diff --git "a/checkpoint-1274/trainer_state.json" "b/checkpoint-1274/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1274/trainer_state.json"
@@ -0,0 +1,9071 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.4606934431857193,
+  "eval_steps": 91,
+  "global_step": 1274,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0027463096464126332,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 3e-06,
+      "loss": 0.6232,
+      "step": 1
+    },
+    {
+      "epoch": 0.0027463096464126332,
+      "eval_loss": 0.6296440362930298,
+      "eval_runtime": 599.6568,
+      "eval_samples_per_second": 15.285,
+      "eval_steps_per_second": 15.285,
+      "step": 1
+    },
+    {
+      "epoch": 0.0054926192928252664,
+      "grad_norm": 0.060546875,
+      "learning_rate": 6e-06,
+      "loss": 0.5596,
+      "step": 2
+    },
+    {
+      "epoch": 0.008238928939237899,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9e-06,
+      "loss": 0.7123,
+      "step": 3
+    },
+    {
+      "epoch": 0.010985238585650533,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6337,
+      "step": 4
+    },
+    {
+      "epoch": 0.013731548232063165,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.5e-05,
+      "loss": 0.5764,
+      "step": 5
+    },
+    {
+      "epoch": 0.016477857878475798,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.8e-05,
+      "loss": 0.7453,
+      "step": 6
+    },
+    {
+      "epoch": 0.01922416752488843,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 2.1e-05,
+      "loss": 0.7076,
+      "step": 7
+    },
+    {
+      "epoch": 0.021970477171301066,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 2.4e-05,
+      "loss": 0.5094,
+      "step": 8
+    },
+    {
+      "epoch": 0.024716786817713696,
+      "grad_norm": 0.05859375,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.6062,
+      "step": 9
+    },
+    {
+      "epoch": 0.02746309646412633,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 3e-05,
+      "loss": 0.6501,
+      "step": 10
+    },
+    {
+      "epoch": 0.030209406110538965,
+      "grad_norm": 0.060546875,
+      "learning_rate": 2.9999964598289033e-05,
+      "loss": 0.6403,
+      "step": 11
+    },
+    {
+      "epoch": 0.032955715756951595,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 2.999985839332323e-05,
+      "loss": 0.6464,
+      "step": 12
+    },
+    {
+      "epoch": 0.03570202540336423,
+      "grad_norm": 0.05859375,
+      "learning_rate": 2.9999681385603907e-05,
+      "loss": 0.6528,
+      "step": 13
+    },
+    {
+      "epoch": 0.03844833504977686,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 2.9999433575966585e-05,
+      "loss": 0.7109,
+      "step": 14
+    },
+    {
+      "epoch": 0.0411946446961895,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 2.999911496558097e-05,
+      "loss": 0.4795,
+      "step": 15
+    },
+    {
+      "epoch": 0.04394095434260213,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 2.9998725555950983e-05,
+      "loss": 0.6743,
+      "step": 16
+    },
+    {
+      "epoch": 0.04668726398901476,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 2.9998265348914726e-05,
+      "loss": 0.6143,
+      "step": 17
+    },
+    {
+      "epoch": 0.04943357363542739,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 2.9997734346644482e-05,
+      "loss": 0.5521,
+      "step": 18
+    },
+    {
+      "epoch": 0.05217988328184003,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 2.99971325516467e-05,
+      "loss": 0.6291,
+      "step": 19
+    },
+    {
+      "epoch": 0.05492619292825266,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 2.9996459966761994e-05,
+      "loss": 0.5828,
+      "step": 20
+    },
+    {
+      "epoch": 0.057672502574665295,
+      "grad_norm": 0.068359375,
+      "learning_rate": 2.9995716595165114e-05,
+      "loss": 0.5432,
+      "step": 21
+    },
+    {
+      "epoch": 0.06041881222107793,
+      "grad_norm": 0.076171875,
+      "learning_rate": 2.9994902440364943e-05,
+      "loss": 0.5769,
+      "step": 22
+    },
+    {
+      "epoch": 0.06316512186749056,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 2.999401750620448e-05,
+      "loss": 0.469,
+      "step": 23
+    },
+    {
+      "epoch": 0.06591143151390319,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 2.999306179686082e-05,
+      "loss": 0.4414,
+      "step": 24
+    },
+    {
+      "epoch": 0.06865774116031582,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.9992035316845125e-05,
+      "loss": 0.7219,
+      "step": 25
+    },
+    {
+      "epoch": 0.07140405080672846,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 2.9990938071002606e-05,
+      "loss": 0.6742,
+      "step": 26
+    },
+    {
+      "epoch": 0.07415036045314109,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.998977006451253e-05,
+      "loss": 0.5633,
+      "step": 27
+    },
+    {
+      "epoch": 0.07689667009955373,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.998853130288814e-05,
+      "loss": 0.5486,
+      "step": 28
+    },
+    {
+      "epoch": 0.07964297974596636,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.9987221791976687e-05,
+      "loss": 0.4064,
+      "step": 29
+    },
+    {
+      "epoch": 0.082389289392379,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.9985841537959345e-05,
+      "loss": 0.5184,
+      "step": 30
+    },
+    {
+      "epoch": 0.08513559903879163,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.9984390547351244e-05,
+      "loss": 0.5407,
+      "step": 31
+    },
+    {
+      "epoch": 0.08788190868520426,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.998286882700138e-05,
+      "loss": 0.5532,
+      "step": 32
+    },
+    {
+      "epoch": 0.09062821833161688,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.9981276384092628e-05,
+      "loss": 1.2318,
+      "step": 33
+    },
+    {
+      "epoch": 0.09337452797802952,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.9979613226141672e-05,
+      "loss": 0.5457,
+      "step": 34
+    },
+    {
+      "epoch": 0.09612083762444215,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.9977879360999007e-05,
+      "loss": 0.5391,
+      "step": 35
+    },
+    {
+      "epoch": 0.09886714727085479,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.9976074796848866e-05,
+      "loss": 0.5919,
+      "step": 36
+    },
+    {
+      "epoch": 0.10161345691726742,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 2.99741995422092e-05,
+      "loss": 0.6211,
+      "step": 37
+    },
+    {
+      "epoch": 0.10435976656368005,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 2.997225360593165e-05,
+      "loss": 0.5296,
+      "step": 38
+    },
+    {
+      "epoch": 0.10710607621009269,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.997023699720147e-05,
+      "loss": 0.5469,
+      "step": 39
+    },
+    {
+      "epoch": 0.10985238585650532,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.9968149725537515e-05,
+      "loss": 0.6447,
+      "step": 40
+    },
+    {
+      "epoch": 0.11259869550291796,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9965991800792185e-05,
+      "loss": 0.5941,
+      "step": 41
+    },
+    {
+      "epoch": 0.11534500514933059,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.9963763233151377e-05,
+      "loss": 0.5228,
+      "step": 42
+    },
+    {
+      "epoch": 0.11809131479574322,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9961464033134444e-05,
+      "loss": 0.6166,
+      "step": 43
+    },
+    {
+      "epoch": 0.12083762444215586,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.9959094211594122e-05,
+      "loss": 0.6156,
+      "step": 44
+    },
+    {
+      "epoch": 0.12358393408856849,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 2.9956653779716517e-05,
+      "loss": 0.5492,
+      "step": 45
+    },
+    {
+      "epoch": 0.1263302437349811,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9954142749021023e-05,
+      "loss": 0.5405,
+      "step": 46
+    },
+    {
+      "epoch": 0.12907655338139376,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.9951561131360278e-05,
+      "loss": 0.584,
+      "step": 47
+    },
+    {
+      "epoch": 0.13182286302780638,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.994890893892011e-05,
+      "loss": 0.62,
+      "step": 48
+    },
+    {
+      "epoch": 0.13456917267421903,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 2.994618618421946e-05,
+      "loss": 0.4026,
+      "step": 49
+    },
+    {
+      "epoch": 0.13731548232063165,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.994339288011037e-05,
+      "loss": 0.6126,
+      "step": 50
+    },
+    {
+      "epoch": 0.1400617919670443,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.9940529039777855e-05,
+      "loss": 0.6316,
+      "step": 51
+    },
+    {
+      "epoch": 0.14280810161345692,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 2.9937594676739907e-05,
+      "loss": 0.3887,
+      "step": 52
+    },
+    {
+      "epoch": 0.14555441125986954,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9934589804847382e-05,
+      "loss": 0.5214,
+      "step": 53
+    },
+    {
+      "epoch": 0.14830072090628219,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 2.9931514438283966e-05,
+      "loss": 0.5914,
+      "step": 54
+    },
+    {
+      "epoch": 0.1510470305526948,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9928368591566085e-05,
+      "loss": 0.5443,
+      "step": 55
+    },
+    {
+      "epoch": 0.15379334019910745,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.9925152279542856e-05,
+      "loss": 0.5522,
+      "step": 56
+    },
+    {
+      "epoch": 0.15653964984552007,
+      "grad_norm": 0.031494140625,
+      "learning_rate": 2.9921865517396008e-05,
+      "loss": 0.5146,
+      "step": 57
+    },
+    {
+      "epoch": 0.15928595949193272,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 2.9918508320639803e-05,
+      "loss": 0.5396,
+      "step": 58
+    },
+    {
+      "epoch": 0.16203226913834534,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9915080705120976e-05,
+      "loss": 0.5118,
+      "step": 59
+    },
+    {
+      "epoch": 0.164778578784758,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.991158268701866e-05,
+      "loss": 0.6652,
+      "step": 60
+    },
+    {
+      "epoch": 0.1675248884311706,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.9908014282844295e-05,
+      "loss": 0.4211,
+      "step": 61
+    },
+    {
+      "epoch": 0.17027119807758326,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.9904375509441562e-05,
+      "loss": 0.4445,
+      "step": 62
+    },
+    {
+      "epoch": 0.17301750772399588,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.9900666383986303e-05,
+      "loss": 0.588,
+      "step": 63
+    },
+    {
+      "epoch": 0.17576381737040853,
+      "grad_norm": 0.033203125,
+      "learning_rate": 2.9896886923986433e-05,
+      "loss": 0.5705,
+      "step": 64
+    },
+    {
+      "epoch": 0.17851012701682115,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.989303714728187e-05,
+      "loss": 0.5068,
+      "step": 65
+    },
+    {
+      "epoch": 0.18125643666323377,
+      "grad_norm": 0.033203125,
+      "learning_rate": 2.9889117072044436e-05,
+      "loss": 0.5196,
+      "step": 66
+    },
+    {
+      "epoch": 0.18400274630964641,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.9885126716777776e-05,
+      "loss": 0.5952,
+      "step": 67
+    },
+    {
+      "epoch": 0.18674905595605903,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 2.9881066100317288e-05,
+      "loss": 0.6194,
+      "step": 68
+    },
+    {
+      "epoch": 0.18949536560247168,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.987693524183e-05,
+      "loss": 0.4453,
+      "step": 69
+    },
+    {
+      "epoch": 0.1922416752488843,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.987273416081451e-05,
+      "loss": 0.524,
+      "step": 70
+    },
+    {
+      "epoch": 0.19498798489529695,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.9868462877100875e-05,
+      "loss": 0.5899,
+      "step": 71
+    },
+    {
+      "epoch": 0.19773429454170957,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9864121410850527e-05,
+      "loss": 0.4603,
+      "step": 72
+    },
+    {
+      "epoch": 0.20048060418812222,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.9859709782556185e-05,
+      "loss": 0.4829,
+      "step": 73
+    },
+    {
+      "epoch": 0.20322691383453484,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 2.9855228013041737e-05,
+      "loss": 0.5735,
+      "step": 74
+    },
