.8863091947669064,
+      "grad_norm": 2.0281886580106177,
+      "learning_rate": 2.748982140039902e-05,
+      "loss": 0.896,
+      "step": 1958
+    },
+    {
+      "epoch": 2.8877833056937536,
+      "grad_norm": 2.0649616694183437,
+      "learning_rate": 2.747012572850528e-05,
+      "loss": 0.8909,
+      "step": 1959
+    },
+    {
+      "epoch": 2.889257416620601,
+      "grad_norm": 1.9376862959358307,
+      "learning_rate": 2.7450428508239024e-05,
+      "loss": 0.9517,
+      "step": 1960
+    },
+    {
+      "epoch": 2.890731527547448,
+      "grad_norm": 2.024468472269905,
+      "learning_rate": 2.743072975194723e-05,
+      "loss": 0.9304,
+      "step": 1961
+    },
+    {
+      "epoch": 2.8922056384742953,
+      "grad_norm": 2.197833693928408,
+      "learning_rate": 2.741102947197789e-05,
+      "loss": 0.9326,
+      "step": 1962
+    },
+    {
+      "epoch": 2.8936797494011426,
+      "grad_norm": 2.205512842757133,
+      "learning_rate": 2.7391327680679895e-05,
+      "loss": 0.9714,
+      "step": 1963
+    },
+    {
+      "epoch": 2.89515386032799,
+      "grad_norm": 2.124966452838655,
+      "learning_rate": 2.7371624390403116e-05,
+      "loss": 1.0,
+      "step": 1964
+    },
+    {
+      "epoch": 2.896627971254837,
+      "grad_norm": 2.164118686869307,
+      "learning_rate": 2.735191961349835e-05,
+      "loss": 0.9994,
+      "step": 1965
+    },
+    {
+      "epoch": 2.8981020821816843,
+      "grad_norm": 2.190263691708167,
+      "learning_rate": 2.7332213362317328e-05,
+      "loss": 0.9316,
+      "step": 1966
+    },
+    {
+      "epoch": 2.8995761931085315,
+      "grad_norm": 1.948747535885078,
+      "learning_rate": 2.7312505649212722e-05,
+      "loss": 0.8285,
+      "step": 1967
+    },
+    {
+      "epoch": 2.901050304035379,
+      "grad_norm": 2.1955565184596666,
+      "learning_rate": 2.7292796486538093e-05,
+      "loss": 1.0584,
+      "step": 1968
+    },
+    {
+      "epoch": 2.902524414962226,
+      "grad_norm": 1.936813048128675,
+      "learning_rate": 2.727308588664793e-05,
+      "loss": 0.7911,
+      "step": 1969
+    },
+    {
+      "epoch": 2.9039985258890733,
+      "grad_norm": 2.351214513029234,
+      "learning_rate": 2.725337386189761e-05,
+      "loss": 1.0626,
+      "step": 1970
+    },
+    {
+      "epoch": 2.9054726368159205,
+      "grad_norm": 2.072243764979541,
+      "learning_rate": 2.723366042464342e-05,
+      "loss": 0.9979,
+      "step": 1971
+    },
+    {
+      "epoch": 2.9069467477427677,
+      "grad_norm": 2.144502010736748,
+      "learning_rate": 2.7213945587242508e-05,
+      "loss": 1.1117,
+      "step": 1972
+    },
+    {
+      "epoch": 2.908420858669615,
+      "grad_norm": 2.3090684535013946,
+      "learning_rate": 2.7194229362052924e-05,
+      "loss": 0.9736,
+      "step": 1973
+    },
+    {
+      "epoch": 2.909894969596462,
+      "grad_norm": 2.0226378757177077,
+      "learning_rate": 2.7174511761433585e-05,
+      "loss": 1.1498,
+      "step": 1974
+    },
+    {
+      "epoch": 2.9113690805233094,
+      "grad_norm": 2.114863284767324,
+      "learning_rate": 2.715479279774425e-05,
+      "loss": 0.9978,
+      "step": 1975
+    },
+    {
+      "epoch": 2.9128431914501567,
+      "grad_norm": 2.0441151424010013,
+      "learning_rate": 2.7135072483345552e-05,
+      "loss": 0.8907,
+      "step": 1976
+    },
+    {
+      "epoch": 2.914317302377004,
+      "grad_norm": 2.159377663109734,
+      "learning_rate": 2.7115350830598958e-05,
+      "loss": 0.948,
+      "step": 1977
+    },
+    {
+      "epoch": 2.915791413303851,
+      "grad_norm": 2.074035751402493,
+      "learning_rate": 2.709562785186679e-05,
+      "loss": 0.9448,
+      "step": 1978
+    },
+    {
+      "epoch": 2.9172655242306984,
+      "grad_norm": 2.4391524753433584,
+      "learning_rate": 2.7075903559512178e-05,
+      "loss": 0.9926,
+      "step": 1979
+    },
+    {
+      "epoch": 2.9187396351575456,
+      "grad_norm": 2.25316259852769,
+      "learning_rate": 2.7056177965899097e-05,
+      "loss": 0.9782,
+      "step": 1980
+    },
+    {
+      "epoch": 2.920213746084393,