+    {
+      "epoch": 0.2059732234809475,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.9850676123462157e-05,
+      "loss": 0.5104,
+      "step": 75
+    },
+    {
+      "epoch": 0.2087195331273601,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.98460541353034e-05,
+      "loss": 0.5501,
+      "step": 76
+    },
+    {
+      "epoch": 0.21146584277377276,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 2.9841362070382307e-05,
+      "loss": 0.5119,
+      "step": 77
+    },
+    {
+      "epoch": 0.21421215242018538,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.9836599950846493e-05,
+      "loss": 0.589,
+      "step": 78
+    },
+    {
+      "epoch": 0.216958462066598,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 2.9831767799174255e-05,
+      "loss": 0.4544,
+      "step": 79
+    },
+    {
+      "epoch": 0.21970477171301064,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9826865638174445e-05,
+      "loss": 0.4294,
+      "step": 80
+    },
+    {
+      "epoch": 0.22245108135942326,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.9821893490986382e-05,
+      "loss": 0.5649,
+      "step": 81
+    },
+    {
+      "epoch": 0.2251973910058359,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.981685138107974e-05,
+      "loss": 0.532,
+      "step": 82
+    },
+    {
+      "epoch": 0.22794370065224853,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.9811739332254418e-05,
+      "loss": 0.6026,
+      "step": 83
+    },
+    {
+      "epoch": 0.23069001029866118,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.9806557368640457e-05,
+      "loss": 0.5516,
+      "step": 84
+    },
+    {
+      "epoch": 0.2334363199450738,
+      "grad_norm": 0.033203125,
+      "learning_rate": 2.9801305514697913e-05,
+      "loss": 0.4544,
+      "step": 85
+    },
+    {
+      "epoch": 0.23618262959148645,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 2.9795983795216727e-05,
+      "loss": 0.5327,
+      "step": 86
+    },
+    {
+      "epoch": 0.23892893923789907,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.979059223531664e-05,
+      "loss": 0.5217,
+      "step": 87
+    },
+    {
+      "epoch": 0.24167524888431172,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.978513086044703e-05,
+      "loss": 0.4562,
+      "step": 88
+    },
+    {
+      "epoch": 0.24442155853072434,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.9779599696386846e-05,
+      "loss": 0.763,
+      "step": 89
+    },
+    {
+      "epoch": 0.24716786817713698,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 2.9773998769244434e-05,
+      "loss": 0.4698,
+      "step": 90
+    },
+    {
+      "epoch": 0.2499141778235496,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.976832810545745e-05,
+      "loss": 0.5602,
+      "step": 91
+    },
+    {
+      "epoch": 0.2499141778235496,
+      "eval_loss": 0.5245938897132874,
+      "eval_runtime": 620.8292,
+      "eval_samples_per_second": 14.764,
+      "eval_steps_per_second": 14.764,
+      "step": 91
+    },
+    {
+      "epoch": 0.2526604874699622,
+      "grad_norm": 0.030517578125,
+      "learning_rate": 2.9762587731792725e-05,
+      "loss": 0.477,
+      "step": 92
+    },
+    {
+      "epoch": 0.2554067971163749,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 2.9756777675346128e-05,
+      "loss": 0.5536,
+      "step": 93
+    },
+    {
+      "epoch": 0.2581531067627875,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9750897963542453e-05,
+      "loss": 0.5581,
+      "step": 94
+    },
+    {
+      "epoch": 0.2608994164092001,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.974494862413528e-05,
+      "loss": 0.5737,
+      "step": 95
+    },
+    {
+      "epoch": 0.26364572605561276,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.973892968520685e-05,
+      "loss": 0.5191,
+      "step": 96
+    },
+    {
+      "epoch": 0.2663920357020254,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9732841175167924e-05,
+      "loss": 0.6794,
+      "step": 97
+    },
+    {
+      "epoch": 0.26913834534843806,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9726683122757664e-05,
+      "loss": 0.5615,
+      "step": 98
+    },
+    {
+      "epoch": 0.27188465499485065,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.972045555704348e-05,
+      "loss": 0.521,
+      "step": 99
+    },
+    {
+      "epoch": 0.2746309646412633,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.97141585074209e-05,
+      "loss": 0.4473,
+      "step": 100
+    },
+    {
+      "epoch": 0.27737727428767595,
+      "grad_norm": 0.033203125,
+      "learning_rate": 2.9707792003613434e-05,
+      "loss": 0.6017,
+      "step": 101
+    },
+    {
+      "epoch": 0.2801235839340886,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.9701356075672442e-05,
+      "loss": 0.5079,
+      "step": 102
+    },
+    {
+      "epoch": 0.2828698935805012,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.969485075397696e-05,
+      "loss": 0.5738,
+      "step": 103
+    },
+    {
+      "epoch": 0.28561620322691383,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 2.9688276069233596e-05,
+      "loss": 0.4251,
+      "step": 104
+    },
+    {
+      "epoch": 0.2883625128733265,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.968163205247636e-05,
+      "loss": 0.5902,
+      "step": 105
+    },
+    {
+      "epoch": 0.2911088225197391,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.9674918735066534e-05,
+      "loss": 0.4307,
+      "step": 106
+    },
+    {
+      "epoch": 0.2938551321661517,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.9668136148692497e-05,
+      "loss": 0.4018,
+      "step": 107
+    },
+    {
+      "epoch": 0.29660144181256437,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.966128432536961e-05,
+      "loss": 0.5109,
+      "step": 108
+    },
+    {
+      "epoch": 0.299347751458977,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.9654363297440045e-05,
+      "loss": 0.6136,
+      "step": 109
+    },
+    {
+      "epoch": 0.3020940611053896,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.964737309757262e-05,
+      "loss": 0.4161,
+      "step": 110
+    },
+    {
+      "epoch": 0.30484037075180226,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.9640313758762692e-05,
+      "loss": 0.4268,
+      "step": 111
+    },
+    {
+      "epoch": 0.3075866803982149,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.9633185314331933e-05,
+      "loss": 0.4809,
+      "step": 112
+    },
+    {
+      "epoch": 0.31033299004462755,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.9625987797928237e-05,
+      "loss": 0.4976,
+      "step": 113
+    },
+    {
+      "epoch": 0.31307929969104015,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.9618721243525522e-05,
+      "loss": 0.5508,
+      "step": 114
+    },
+    {
+      "epoch": 0.3158256093374528,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.9611385685423582e-05,
+      "loss": 0.4852,
+      "step": 115
+    },
+    {
+      "epoch": 0.31857191898386544,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.9603981158247918e-05,
+      "loss": 0.4301,
+      "step": 116
+    },
+    {
+      "epoch": 0.3213182286302781,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.9596507696949598e-05,
+      "loss": 0.4456,
+      "step": 117
+    },
+    {
+      "epoch": 0.3240645382766907,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.9588965336805065e-05,
+      "loss": 0.6092,
+      "step": 118
+    },
+    {
+      "epoch": 0.32681084792310333,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 2.958135411341597e-05,
+      "loss": 0.4823,
+      "step": 119
+    },
+    {
+      "epoch": 0.329557157569516,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.9573674062709024e-05,
+      "loss": 0.4666,
+      "step": 120
+    },
+    {
+      "epoch": 0.33230346721592857,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9565925220935828e-05,
+      "loss": 0.4868,
+      "step": 121
+    },
+    {
+      "epoch": 0.3350497768623412,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 2.9558107624672673e-05,
+      "loss": 0.529,
+      "step": 122
+    },
+    {
+      "epoch": 0.33779608650875387,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.9550221310820405e-05,
+      "loss": 0.4308,
+      "step": 123
+    },
+    {
+      "epoch": 0.3405423961551665,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 2.9542266316604213e-05,
+      "loss": 0.4058,
+      "step": 124
+    },
+    {
+      "epoch": 0.3432887058015791,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.95342426795735e-05,
+      "loss": 0.4788,
+      "step": 125
+    },
+    {
+      "epoch": 0.34603501544799176,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.952615043760165e-05,
+      "loss": 1.2963,
+      "step": 126
+    },
+    {
+      "epoch": 0.3487813250944044,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.95179896288859e-05,
+      "loss": 0.5734,
+      "step": 127
+    },
+    {
+      "epoch": 0.35152763474081705,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.9509760291947128e-05,
+      "loss": 0.4352,
+      "step": 128
+    },
+    {
+      "epoch": 0.35427394438722964,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9501462465629672e-05,
+      "loss": 0.6082,
+      "step": 129
+    },
+    {
+      "epoch": 0.3570202540336423,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.949309618910118e-05,
+      "loss": 0.4699,
+      "step": 130
+    },
+    {
+      "epoch": 0.35976656368005494,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.9484661501852373e-05,
+      "loss": 0.5504,
+      "step": 131
+    },
+    {
+      "epoch": 0.36251287332646753,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 2.94761584436969e-05,
+      "loss": 0.545,
+      "step": 132
+    },
+    {
+      "epoch": 0.3652591829728802,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.9467587054771146e-05,
+      "loss": 0.445,
+      "step": 133
+    },
+    {
+      "epoch": 0.36800549261929283,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.945894737553401e-05,
+      "loss": 1.1891,
+      "step": 134
+    },
+    {
+      "epoch": 0.3707518022657055,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.945023944676676e-05,
+      "loss": 0.565,
+      "step": 135
+    },
+    {
+      "epoch": 0.37349811191211807,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.9441463309572797e-05,
+      "loss": 0.6599,
+      "step": 136
+    },
+    {
+      "epoch": 0.3762444215585307,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.9432619005377496e-05,
+      "loss": 0.4754,
+      "step": 137
+    },
+    {
+      "epoch": 0.37899073120494337,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9423706575927985e-05,
+      "loss": 0.4966,
+      "step": 138
+    },
+    {
+      "epoch": 0.381737040851356,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.9414726063292974e-05,
+      "loss": 0.4269,
+      "step": 139
+    },
+    {
+      "epoch": 0.3844833504977686,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.940567750986252e-05,
+      "loss": 0.5516,
+      "step": 140
+    },
+    {
+      "epoch": 0.38722966014418125,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.9396560958347865e-05,
+      "loss": 0.486,
+      "step": 141
+    },
+    {
+      "epoch": 0.3899759697905939,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 2.9387376451781215e-05,
+      "loss": 0.4506,
+      "step": 142
+    },
+    {
+      "epoch": 0.39272227943700655,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.9378124033515533e-05,
+      "loss": 0.6122,
+      "step": 143
+    },
+    {
+      "epoch": 0.39546858908341914,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.936880374722434e-05,
+      "loss": 0.4776,
+      "step": 144
+    },
+    {
+      "epoch": 0.3982148987298318,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.9359415636901522e-05,
+      "loss": 0.5574,
+      "step": 145
+    },
+    {
+      "epoch": 0.40096120837624444,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9349959746861093e-05,
+      "loss": 0.5289,
+      "step": 146
+    },
+    {
+      "epoch": 0.40370751802265703,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.9340436121737018e-05,
+      "loss": 0.4664,
+      "step": 147
+    },
+    {
+      "epoch": 0.4064538276690697,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.9330844806482974e-05,
+      "loss": 0.5322,
+      "step": 148
+    },
+    {
+      "epoch": 0.4092001373154823,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.9321185846372162e-05,
+      "loss": 0.4143,
+      "step": 149
+    },
+    {
+      "epoch": 0.411946446961895,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.9311459286997073e-05,
+      "loss": 0.5038,
+      "step": 150
+    },
+    {
+      "epoch": 0.41469275660830757,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.930166517426929e-05,
+      "loss": 0.4905,
+      "step": 151
+    },
+    {
+      "epoch": 0.4174390662547202,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.929180355441926e-05,
+      "loss": 0.5357,
+      "step": 152
+    },
+    {
+      "epoch": 0.42018537590113286,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.9281874473996077e-05,
+      "loss": 0.5449,
+      "step": 153
+    },
+    {
+      "epoch": 0.4229316855475455,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.9271877979867263e-05,
+      "loss": 1.3347,
+      "step": 154
+    },
+    {
+      "epoch": 0.4256779951939581,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.926181411921855e-05,
+      "loss": 0.4532,
+      "step": 155
+    },
+    {
+      "epoch": 0.42842430484037075,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.9251682939553662e-05,
+      "loss": 0.5425,
+      "step": 156
+    },
+    {
+      "epoch": 0.4311706144867834,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.9241484488694074e-05,
+      "loss": 0.4875,
+      "step": 157
+    },
+    {
+      "epoch": 0.433916924133196,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.92312188147788e-05,
+      "loss": 0.4574,
+      "step": 158
+    },
+    {
+      "epoch": 0.43666323377960864,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.9220885966264174e-05,
+      "loss": 0.5003,
+      "step": 159
+    },
+    {
+      "epoch": 0.4394095434260213,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.9210485991923577e-05,
+      "loss": 0.4766,
+      "step": 160
+    },
+    {
+      "epoch": 0.44215585307243394,
+      "grad_norm": 0.03515625,
+      "learning_rate": 2.9200018940847278e-05,
+      "loss": 0.3866,
+      "step": 161
+    },
+    {
+      "epoch": 0.44490216271884653,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.918948486244214e-05,
+      "loss": 0.4401,
+      "step": 162
+    },
+    {
+      "epoch": 0.4476484723652592,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.917888380643142e-05,
+      "loss": 0.5193,
+      "step": 163
+    },
+    {
+      "epoch": 0.4503947820116718,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.916821582285451e-05,
+      "loss": 0.4802,
+      "step": 164
+    },
+    {
+      "epoch": 0.45314109165808447,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 2.915748096206674e-05,
+      "loss": 0.4693,
+      "step": 165
+    },
+    {
+      "epoch": 0.45588740130449706,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.914667927473909e-05,
+      "loss": 0.4949,
+      "step": 166
+    },
+    {
+      "epoch": 0.4586337109509097,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9135810811857994e-05,
+      "loss": 0.5453,
+      "step": 167
+    },
+    {
+      "epoch": 0.46138002059732236,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.912487562472508e-05,
+      "loss": 0.4653,
+      "step": 168
+    },
+    {
+      "epoch": 0.464126330243735,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.9113873764956917e-05,
+      "loss": 0.5032,
+      "step": 169
+    },
+    {
+      "epoch": 0.4668726398901476,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.91028052844848e-05,
+      "loss": 0.4736,
+      "step": 170
+    },
+    {
+      "epoch": 0.46961894953656025,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.9091670235554478e-05,
+      "loss": 0.4773,
+      "step": 171
+    },
+    {
+      "epoch": 0.4723652591829729,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.9080468670725922e-05,
+      "loss": 0.5689,
+      "step": 172
+    },
+    {
+      "epoch": 0.4751115688293855,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.906920064287308e-05,
+      "loss": 0.51,
+      "step": 173
+    },
+    {
+      "epoch": 0.47785787847579814,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 2.9057866205183606e-05,
+      "loss": 0.446,
+      "step": 174
+    },
+    {
+      "epoch": 0.4806041881222108,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 2.9046465411158634e-05,
+      "loss": 0.4956,
+      "step": 175
+    },
+    {
+      "epoch": 0.48335049776862343,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.9034998314612516e-05,
+      "loss": 0.4963,
+      "step": 176
+    },
+    {
+      "epoch": 0.486096807415036,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.902346496967256e-05,
+      "loss": 0.4928,
+      "step": 177
+    },
+    {
+      "epoch": 0.4888431170614487,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.9011865430778782e-05,
+      "loss": 0.4731,
+      "step": 178
+    },
+    {
+      "epoch": 0.4915894267078613,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.9000199752683663e-05,
+      "loss": 0.5374,
+      "step": 179
+    },
+    {
+      "epoch": 0.49433573635427397,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.8988467990451853e-05,
+      "loss": 0.6108,
+      "step": 180
+    },
+    {
+      "epoch": 0.49708204600068656,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.8976670199459953e-05,
+      "loss": 0.4189,
+      "step": 181
+    },
+    {
+      "epoch": 0.4998283556470992,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.8964806435396227e-05,
+      "loss": 0.4773,
+      "step": 182
+    },
+    {
+      "epoch": 0.4998283556470992,
+      "eval_loss": 0.5154861807823181,
+      "eval_runtime": 620.9342,
+      "eval_samples_per_second": 14.762,
+      "eval_steps_per_second": 14.762,
+      "step": 182
+    },
+    {
+      "epoch": 0.5025746652935118,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.8952876754260342e-05,
+      "loss": 0.5624,
+      "step": 183
+    },
+    {
+      "epoch": 0.5053209749399244,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.8940881212363124e-05,
+      "loss": 1.2595,
+      "step": 184
+    },
+    {
+      "epoch": 0.5080672845863371,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.8928819866326262e-05,
+      "loss": 0.6287,
+      "step": 185
+    },
+    {
+      "epoch": 0.5108135942327497,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.891669277308206e-05,
+      "loss": 0.4508,
+      "step": 186
+    },
+    {
+      "epoch": 0.5135599038791624,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.8904499989873166e-05,
+      "loss": 0.5141,
+      "step": 187
+    },
+    {
+      "epoch": 0.516306213525575,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.88922415742523e-05,
+      "loss": 0.4496,
+      "step": 188
+    },
+    {
+      "epoch": 0.5190525231719877,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.8879917584081975e-05,
+      "loss": 0.5467,
+      "step": 189
+    },
+    {
+      "epoch": 0.5217988328184002,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.886752807753424e-05,
+      "loss": 0.4188,
+      "step": 190
+    },
+    {
+      "epoch": 0.5245451424648129,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.8855073113090395e-05,
+      "loss": 0.5347,
+      "step": 191
+    },
+    {
+      "epoch": 0.5272914521112255,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.8842552749540708e-05,
+      "loss": 0.4117,
+      "step": 192
+    },
+    {
+      "epoch": 0.5300377617576382,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.8829967045984155e-05,
+      "loss": 0.5413,
+      "step": 193
+    },
+    {
+      "epoch": 0.5327840714040508,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 2.8817316061828126e-05,
+      "loss": 0.5683,
+      "step": 194
+    },
+    {
+      "epoch": 0.5355303810504635,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.8804599856788154e-05,
+      "loss": 0.3851,
+      "step": 195
+    },
+    {
+      "epoch": 0.5382766906968761,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.8791818490887628e-05,
+      "loss": 0.42,
+      "step": 196
+    },
+    {
+      "epoch": 0.5410230003432887,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.8778972024457504e-05,
+      "loss": 0.5491,
+      "step": 197
+    },
+    {
+      "epoch": 0.5437693099897013,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.876606051813604e-05,
+      "loss": 0.5299,
+      "step": 198
+    },
+    {
+      "epoch": 0.546515619636114,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.8753084032868494e-05,
+      "loss": 0.4881,
+      "step": 199
+    },
+    {
+      "epoch": 0.5492619292825266,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8740042629906833e-05,
+      "loss": 0.4698,
+      "step": 200
+    },
+    {
+      "epoch": 0.5520082389289392,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.8726936370809455e-05,
+      "loss": 0.5685,
+      "step": 201
+    },
+    {
+      "epoch": 0.5547545485753519,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.8713765317440895e-05,
+      "loss": 0.5536,
+      "step": 202
+    },
+    {
+      "epoch": 0.5575008582217645,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.870052953197152e-05,
+      "loss": 0.4891,
+      "step": 203
+    },
+    {
+      "epoch": 0.5602471678681772,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.8687229076877274e-05,
+      "loss": 0.4182,
+      "step": 204
+    },
+    {
+      "epoch": 0.5629934775145897,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.867386401493932e-05,
+      "loss": 0.507,
+      "step": 205
+    },
+    {
+      "epoch": 0.5657397871610024,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 2.8660434409243817e-05,
+      "loss": 0.4052,
+      "step": 206
+    },
+    {
+      "epoch": 0.568486096807415,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.8646940323181553e-05,
+      "loss": 0.4503,
+      "step": 207
+    },
+    {
+      "epoch": 0.5712324064538277,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.86333818204477e-05,
+      "loss": 0.4234,
+      "step": 208
+    },
+    {
+      "epoch": 0.5739787161002403,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8619758965041488e-05,
+      "loss": 0.5319,
+      "step": 209
+    },
+    {
+      "epoch": 0.576725025746653,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.8606071821265888e-05,
+      "loss": 0.5282,
+      "step": 210
+    },
+    {
+      "epoch": 0.5794713353930656,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.8592320453727356e-05,
+      "loss": 0.4596,
+      "step": 211
+    },
+    {
+      "epoch": 0.5822176450394781,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.857850492733548e-05,
+      "loss": 0.5258,
+      "step": 212
+    },
+    {
+      "epoch": 0.5849639546858908,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.856462530730269e-05,
+      "loss": 0.4836,
+      "step": 213
+    },
+    {
+      "epoch": 0.5877102643323034,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.855068165914397e-05,
+      "loss": 0.4973,
+      "step": 214
+    },
+    {
+      "epoch": 0.5904565739787161,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8536674048676506e-05,
+      "loss": 0.5643,
+      "step": 215
+    },
+    {
+      "epoch": 0.5932028836251287,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.8522602542019425e-05,
+      "loss": 0.476,
+      "step": 216
+    },
+    {
+      "epoch": 0.5959491932715414,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.850846720559345e-05,
+      "loss": 0.4767,
+      "step": 217
+    },
+    {
+      "epoch": 0.598695502917954,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.8494268106120586e-05,
+      "loss": 0.5567,
+      "step": 218
+    },
+    {
+      "epoch": 0.6014418125643667,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8480005310623823e-05,
+      "loss": 0.536,
+      "step": 219
+    },
+    {
+      "epoch": 0.6041881222107792,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.8465678886426814e-05,
+      "loss": 0.4813,
+      "step": 220
+    },
+    {
+      "epoch": 0.6069344318571919,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.845128890115355e-05,
+      "loss": 0.4215,
+      "step": 221
+    },
+    {
+      "epoch": 0.6096807415036045,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8436835422728036e-05,
+      "loss": 0.547,
+      "step": 222
+    },
+    {
+      "epoch": 0.6124270511500172,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.8422318519373996e-05,
+      "loss": 0.4629,
+      "step": 223
+    },
+    {
+      "epoch": 0.6151733607964298,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 2.8407738259614524e-05,
+      "loss": 0.4823,
+      "step": 224
+    },
+    {
+      "epoch": 0.6179196704428425,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.8393094712271772e-05,