+      "grad_norm": 2.237264512099727,
+      "learning_rate": 2.7036451083392332e-05,
+      "loss": 0.9497,
+      "step": 1981
+    },
+    {
+      "epoch": 2.92168785701124,
+      "grad_norm": 2.232577367725665,
+      "learning_rate": 2.701672292435747e-05,
+      "loss": 0.9958,
+      "step": 1982
+    },
+    {
+      "epoch": 2.9231619679380874,
+      "grad_norm": 2.2715650890176575,
+      "learning_rate": 2.69969935011609e-05,
+      "loss": 1.0359,
+      "step": 1983
+    },
+    {
+      "epoch": 2.9246360788649346,
+      "grad_norm": 2.1787683880326028,
+      "learning_rate": 2.6977262826169807e-05,
+      "loss": 1.0161,
+      "step": 1984
+    },
+    {
+      "epoch": 2.926110189791782,
+      "grad_norm": 2.195982155531071,
+      "learning_rate": 2.695753091175216e-05,
+      "loss": 0.9379,
+      "step": 1985
+    },
+    {
+      "epoch": 2.927584300718629,
+      "grad_norm": 2.138687593585216,
+      "learning_rate": 2.6937797770276702e-05,
+      "loss": 1.0092,
+      "step": 1986
+    },
+    {
+      "epoch": 2.9290584116454763,
+      "grad_norm": 2.00690222860402,
+      "learning_rate": 2.6918063414112942e-05,
+      "loss": 0.8235,
+      "step": 1987
+    },
+    {
+      "epoch": 2.9305325225723236,
+      "grad_norm": 2.1371470689804912,
+      "learning_rate": 2.6898327855631155e-05,
+      "loss": 1.0929,
+      "step": 1988
+    },
+    {
+      "epoch": 2.932006633499171,
+      "grad_norm": 2.004136100213434,
+      "learning_rate": 2.6878591107202383e-05,
+      "loss": 0.9106,
+      "step": 1989
+    },
+    {
+      "epoch": 2.933480744426018,
+      "grad_norm": 2.084505539771222,
+      "learning_rate": 2.685885318119839e-05,
+      "loss": 0.9831,
+      "step": 1990
+    },
+    {
+      "epoch": 2.9349548553528653,
+      "grad_norm": 2.2400347114325823,
+      "learning_rate": 2.683911408999169e-05,
+      "loss": 0.9796,
+      "step": 1991
+    },
+    {
+      "epoch": 2.9364289662797125,
+      "grad_norm": 2.0879904256579245,
+      "learning_rate": 2.6819373845955527e-05,
+      "loss": 0.9937,
+      "step": 1992
+    },
+    {
+      "epoch": 2.9379030772065597,
+      "grad_norm": 2.097440172144683,
+      "learning_rate": 2.6799632461463862e-05,
+      "loss": 1.0235,
+      "step": 1993
+    },
+    {
+      "epoch": 2.939377188133407,
+      "grad_norm": 2.122542924211053,
+      "learning_rate": 2.6779889948891384e-05,
+      "loss": 1.132,
+      "step": 1994
+    },
+    {
+      "epoch": 2.940851299060254,
+      "grad_norm": 2.0540627809864143,
+      "learning_rate": 2.676014632061347e-05,
+      "loss": 0.8272,
+      "step": 1995
+    },
+    {
+      "epoch": 2.9423254099871015,
+      "grad_norm": 2.1484950942209196,
+      "learning_rate": 2.674040158900622e-05,
+      "loss": 0.8859,
+      "step": 1996
+    },
+    {
+      "epoch": 2.9437995209139487,
+      "grad_norm": 2.2527703078784116,
+      "learning_rate": 2.6720655766446412e-05,
+      "loss": 1.0119,
+      "step": 1997
+    },
+    {
+      "epoch": 2.945273631840796,
+      "grad_norm": 2.24887438274793,
+      "learning_rate": 2.6700908865311497e-05,
+      "loss": 1.1303,
+      "step": 1998
+    },
+    {
+      "epoch": 2.946747742767643,
+      "grad_norm": 1.9380307854883236,
+      "learning_rate": 2.6681160897979623e-05,
+      "loss": 0.8704,
+      "step": 1999
+    },
+    {
+      "epoch": 2.9482218536944904,
+      "grad_norm": 2.116668524954073,
+      "learning_rate": 2.6661411876829596e-05,
+      "loss": 0.9759,
+      "step": 2000
+    },
+    {
+      "epoch": 2.9482218536944904,
+      "eval_bleu": 0.04787357154563729,
+      "eval_bleu_1gram": 0.3421104118417729,
+      "eval_bleu_2gram": 0.11816282651259664,
+      "eval_bleu_3gram": 0.04741820328686351,
+      "eval_bleu_4gram": 0.022146725466342863,
+      "eval_rag_val_loss": 1.6265630154840407,
+      "eval_rouge1": 0.33128632764598065,
+      "eval_rouge2": 0.11283687684156013,
+      "eval_rougeL": 0.31066768988855775,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4068,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": true,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}