+      "loss": 0.5568,
+      "step": 225
+    },
+    {
+      "epoch": 0.6206659800892551,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8378387946466623e-05,
+      "loss": 0.4709,
+      "step": 226
+    },
+    {
+      "epoch": 0.6234122897356676,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8363618031618364e-05,
+      "loss": 0.4205,
+      "step": 227
+    },
+    {
+      "epoch": 0.6261585993820803,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.8348785037444366e-05,
+      "loss": 0.4985,
+      "step": 228
+    },
+    {
+      "epoch": 0.6289049090284929,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.8333889033959746e-05,
+      "loss": 0.4527,
+      "step": 229
+    },
+    {
+      "epoch": 0.6316512186749056,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8318930091477037e-05,
+      "loss": 0.582,
+      "step": 230
+    },
+    {
+      "epoch": 0.6343975283213182,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.8303908280605854e-05,
+      "loss": 0.5028,
+      "step": 231
+    },
+    {
+      "epoch": 0.6371438379677309,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8288823672252586e-05,
+      "loss": 0.5349,
+      "step": 232
+    },
+    {
+      "epoch": 0.6398901476141435,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.827367633762001e-05,
+      "loss": 0.4251,
+      "step": 233
+    },
+    {
+      "epoch": 0.6426364572605562,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.825846634820701e-05,
+      "loss": 0.5079,
+      "step": 234
+    },
+    {
+      "epoch": 0.6453827669069687,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.824319377580821e-05,
+      "loss": 1.2174,
+      "step": 235
+    },
+    {
+      "epoch": 0.6481290765533814,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.8227858692513626e-05,
+      "loss": 0.4188,
+      "step": 236
+    },
+    {
+      "epoch": 0.650875386199794,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.821246117070835e-05,
+      "loss": 0.4767,
+      "step": 237
+    },
+    {
+      "epoch": 0.6536216958462067,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.8197001283072205e-05,
+      "loss": 0.4736,
+      "step": 238
+    },
+    {
+      "epoch": 0.6563680054926193,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 2.8181479102579383e-05,
+      "loss": 0.388,
+      "step": 239
+    },
+    {
+      "epoch": 0.659114315139032,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8165894702498116e-05,
+      "loss": 0.6023,
+      "step": 240
+    },
+    {
+      "epoch": 0.6618606247854446,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.8150248156390327e-05,
+      "loss": 0.5319,
+      "step": 241
+    },
+    {
+      "epoch": 0.6646069344318571,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.8134539538111286e-05,
+      "loss": 0.5133,
+      "step": 242
+    },
+    {
+      "epoch": 0.6673532440782698,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.8118768921809258e-05,
+      "loss": 0.4813,
+      "step": 243
+    },
+    {
+      "epoch": 0.6700995537246824,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.8102936381925143e-05,
+      "loss": 0.5085,
+      "step": 244
+    },
+    {
+      "epoch": 0.6728458633710951,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.8087041993192148e-05,
+      "loss": 0.4245,
+      "step": 245
+    },
+    {
+      "epoch": 0.6755921730175077,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.8071085830635404e-05,
+      "loss": 0.5026,
+      "step": 246
+    },
+    {
+      "epoch": 0.6783384826639204,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8055067969571647e-05,
+      "loss": 0.5615,
+      "step": 247
+    },
+    {
+      "epoch": 0.681084792310333,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.803898848560883e-05,
+      "loss": 0.4929,
+      "step": 248
+    },
+    {
+      "epoch": 0.6838311019567456,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.802284745464579e-05,
+      "loss": 0.5747,
+      "step": 249
+    },
+    {
+      "epoch": 0.6865774116031582,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.800664495287187e-05,
+      "loss": 0.4181,
+      "step": 250
+    },
+    {
+      "epoch": 0.6893237212495709,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.7990381056766583e-05,
+      "loss": 0.548,
+      "step": 251
+    },
+    {
+      "epoch": 0.6920700308959835,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.797405584309922e-05,
+      "loss": 0.5344,
+      "step": 252
+    },
+    {
+      "epoch": 0.6948163405423962,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.7957669388928517e-05,
+      "loss": 0.4484,
+      "step": 253
+    },
+    {
+      "epoch": 0.6975626501888088,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 2.7941221771602278e-05,
+      "loss": 0.5194,
+      "step": 254
+    },
+    {
+      "epoch": 0.7003089598352215,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.7924713068757004e-05,
+      "loss": 0.4297,
+      "step": 255
+    },
+    {
+      "epoch": 0.7030552694816341,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.7908143358317545e-05,
+      "loss": 0.4723,
+      "step": 256
+    },
+    {
+      "epoch": 0.7058015791280466,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.7891512718496712e-05,
+      "loss": 0.4401,
+      "step": 257
+    },
+    {
+      "epoch": 0.7085478887744593,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.7874821227794915e-05,
+      "loss": 0.5961,
+      "step": 258
+    },
+    {
+      "epoch": 0.7112941984208719,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.78580689649998e-05,
+      "loss": 0.5483,
+      "step": 259
+    },
+    {
+      "epoch": 0.7140405080672846,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.7841256009185876e-05,
+      "loss": 0.493,
+      "step": 260
+    },
+    {
+      "epoch": 0.7167868177136972,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.782438243971412e-05,
+      "loss": 0.5366,
+      "step": 261
+    },
+    {
+      "epoch": 0.7195331273601099,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.7807448336231635e-05,
+      "loss": 0.3991,
+      "step": 262
+    },
+    {
+      "epoch": 0.7222794370065225,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.7790453778671248e-05,
+      "loss": 0.528,
+      "step": 263
+    },
+    {
+      "epoch": 0.7250257466529351,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.7773398847251152e-05,
+      "loss": 0.4221,
+      "step": 264
+    },
+    {
+      "epoch": 0.7277720562993477,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.7756283622474515e-05,
+      "loss": 0.4483,
+      "step": 265
+    },
+    {
+      "epoch": 0.7305183659457604,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.77391081851291e-05,
+      "loss": 0.4633,
+      "step": 266
+    },
+    {
+      "epoch": 0.733264675592173,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.7721872616286888e-05,
+      "loss": 0.5595,
+      "step": 267
+    },
+    {
+      "epoch": 0.7360109852385857,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.7704576997303694e-05,
+      "loss": 0.5091,
+      "step": 268
+    },
+    {
+      "epoch": 0.7387572948849983,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 2.768722140981879e-05,
+      "loss": 0.4357,
+      "step": 269
+    },
+    {
+      "epoch": 0.741503604531411,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.766980593575451e-05,
+      "loss": 0.4608,
+      "step": 270
+    },
+    {
+      "epoch": 0.7442499141778236,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 2.765233065731586e-05,
+      "loss": 0.4593,
+      "step": 271
+    },
+    {
+      "epoch": 0.7469962238242361,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.7634795656990143e-05,
+      "loss": 0.5097,
+      "step": 272
+    },
+    {
+      "epoch": 0.7497425334706488,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.761720101754656e-05,
+      "loss": 0.4375,
+      "step": 273
+    },
+    {
+      "epoch": 0.7497425334706488,
+      "eval_loss": 0.5116191506385803,
+      "eval_runtime": 620.1922,
+      "eval_samples_per_second": 14.779,
+      "eval_steps_per_second": 14.779,
+      "step": 273
+    },
+    {
+      "epoch": 0.7524888431170614,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.7599546822035817e-05,
+      "loss": 0.5089,
+      "step": 274
+    },
+    {
+      "epoch": 0.7552351527634741,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.758183315378976e-05,
+      "loss": 0.5961,
+      "step": 275
+    },
+    {
+      "epoch": 0.7579814624098867,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.7564060096420925e-05,
+      "loss": 0.4763,
+      "step": 276
+    },
+    {
+      "epoch": 0.7607277720562994,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.754622773382221e-05,
+      "loss": 0.5076,
+      "step": 277
+    },
+    {
+      "epoch": 0.763474081702712,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.7528336150166436e-05,
+      "loss": 0.4411,
+      "step": 278
+    },
+    {
+      "epoch": 0.7662203913491246,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.751038542990595e-05,
+      "loss": 0.5316,
+      "step": 279
+    },
+    {
+      "epoch": 0.7689667009955372,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.7492375657772254e-05,
+      "loss": 0.4153,
+      "step": 280
+    },
+    {
+      "epoch": 0.7717130106419499,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.7474306918775576e-05,
+      "loss": 0.5106,
+      "step": 281
+    },
+    {
+      "epoch": 0.7744593202883625,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.745617929820449e-05,
+      "loss": 0.474,
+      "step": 282
+    },
+    {
+      "epoch": 0.7772056299347752,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.74379928816255e-05,
+      "loss": 1.2147,
+      "step": 283
+    },
+    {
+      "epoch": 0.7799519395811878,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.7419747754882637e-05,
+      "loss": 0.5727,
+      "step": 284
+    },
+    {
+      "epoch": 0.7826982492276005,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.740144400409707e-05,
+      "loss": 0.5203,
+      "step": 285
+    },
+    {
+      "epoch": 0.7854445588740131,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.738308171566667e-05,
+      "loss": 0.5998,
+      "step": 286
+    },
+    {
+      "epoch": 0.7881908685204256,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.7364660976265624e-05,
+      "loss": 0.5133,
+      "step": 287
+    },
+    {
+      "epoch": 0.7909371781668383,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.7346181872844037e-05,
+      "loss": 0.4711,
+      "step": 288
+    },
+    {
+      "epoch": 0.7936834878132509,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.7327644492627487e-05,
+      "loss": 0.5563,
+      "step": 289
+    },
+    {
+      "epoch": 0.7964297974596636,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.7309048923116635e-05,
+      "loss": 0.4684,
+      "step": 290
+    },
+    {
+      "epoch": 0.7991761071060762,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.729039525208682e-05,
+      "loss": 0.4581,
+      "step": 291
+    },
+    {
+      "epoch": 0.8019224167524889,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.7271683567587608e-05,
+      "loss": 0.4502,
+      "step": 292
+    },
+    {
+      "epoch": 0.8046687263989015,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.7252913957942435e-05,
+      "loss": 0.564,
+      "step": 293
+    },
+    {
+      "epoch": 0.8074150360453141,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.723408651174813e-05,
+      "loss": 0.4386,
+      "step": 294
+    },
+    {
+      "epoch": 0.8101613456917267,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.7215201317874537e-05,
+      "loss": 0.5623,
+      "step": 295
+    },
+    {
+      "epoch": 0.8129076553381394,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.7196258465464087e-05,
+      "loss": 0.5303,
+      "step": 296
+    },
+    {
+      "epoch": 0.815653964984552,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.7177258043931354e-05,
+      "loss": 0.5094,
+      "step": 297
+    },
+    {
+      "epoch": 0.8184002746309647,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.7158200142962665e-05,
+      "loss": 0.502,
+      "step": 298
+    },
+    {
+      "epoch": 0.8211465842773773,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.7139084852515665e-05,
+      "loss": 0.4744,
+      "step": 299
+    },
+    {
+      "epoch": 0.82389289392379,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.7119912262818878e-05,
+      "loss": 0.5895,
+      "step": 300
+    },
+    {
+      "epoch": 0.8266392035702025,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.7100682464371306e-05,
+      "loss": 0.3948,
+      "step": 301
+    },
+    {
+      "epoch": 0.8293855132166151,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.7081395547941986e-05,
+      "loss": 0.4514,
+      "step": 302
+    },
+    {
+      "epoch": 0.8321318228630278,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.7062051604569562e-05,
+      "loss": 0.4525,
+      "step": 303
+    },
+    {
+      "epoch": 0.8348781325094404,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.7042650725561854e-05,
+      "loss": 0.4161,
+      "step": 304
+    },
+    {
+      "epoch": 0.8376244421558531,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.7023193002495447e-05,
+      "loss": 0.5065,
+      "step": 305
+    },
+    {
+      "epoch": 0.8403707518022657,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.7003678527215224e-05,
+      "loss": 1.3831,
+      "step": 306
+    },
+    {
+      "epoch": 0.8431170614486784,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.6984107391833972e-05,
+      "loss": 0.5368,
+      "step": 307
+    },
+    {
+      "epoch": 0.845863371095091,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 2.6964479688731897e-05,
+      "loss": 0.4434,
+      "step": 308
+    },
+    {
+      "epoch": 0.8486096807415036,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.694479551055625e-05,
+      "loss": 0.5286,
+      "step": 309
+    },
+    {
+      "epoch": 0.8513559903879162,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.6925054950220834e-05,
+      "loss": 0.4054,
+      "step": 310
+    },
+    {
+      "epoch": 0.8541023000343289,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.69052581009056e-05,
+      "loss": 0.3735,
+      "step": 311
+    },
+    {
+      "epoch": 0.8568486096807415,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.68854050560562e-05,
+      "loss": 0.5696,
+      "step": 312
+    },
+    {
+      "epoch": 0.8595949193271541,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.6865495909383525e-05,
+      "loss": 0.4851,
+      "step": 313
+    },
+    {
+      "epoch": 0.8623412289735668,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.684553075486329e-05,
+      "loss": 0.5755,
+      "step": 314
+    },
+    {
+      "epoch": 0.8650875386199794,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.682550968673558e-05,
+      "loss": 0.5376,
+      "step": 315
+    },
+    {
+      "epoch": 0.867833848266392,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.6805432799504407e-05,
+      "loss": 0.5374,
+      "step": 316
+    },
+    {
+      "epoch": 0.8705801579128046,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.6785300187937264e-05,
+      "loss": 0.421,
+      "step": 317
+    },
+    {
+      "epoch": 0.8733264675592173,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 2.6765111947064654e-05,
+      "loss": 0.4206,
+      "step": 318
+    },
+    {
+      "epoch": 0.8760727772056299,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.6744868172179692e-05,
+      "loss": 0.5895,
+      "step": 319
+    },
+    {
+      "epoch": 0.8788190868520426,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.672456895883761e-05,
+      "loss": 0.4784,
+      "step": 320
+    },
+    {
+      "epoch": 0.8815653964984552,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.670421440285533e-05,
+      "loss": 0.4898,
+      "step": 321
+    },
+    {
+      "epoch": 0.8843117061448679,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.6683804600310997e-05,
+      "loss": 0.6258,
+      "step": 322
+    },
+    {
+      "epoch": 0.8870580157912805,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.6663339647543528e-05,
+      "loss": 0.5587,
+      "step": 323
+    },
+    {
+      "epoch": 0.8898043254376931,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.664281964115218e-05,
+      "loss": 0.4539,
+      "step": 324
+    },
+    {
+      "epoch": 0.8925506350841057,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 2.6622244677996058e-05,
+      "loss": 0.4652,
+      "step": 325
+    },
+    {
+      "epoch": 0.8952969447305184,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 2.660161485519368e-05,
+      "loss": 0.4624,
+      "step": 326
+    },
+    {
+      "epoch": 0.898043254376931,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.6580930270122524e-05,
+      "loss": 0.5089,
+      "step": 327
+    },
+    {
+      "epoch": 0.9007895640233436,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.6560191020418545e-05,
+      "loss": 0.4246,
+      "step": 328
+    },
+    {
+      "epoch": 0.9035358736697563,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.6539397203975732e-05,
+      "loss": 0.516,
+      "step": 329
+    },
+    {
+      "epoch": 0.9062821833161689,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.6518548918945646e-05,
+      "loss": 0.5008,
+      "step": 330
+    },
+    {
+      "epoch": 0.9090284929625815,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.6497646263736943e-05,
+      "loss": 0.5195,
+      "step": 331
+    },
+    {
+      "epoch": 0.9117748026089941,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.6476689337014925e-05,
+      "loss": 0.5701,
+      "step": 332
+    },
+    {
+      "epoch": 0.9145211122554068,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.6455678237701072e-05,
+      "loss": 0.5766,
+      "step": 333
+    },
+    {
+      "epoch": 0.9172674219018194,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.643461306497256e-05,
+      "loss": 0.4613,
+      "step": 334
+    },
+    {
+      "epoch": 0.9200137315482321,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.641349391826182e-05,
+      "loss": 0.4347,
+      "step": 335
+    },
+    {
+      "epoch": 0.9227600411946447,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.6392320897256034e-05,
+      "loss": 0.4371,
+      "step": 336
+    },
+    {
+      "epoch": 0.9255063508410574,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.637109410189669e-05,
+      "loss": 0.5219,
+      "step": 337
+    },
+    {
+      "epoch": 0.92825266048747,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.6349813632379103e-05,
+      "loss": 0.5435,
+      "step": 338
+    },
+    {
+      "epoch": 0.9309989701338826,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.6328479589151953e-05,
+      "loss": 0.4764,
+      "step": 339
+    },
+    {
+      "epoch": 0.9337452797802952,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.6307092072916786e-05,
+      "loss": 0.4664,
+      "step": 340
+    },
+    {
+      "epoch": 0.9364915894267078,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.628565118462756e-05,
+      "loss": 0.4723,
+      "step": 341
+    },
+    {
+      "epoch": 0.9392378990731205,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.626415702549015e-05,
+      "loss": 0.5179,
+      "step": 342
+    },
+    {
+      "epoch": 0.9419842087195331,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.62426096969619e-05,
+      "loss": 0.5736,
+      "step": 343
+    },
+    {
+      "epoch": 0.9447305183659458,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.6221009300751113e-05,
+      "loss": 0.5238,
+      "step": 344
+    },
+    {
+      "epoch": 0.9474768280123584,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.6199355938816586e-05,
+      "loss": 0.4591,
+      "step": 345
+    },
+    {
+      "epoch": 0.950223137658771,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.6177649713367136e-05,
+      "loss": 0.5288,
+      "step": 346
+    },
+    {
+      "epoch": 0.9529694473051836,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.6155890726861084e-05,
+      "loss": 0.5066,
+      "step": 347
+    },
+    {
+      "epoch": 0.9557157569515963,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 2.613407908200582e-05,
+      "loss": 0.4485,
+      "step": 348
+    },
+    {
+      "epoch": 0.9584620665980089,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.6112214881757285e-05,
+      "loss": 0.5076,
+      "step": 349
+    },
+    {
+      "epoch": 0.9612083762444216,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.6090298229319477e-05,
+      "loss": 0.5024,
+      "step": 350
+    },
+    {
+      "epoch": 0.9639546858908342,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.6068329228144016e-05,
+      "loss": 0.4839,
+      "step": 351
+    },
+    {
+      "epoch": 0.9667009955372469,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.604630798192959e-05,
+      "loss": 0.5425,
+      "step": 352
+    },
+    {
+      "epoch": 0.9694473051836594,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.60242345946215e-05,
+      "loss": 0.4468,
+      "step": 353
+    },
+    {
+      "epoch": 0.972193614830072,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.6002109170411178e-05,
+      "loss": 0.5624,
+      "step": 354
+    },
+    {
+      "epoch": 0.9749399244764847,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 2.597993181373567e-05,
+      "loss": 0.3949,
+      "step": 355
+    },
+    {
+      "epoch": 0.9776862341228973,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.5957702629277154e-05,
+      "loss": 0.5243,
+      "step": 356
+    },
+    {
+      "epoch": 0.98043254376931,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.593542172196246e-05,
+      "loss": 0.574,
+      "step": 357
+    },
+    {
+      "epoch": 0.9831788534157226,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.5913089196962547e-05,
+      "loss": 0.4708,
+      "step": 358
+    },
+    {
+      "epoch": 0.9859251630621353,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.5890705159692036e-05,
+      "loss": 0.4344,
+      "step": 359
+    },
+    {
+      "epoch": 0.9886714727085479,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.5868269715808685e-05,
+      "loss": 0.4977,
+      "step": 360
+    },
+    {
+      "epoch": 0.9914177823549605,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.58457829712129e-05,
+      "loss": 0.551,
+      "step": 361
+    },
+    {
+      "epoch": 0.9941640920013731,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.5823245032047255e-05,
+      "loss": 0.5069,
+      "step": 362
+    },
+    {
+      "epoch": 0.9969104016477858,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.5800656004695962e-05,
+      "loss": 0.5246,
+      "step": 363
+    },
+    {
+      "epoch": 0.9996567112941984,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.5778015995784385e-05,
+      "loss": 0.6325,
+      "step": 364
+    },
+    {
+      "epoch": 0.9996567112941984,
+      "eval_loss": 0.509181559085846,
+      "eval_runtime": 618.8303,
+      "eval_samples_per_second": 14.812,
+      "eval_steps_per_second": 14.812,
+      "step": 364
+    },
+    {
+      "epoch": 1.002403020940611,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.575532511217852e-05,
+      "loss": 0.607,
+      "step": 365
+    },
+    {
+      "epoch": 1.0051493305870236,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.5732583460984527e-05,
+      "loss": 0.5572,
+      "step": 366
+    },
+    {
+      "epoch": 1.0078956402334363,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.5709791149548184e-05,
+      "loss": 1.256,
+      "step": 367
+    },
+    {
+      "epoch": 1.010641949879849,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.56869482854544e-05,
+      "loss": 0.4604,
+      "step": 368
+    },
+    {
+      "epoch": 1.0020597322348095,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.5664054976526702e-05,
+      "loss": 0.5396,
+      "step": 369
+    },
+    {
+      "epoch": 1.0048060418812221,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.564111133082674e-05,
+      "loss": 0.4803,
+      "step": 370
+    },
+    {
+      "epoch": 1.0075523515276348,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 2.561811745665374e-05,
+      "loss": 0.3781,
+      "step": 371
+    },
+    {
+      "epoch": 1.0102986611740474,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.5595073462544046e-05,
+      "loss": 0.4143,
+      "step": 372
+    },
+    {
+      "epoch": 1.01304497082046,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.5571979457270565e-05,
+      "loss": 0.4698,
+      "step": 373
+    },
+    {
+      "epoch": 1.0157912804668727,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.5548835549842274e-05,
+      "loss": 0.5101,
+      "step": 374
+    },
+    {
+      "epoch": 1.0185375901132854,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 2.5525641849503685e-05,
+      "loss": 0.4252,
+      "step": 375
+    },
+    {
+      "epoch": 1.0212838997596978,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.5502398465734357e-05,
+      "loss": 0.5116,
+      "step": 376
+    },
+    {
+      "epoch": 1.0240302094061104,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.5479105508248373e-05,
+      "loss": 0.4816,
+      "step": 377
+    },
+    {
+      "epoch": 1.026776519052523,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.54557630869938e-05,
+      "loss": 0.4521,
+      "step": 378
+    },
+    {
+      "epoch": 1.0295228286989357,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.543237131215219e-05,
+      "loss": 0.4769,
+      "step": 379
+    },
+    {
+      "epoch": 1.0322691383453484,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.5408930294138065e-05,
+      "loss": 0.5011,
+      "step": 380
+    },
+    {
+      "epoch": 1.035015447991761,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.538544014359837e-05,
+      "loss": 0.407,
+      "step": 381
+    },
+    {
+      "epoch": 1.0377617576381737,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.536190097141197e-05,
+      "loss": 0.4991,
+      "step": 382
+    },
+    {
+      "epoch": 1.0405080672845863,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.5338312888689137e-05,
+      "loss": 0.5129,
+      "step": 383
+    },
+    {
+      "epoch": 1.043254376930999,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.5314676006771e-05,
+      "loss": 0.4409,
+      "step": 384
+    },
+    {
+      "epoch": 1.0460006865774116,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 2.529099043722903e-05,
+      "loss": 0.542,
+      "step": 385
+    },
+    {
+      "epoch": 1.0487469962238243,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.526725629186452e-05,
+      "loss": 0.5767,
+      "step": 386
+    },
+    {
+      "epoch": 1.051493305870237,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.5243473682708057e-05,
+      "loss": 0.5457,
+      "step": 387
+    },
+    {
+      "epoch": 1.0542396155166496,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 2.5219642722018975e-05,
+      "loss": 0.4768,
+      "step": 388
+    },
+    {
+      "epoch": 1.0569859251630622,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.5195763522284848e-05,
+      "loss": 0.58,
+      "step": 389
+    },
+    {
+      "epoch": 1.0597322348094749,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.5171836196220946e-05,
+      "loss": 0.5176,
+      "step": 390
+    },
+    {
+      "epoch": 1.0624785444558873,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.51478608567697e-05,
+      "loss": 0.4992,
+      "step": 391
+    },
+    {
+      "epoch": 1.0652248541023,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 2.512383761710019e-05,
+      "loss": 0.5167,
+      "step": 392
+    },
+    {
+      "epoch": 1.0679711637487126,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.5099766590607587e-05,
+      "loss": 1.119,
+      "step": 393
+    },
+    {
+      "epoch": 1.0707174733951252,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.5075647890912628e-05,
+      "loss": 0.4643,
+      "step": 394
+    },
+    {
+      "epoch": 1.0734637830415379,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.505148163186107e-05,
+      "loss": 0.5572,
+      "step": 395
+    },
+    {
+      "epoch": 1.0762100926879505,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.5027267927523178e-05,
+      "loss": 0.4685,
+      "step": 396
+    },
+    {
+      "epoch": 1.0789564023343632,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.500300689219315e-05,
+      "loss": 0.5597,
+      "step": 397
+    },
+    {
+      "epoch": 1.0817027119807758,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.4978698640388617e-05,
+      "loss": 0.47,
+      "step": 398
+    },
+    {
+      "epoch": 1.0844490216271885,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.495434328685007e-05,
+      "loss": 0.5364,
+      "step": 399
+    },
+    {
+      "epoch": 1.0871953312736011,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.492994094654033e-05,
+      "loss": 0.4303,
+      "step": 400
+    },
+    {
+      "epoch": 1.0899416409200138,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.490549173464402e-05,
+      "loss": 1.1982,
+      "step": 401
+    },
+    {
+      "epoch": 1.0926879505664264,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.4880995766566986e-05,
+      "loss": 0.5137,
+      "step": 402
+    },
+    {
+      "epoch": 1.095434260212839,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.4856453157935795e-05,
+      "loss": 0.4997,
+      "step": 403
+    },
+    {
+      "epoch": 1.0981805698592517,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.483186402459715e-05,
+      "loss": 0.5209,
+      "step": 404
+    },
+    {
+      "epoch": 1.1009268795056641,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.4807228482617376e-05,
+      "loss": 0.483,
+      "step": 405
+    },
+    {
+      "epoch": 1.1036731891520768,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.4782546648281848e-05,
+      "loss": 0.5055,
+      "step": 406
+    },
+    {
+      "epoch": 1.1064194987984894,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.4757818638094457e-05,
+      "loss": 0.462,
+      "step": 407
+    },
+    {
+      "epoch": 1.109165808444902,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.473304456877705e-05,
+      "loss": 0.4663,
+      "step": 408
+    },
+    {
+      "epoch": 1.1119121180913147,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.470822455726889e-05,
+      "loss": 0.5343,
+      "step": 409
+    },
+    {
+      "epoch": 1.1146584277377274,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.468335872072609e-05,
+      "loss": 0.4854,
+      "step": 410
+    },
+    {
+      "epoch": 1.11740473738414,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.4658447176521076e-05,
+      "loss": 0.5206,
+      "step": 411
+    },
+    {
+      "epoch": 1.1201510470305527,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.463349004224201e-05,
+      "loss": 0.4738,
+      "step": 412
+    },
+    {
+      "epoch": 1.1228973566769653,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.460848743569227e-05,
+      "loss": 0.5632,
+      "step": 413
+    },
+    {
+      "epoch": 1.125643666323378,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.458343947488985e-05,
+      "loss": 0.6056,
+      "step": 414
+    },
+    {
+      "epoch": 1.1283899759697906,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.4558346278066853e-05,
+      "loss": 1.1007,
+      "step": 415
+    },
+    {
+      "epoch": 1.1311362856162033,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.4533207963668883e-05,
+      "loss": 0.4747,
+      "step": 416
+    },
+    {
+      "epoch": 1.133882595262616,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.4508024650354525e-05,
+      "loss": 0.439,
+      "step": 417
+    },
+    {
+      "epoch": 1.1366289049090286,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.4482796456994757e-05,
+      "loss": 0.4913,
+      "step": 418
+    },
+    {
+      "epoch": 1.1393752145554412,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.4457523502672415e-05,
+      "loss": 0.5722,
+      "step": 419
+    },
+    {
+      "epoch": 1.1421215242018539,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.44322059066816e-05,
+      "loss": 0.3971,
+      "step": 420
+    },
+    {
+      "epoch": 1.1448678338482665,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.440684378852714e-05,
+      "loss": 0.4724,
+      "step": 421
+    },
+    {
+      "epoch": 1.147614143494679,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.438143726792403e-05,
+      "loss": 0.5305,
+      "step": 422
+    },
+    {
+      "epoch": 1.1503604531410916,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 2.435598646479683e-05,
+      "loss": 0.4924,
+      "step": 423
+    },
+    {
+      "epoch": 1.1531067627875042,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.4330491499279148e-05,
+      "loss": 0.4927,
+      "step": 424
+    },
+    {
+      "epoch": 1.1558530724339169,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.4304952491713035e-05,
+      "loss": 0.45,
+      "step": 425
+    },
+    {
+      "epoch": 1.1585993820803295,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.4279369562648424e-05,
+      "loss": 0.5892,
+      "step": 426
+    },
+    {
+      "epoch": 1.1613456917267422,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.4253742832842583e-05,
+      "loss": 0.4727,
+      "step": 427
+    },
+    {
+      "epoch": 1.1640920013731548,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.4228072423259527e-05,
+      "loss": 0.5063,
+      "step": 428
+    },
+    {
+      "epoch": 1.1668383110195675,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 2.420235845506944e-05,
+      "loss": 0.4872,
+      "step": 429
+    },
+    {
+      "epoch": 1.1695846206659801,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.4176601049648116e-05,
+      "loss": 0.3843,
+      "step": 430
+    },
+    {
+      "epoch": 1.1723309303123928,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 2.415080032857639e-05,
+      "loss": 0.4478,
+      "step": 431
+    },
+    {
+      "epoch": 1.1750772399588054,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.4124956413639548e-05,
+      "loss": 0.4964,
+      "step": 432
+    },
+    {
+      "epoch": 1.177823549605218,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.4099069426826766e-05,
+      "loss": 0.5176,
+      "step": 433
+    },
+    {
+      "epoch": 1.1805698592516307,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.4073139490330526e-05,
+      "loss": 0.5596,
+      "step": 434
+    },
+    {
+      "epoch": 1.1833161688980431,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.4047166726546047e-05,
+      "loss": 0.485,
+      "step": 435
+    },
+    {
+      "epoch": 1.1860624785444558,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.4021151258070694e-05,
+      "loss": 0.4768,
+      "step": 436
+    },
+    {
+      "epoch": 1.1888087881908684,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 2.3995093207703413e-05,
+      "loss": 0.5097,
+      "step": 437
+    },
+    {
+      "epoch": 1.191555097837281,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.3968992698444153e-05,
+      "loss": 0.5401,
+      "step": 438
+    },
+    {
+      "epoch": 1.1943014074836937,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.394284985349327e-05,
+      "loss": 0.425,
+      "step": 439
+    },
+    {
+      "epoch": 1.1970477171301064,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.3916664796250946e-05,
+      "loss": 0.3752,
+      "step": 440
+    },
+    {
+      "epoch": 1.199794026776519,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.389043765031664e-05,
+      "loss": 0.4724,
+      "step": 441
+    },
+    {
+      "epoch": 1.2025403364229317,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.386416853948845e-05,
+      "loss": 0.5598,
+      "step": 442
+    },
+    {
+      "epoch": 1.2052866460693443,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.3837857587762583e-05,
+      "loss": 0.3885,
+      "step": 443
+    },
+    {
+      "epoch": 1.208032955715757,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.3811504919332727e-05,
+      "loss": 0.4608,
+      "step": 444
+    },
+    {
+      "epoch": 1.2107792653621696,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.378511065858949e-05,
+      "loss": 0.4457,
+      "step": 445
+    },
+    {
+      "epoch": 1.2135255750085823,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.3758674930119807e-05,
+      "loss": 0.4162,
+      "step": 446
+    },
+    {
+      "epoch": 1.216271884654995,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 2.3732197858706343e-05,
+      "loss": 0.4656,
+      "step": 447
+    },
+    {
+      "epoch": 1.2190181943014076,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.370567956932692e-05,
+      "loss": 0.4525,
+      "step": 448
+    },
+    {
+      "epoch": 1.2217645039478202,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.367912018715391e-05,
+      "loss": 0.498,
+      "step": 449
+    },
+    {
+      "epoch": 1.2245108135942329,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.3652519837553655e-05,
+      "loss": 0.3724,
+      "step": 450
+    },
+    {
+      "epoch": 1.2272571232406453,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.3625878646085873e-05,
+      "loss": 0.3611,
+      "step": 451
+    },
+    {
+      "epoch": 1.230003432887058,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.3599196738503068e-05,
+      "loss": 0.4002,
+      "step": 452
+    },
+    {
+      "epoch": 1.2327497425334706,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.3572474240749932e-05,
+      "loss": 0.5691,
+      "step": 453
+    },
+    {
+      "epoch": 1.2354960521798832,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.354571127896275e-05,
+      "loss": 0.536,
+      "step": 454
+    },
+    {
+      "epoch": 1.2382423618262959,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.3518907979468807e-05,
+      "loss": 0.4385,
+      "step": 455
+    },
+    {
+      "epoch": 1.2382423618262959,
+      "eval_loss": 0.5073373913764954,
+      "eval_runtime": 627.5271,
+      "eval_samples_per_second": 14.607,
+      "eval_steps_per_second": 14.607,
+      "step": 455
+    },
+    {
+      "epoch": 1.2409886714727085,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.349206446878578e-05,
+      "loss": 0.5131,
+      "step": 456
+    },
+    {
+      "epoch": 1.2437349811191212,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.346518087362118e-05,
+      "loss": 0.4821,
+      "step": 457
+    },
+    {
+      "epoch": 1.2464812907655338,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.3438257320871704e-05,
+      "loss": 0.5344,
+      "step": 458
+    },
+    {
+      "epoch": 1.2492276004119465,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.3411293937622658e-05,
+      "loss": 0.4752,
+      "step": 459
+    },
+    {
+      "epoch": 1.2519739100583591,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.338429085114737e-05,
+      "loss": 0.4887,
+      "step": 460
+    },
+    {
+      "epoch": 1.2547202197047718,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.335724818890656e-05,
+      "loss": 0.4445,
+      "step": 461
+    },
+    {
+      "epoch": 1.2574665293511844,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.3330166078547763e-05,
+      "loss": 0.5841,
+      "step": 462
+    },
+    {
+      "epoch": 1.2602128389975968,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.3303044647904725e-05,
+      "loss": 0.519,
+      "step": 463
+    },
+    {
+      "epoch": 1.2629591486440095,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.3275884024996784e-05,
+      "loss": 0.5149,
+      "step": 464
+    },
+    {
+      "epoch": 1.2657054582904221,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.324868433802827e-05,
+      "loss": 0.4681,
+      "step": 465
+    },
+    {
+      "epoch": 1.2684517679368348,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.3221445715387917e-05,
+      "loss": 0.5058,
+      "step": 466
+    },
+    {
+      "epoch": 1.2711980775832474,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.319416828564824e-05,
+      "loss": 0.5142,
+      "step": 467
+    },
+    {
+      "epoch": 1.27394438722966,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.3166852177564925e-05,
+      "loss": 0.4682,
+      "step": 468
+    },
+    {
+      "epoch": 1.2766906968760727,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.3139497520076233e-05,
+      "loss": 0.4361,
+      "step": 469
+    },
+    {
+      "epoch": 1.2794370065224854,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.3112104442302393e-05,
+      "loss": 0.5738,
+      "step": 470
+    },
+    {
+      "epoch": 1.282183316168898,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.3084673073544976e-05,
+      "loss": 0.4828,
+      "step": 471
+    },
+    {
+      "epoch": 1.2849296258153107,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.3057203543286297e-05,
+      "loss": 0.503,
+      "step": 472
+    },
+    {
+      "epoch": 1.2876759354617233,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.3029695981188818e-05,
+      "loss": 0.5526,
+      "step": 473
+    },
+    {
+      "epoch": 1.290422245108136,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.3002150517094496e-05,
+      "loss": 0.4757,
+      "step": 474
+    },
+    {
+      "epoch": 1.2931685547545486,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.297456728102421e-05,
+      "loss": 0.5773,
+      "step": 475
+    },
+    {
+      "epoch": 1.2959148644009613,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.294694640317713e-05,
+      "loss": 0.5248,
+      "step": 476
+    },
+    {
+      "epoch": 1.298661174047374,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.2919288013930094e-05,
+      "loss": 0.4915,
+      "step": 477
+    },
+    {
+      "epoch": 1.3014074836937866,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.2891592243837015e-05,
+      "loss": 0.5389,
+      "step": 478
+    },
+    {
+      "epoch": 1.3041537933401992,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.286385922362824e-05,
+      "loss": 0.4232,
+      "step": 479
+    },
+    {
+      "epoch": 1.3069001029866119,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.2836089084209955e-05,
+      "loss": 0.5072,
+      "step": 480
+    },
+    {
+      "epoch": 1.3096464126330245,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.280828195666355e-05,
+      "loss": 0.54,
+      "step": 481
+    },
+    {
+      "epoch": 1.312392722279437,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.2780437972245014e-05,
+      "loss": 0.5446,
+      "step": 482
+    },
+    {
+      "epoch": 1.3151390319258496,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.2752557262384307e-05,
+      "loss": 0.4725,
+      "step": 483
+    },
+    {
+      "epoch": 1.3178853415722622,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.2724639958684733e-05,
+      "loss": 1.2587,
+      "step": 484
+    },
+    {
+      "epoch": 1.3206316512186749,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.2696686192922342e-05,
+      "loss": 0.4965,
+      "step": 485
+    },
+    {
+      "epoch": 1.3233779608650875,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.2668696097045284e-05,
+      "loss": 0.5382,
+      "step": 486
+    },
+    {
+      "epoch": 1.3261242705115002,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.2640669803173195e-05,
+      "loss": 0.4305,
+      "step": 487
+    },
+    {
+      "epoch": 1.3288705801579128,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.2612607443596572e-05,
+      "loss": 0.4622,
+      "step": 488
+    },
+    {
+      "epoch": 1.3316168898043255,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 2.258450915077616e-05,
+      "loss": 0.4975,
+      "step": 489
+    },
+    {
+      "epoch": 1.3343631994507381,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.2556375057342306e-05,
+      "loss": 0.6356,
+      "step": 490
+    },
+    {
+      "epoch": 1.3371095090971508,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.2528205296094356e-05,
+      "loss": 0.4422,
+      "step": 491
+    },
+    {
+      "epoch": 1.3398558187435634,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.25e-05,
+      "loss": 0.446,
+      "step": 492
+    },
+    {
+      "epoch": 1.3426021283899758,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.247175930219468e-05,
+      "loss": 0.5996,
+      "step": 493
+    },
+    {
+      "epoch": 1.3453484380363885,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.2443483335980924e-05,
+      "loss": 0.5905,
+      "step": 494
+    },
+    {
+      "epoch": 1.3480947476828011,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.2415172234827754e-05,
+      "loss": 0.5824,
+      "step": 495
+    },
+    {
+      "epoch": 1.3508410573292138,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.238682613237001e-05,
+      "loss": 0.4885,
+      "step": 496
+    },
+    {
+      "epoch": 1.3535873669756264,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.2358445162407775e-05,
+      "loss": 0.587,
+      "step": 497
+    },
+    {
+      "epoch": 1.356333676622039,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.2330029458905697e-05,
+      "loss": 0.5453,
+      "step": 498
+    },
+    {
+      "epoch": 1.3590799862684517,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.230157915599238e-05,
+      "loss": 0.4596,
+      "step": 499
+    },
+    {
+      "epoch": 1.3618262959148644,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.2273094387959747e-05,
+      "loss": 0.4349,
+      "step": 500
+    },
+    {
+      "epoch": 1.364572605561277,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.2244575289262394e-05,
+      "loss": 0.4613,
+      "step": 501
+    },
+    {
+      "epoch": 1.3673189152076897,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.221602199451698e-05,
+      "loss": 0.4176,
+      "step": 502
+    },
+    {
+      "epoch": 1.3700652248541023,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.2187434638501564e-05,
+      "loss": 0.4799,
+      "step": 503
+    },
+    {
+      "epoch": 1.372811534500515,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.215881335615499e-05,
+      "loss": 0.4335,
+      "step": 504
+    },
+    {
+      "epoch": 1.3755578441469276,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.2130158282576245e-05,
+      "loss": 0.5999,
+      "step": 505
+    },
+    {
+      "epoch": 1.3783041537933403,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.2101469553023807e-05,
+      "loss": 0.4654,
+      "step": 506
+    },
+    {
+      "epoch": 1.381050463439753,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.2072747302915026e-05,
+      "loss": 0.4423,
+      "step": 507
+    },
+    {
+      "epoch": 1.3837967730861656,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.2043991667825478e-05,
+      "loss": 0.5145,
+      "step": 508
+    },
+    {
+      "epoch": 1.3865430827325782,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 2.2015202783488316e-05,
+      "loss": 0.5894,
+      "step": 509
+    },
+    {
+      "epoch": 1.3892893923789909,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.1986380785793646e-05,
+      "loss": 0.5228,
+      "step": 510
+    },
+    {
+      "epoch": 1.3920357020254035,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.195752581078787e-05,
+      "loss": 0.5529,
+      "step": 511
+    },
+    {
+      "epoch": 1.394782011671816,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.1928637994673053e-05,
+      "loss": 0.5783,
+      "step": 512
+    },
+    {
+      "epoch": 1.3975283213182286,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.1899717473806273e-05,
+      "loss": 0.418,
+      "step": 513
+    },
+    {
+      "epoch": 1.4002746309646412,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.1870764384698992e-05,
+      "loss": 0.4945,
+      "step": 514
+    },
+    {
+      "epoch": 1.4030209406110539,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.1841778864016396e-05,
+      "loss": 0.496,
+      "step": 515
+    },
+    {
+      "epoch": 1.4057672502574665,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.1812761048576752e-05,
+      "loss": 0.5087,
+      "step": 516
+    },
+    {
+      "epoch": 1.4085135599038792,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.1783711075350766e-05,
+      "loss": 0.4898,
+      "step": 517
+    },
+    {
+      "epoch": 1.4112598695502918,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.1754629081460947e-05,
+      "loss": 0.4379,
+      "step": 518
+    },
+    {
+      "epoch": 1.4140061791967045,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.172551520418093e-05,
+      "loss": 0.4827,
+      "step": 519
+    },
+    {
+      "epoch": 1.416752488843117,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.169636958093487e-05,
+      "loss": 0.5007,
+      "step": 520
+    },
+    {
+      "epoch": 1.4194987984895298,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.1667192349296746e-05,
+      "loss": 0.4651,
+      "step": 521
+    },
+    {
+      "epoch": 1.4222451081359424,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.1637983646989758e-05,
+      "loss": 0.4674,
+      "step": 522
+    },
+    {
+      "epoch": 1.4249914177823548,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.1608743611885633e-05,
+      "loss": 0.4794,
+      "step": 523
+    },
+    {
+      "epoch": 1.4277377274287675,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.1579472382004015e-05,
+      "loss": 0.5292,
+      "step": 524
+    },
+    {
+      "epoch": 1.4304840370751801,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.1550170095511784e-05,
+      "loss": 0.4964,
+      "step": 525
+    },
+    {
+      "epoch": 1.4332303467215928,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 2.1520836890722416e-05,
+      "loss": 0.4236,
+      "step": 526
+    },
+    {
+      "epoch": 1.4359766563680054,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.149147290609533e-05,
+      "loss": 0.4859,
+      "step": 527
+    },
+    {
+      "epoch": 1.438722966014418,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.146207828023524e-05,
+      "loss": 0.4659,
+      "step": 528
+    },
+    {
+      "epoch": 1.4414692756608307,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.1432653151891473e-05,
+      "loss": 0.4424,
+      "step": 529
+    },
+    {
+      "epoch": 1.4442155853072434,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.1403197659957356e-05,
+      "loss": 0.4515,
+      "step": 530
+    },
+    {
+      "epoch": 1.446961894953656,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.137371194346953e-05,
+      "loss": 0.4618,
+      "step": 531
+    },
+    {
+      "epoch": 1.4497082046000687,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.1344196141607297e-05,
+      "loss": 0.3928,
+      "step": 532
+    },
+    {
+      "epoch": 1.4524545142464813,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.1314650393691984e-05,
+      "loss": 0.4598,
+      "step": 533
+    },
+    {
+      "epoch": 1.455200823892894,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.1285074839186257e-05,
+      "loss": 0.5646,
+      "step": 534
+    },
+    {
+      "epoch": 1.4579471335393066,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.1255469617693476e-05,
+      "loss": 0.5984,
+      "step": 535
+    },
+    {
+      "epoch": 1.4606934431857193,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.122583486895705e-05,
+      "loss": 0.5419,
+      "step": 536
+    },
+    {
+      "epoch": 1.463439752832132,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.119617073285974e-05,
+      "loss": 0.5481,
+      "step": 537
+    },
+    {
+      "epoch": 1.4661860624785445,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.116647734942305e-05,
+      "loss": 0.5588,
+      "step": 538
+    },
+    {
+      "epoch": 1.4689323721249572,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.113675485880652e-05,
+      "loss": 0.5621,
+      "step": 539
+    },
+    {
+      "epoch": 1.4716786817713698,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.110700340130708e-05,
+      "loss": 0.5056,
+      "step": 540
+    },
+    {
+      "epoch": 1.4744249914177825,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.1077223117358395e-05,
+      "loss": 0.5526,
+      "step": 541
+    },
+    {
+      "epoch": 1.477171301064195,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.104741414753021e-05,
+      "loss": 0.5414,
+      "step": 542
+    },
+    {
+      "epoch": 1.4799176107106076,
+      "grad_norm": 0.056640625,
+      "learning_rate": 2.1017576632527662e-05,
+      "loss": 0.5472,
+      "step": 543
+    },
+    {
+      "epoch": 1.4826639203570202,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.098771071319062e-05,
+      "loss": 0.4568,
+      "step": 544
+    },
+    {
+      "epoch": 1.4854102300034329,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.0957816530493037e-05,
+      "loss": 0.4277,
+      "step": 545
+    },
+    {
+      "epoch": 1.4881565396498455,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.0927894225542282e-05,
+      "loss": 0.4949,
+      "step": 546
+    },
+    {
+      "epoch": 1.4881565396498455,
+      "eval_loss": 0.5060501098632812,
+      "eval_runtime": 630.5882,
+      "eval_samples_per_second": 14.536,
+      "eval_steps_per_second": 14.536,
+      "step": 546
+    },
+    {
+      "epoch": 1.4909028492962582,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.089794393957846e-05,
+      "loss": 0.3558,
+      "step": 547
+    },
+    {
+      "epoch": 1.4936491589426708,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.086796581397374e-05,
+      "loss": 0.4622,
+      "step": 548
+    },
+    {
+      "epoch": 1.4963954685890835,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.083795999023173e-05,
+      "loss": 0.5402,
+      "step": 549
+    },
+    {
+      "epoch": 1.499141778235496,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.080792660998676e-05,
+      "loss": 0.5271,
+      "step": 550
+    },
+    {
+      "epoch": 1.5018880878819085,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.0777865815003234e-05,
+      "loss": 0.5152,
+      "step": 551
+    },
+    {
+      "epoch": 1.5046343975283212,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.074777774717496e-05,
+      "loss": 0.5099,
+      "step": 552
+    },
+    {
+      "epoch": 1.5073807071747338,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.0717662548524482e-05,
+      "loss": 0.4075,
+      "step": 553
+    },
+    {
+      "epoch": 1.5101270168211465,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.068752036120241e-05,
+      "loss": 0.5205,
+      "step": 554
+    },
+    {
+      "epoch": 1.5128733264675591,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.0657351327486745e-05,
+      "loss": 0.5127,
+      "step": 555
+    },
+    {
+      "epoch": 1.5156196361139718,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.0627155589782212e-05,
+      "loss": 0.5399,
+      "step": 556
+    },
+    {
+      "epoch": 1.5183659457603844,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.0596933290619572e-05,
+      "loss": 0.4869,
+      "step": 557
+    },
+    {
+      "epoch": 1.521112255406797,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.0566684572654978e-05,
+      "loss": 0.6318,
+      "step": 558
+    },
+    {
+      "epoch": 1.5238585650532097,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 2.0536409578669277e-05,
+      "loss": 0.4729,
+      "step": 559
+    },
+    {
+      "epoch": 1.5266048746996224,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.0506108451567347e-05,
+      "loss": 0.5059,
+      "step": 560
+    },
+    {
+      "epoch": 1.529351184346035,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.0475781334377426e-05,
+      "loss": 0.3829,
+      "step": 561
+    },
+    {
+      "epoch": 1.5320974939924477,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.044542837025042e-05,
+      "loss": 0.4582,
+      "step": 562
+    },
+    {
+      "epoch": 1.5348438036388603,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.0415049702459244e-05,
+      "loss": 0.5344,
+      "step": 563
+    },
+    {
+      "epoch": 1.537590113285273,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.0384645474398137e-05,
+      "loss": 0.4508,
+      "step": 564
+    },
+    {
+      "epoch": 1.5403364229316856,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 2.0354215829582005e-05,
+      "loss": 0.4973,
+      "step": 565
+    },
+    {
+      "epoch": 1.5430827325780982,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.03237609116457e-05,
+      "loss": 0.5406,
+      "step": 566
+    },
+    {
+      "epoch": 1.545829042224511,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.029328086434339e-05,
+      "loss": 0.4956,
+      "step": 567
+    },
+    {
+      "epoch": 1.5485753518709235,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 2.0262775831547847e-05,
+      "loss": 0.5642,
+      "step": 568
+    },
+    {
+      "epoch": 1.5513216615173362,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.0232245957249788e-05,
+      "loss": 0.5424,
+      "step": 569
+    },
+    {
+      "epoch": 1.5540679711637488,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.020169138555718e-05,
+      "loss": 0.4972,
+      "step": 570
+    },
+    {
+      "epoch": 1.5568142808101615,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.0171112260694576e-05,
+      "loss": 0.4511,
+      "step": 571
+    },
+    {
+      "epoch": 1.5595605904565741,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.0140508727002422e-05,
+      "loss": 0.4669,
+      "step": 572
+    },
+    {
+      "epoch": 1.5623069001029866,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.0109880928936375e-05,
+      "loss": 0.5472,
+      "step": 573
+    },
+    {
+      "epoch": 1.5650532097493992,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.007922901106663e-05,
+      "loss": 0.5493,
+      "step": 574
+    },
+    {
+      "epoch": 1.5677995193958119,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.0048553118077238e-05,
+      "loss": 0.46,
+      "step": 575
+    },
+    {
+      "epoch": 1.5705458290422245,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.0017853394765402e-05,
+      "loss": 0.6062,
+      "step": 576
+    },
+    {
+      "epoch": 1.5732921386886372,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9987129986040825e-05,
+      "loss": 0.5053,
+      "step": 577
+    },
+    {
+      "epoch": 1.5760384483350498,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.9956383036925006e-05,
+      "loss": 0.5205,
+      "step": 578
+    },
+    {
+      "epoch": 1.5787847579814624,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.9925612692550554e-05,
+      "loss": 0.5296,
+      "step": 579
+    },
+    {
+      "epoch": 1.581531067627875,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.989481909816052e-05,
+      "loss": 0.577,
+      "step": 580
+    },
+    {
+      "epoch": 1.5842773772742875,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.986400239910769e-05,
+      "loss": 0.5867,
+      "step": 581
+    },
+    {
+      "epoch": 1.5870236869207002,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9833162740853916e-05,
+      "loss": 0.5371,
+      "step": 582
+    },
+    {
+      "epoch": 1.5897699965671128,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.980230026896942e-05,
+      "loss": 0.4848,
+      "step": 583
+    },
+    {
+      "epoch": 1.5925163062135255,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.977141512913211e-05,
+      "loss": 0.4747,
+      "step": 584
+    },
+    {
+      "epoch": 1.5952626158599381,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.974050746712689e-05,
+      "loss": 0.4296,
+      "step": 585
+    },
+    {
+      "epoch": 1.5980089255063508,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.9709577428844984e-05,
+      "loss": 0.4943,
+      "step": 586
+    },
+    {
+      "epoch": 1.6007552351527634,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.967862516028321e-05,
+      "loss": 0.487,
+      "step": 587
+    },
+    {
+      "epoch": 1.603501544799176,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.9647650807543358e-05,
+      "loss": 0.5275,
+      "step": 588
+    },
+    {
+      "epoch": 1.6062478544455887,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.961665451683143e-05,
+      "loss": 0.557,
+      "step": 589
+    },
+    {
+      "epoch": 1.6089941640920014,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.9585636434456988e-05,
+      "loss": 0.4689,
+      "step": 590
+    },
+    {
+      "epoch": 1.611740473738414,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.9554596706832457e-05,
+      "loss": 0.5351,
+      "step": 591
+    },
+    {
+      "epoch": 1.6144867833848267,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.952353548047243e-05,
+      "loss": 0.5714,
+      "step": 592
+    },
+    {
+      "epoch": 1.6172330930312393,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.9492452901992987e-05,
+      "loss": 0.5468,
+      "step": 593
+    },
+    {
+      "epoch": 1.619979402677652,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.946134911811099e-05,
+      "loss": 0.5812,
+      "step": 594
+    },
+    {
+      "epoch": 1.6227257123240646,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9430224275643388e-05,
+      "loss": 0.5367,
+      "step": 595
+    },
+    {
+      "epoch": 1.6254720219704772,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.9399078521506546e-05,
+      "loss": 0.5746,
+      "step": 596
+    },
+    {
+      "epoch": 1.62821833161689,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.9367912002715524e-05,
+      "loss": 0.4458,
+      "step": 597
+    },
+    {
+      "epoch": 1.6309646412633025,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.93367248663834e-05,
+      "loss": 0.4413,
+      "step": 598
+    },
+    {
+      "epoch": 1.6337109509097152,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.9305517259720573e-05,
+      "loss": 0.5666,
+      "step": 599
+    },
+    {
+      "epoch": 1.6364572605561278,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.9274289330034068e-05,
+      "loss": 0.5282,
+      "step": 600
+    },
+    {
+      "epoch": 1.6392035702025405,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.924304122472683e-05,
+      "loss": 0.5065,
+      "step": 601
+    },
+    {
+      "epoch": 1.6419498798489531,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.9211773091297057e-05,
+      "loss": 0.5519,
+      "step": 602
+    },
+    {
+      "epoch": 1.6446961894953656,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9180485077337462e-05,
+      "loss": 0.5044,
+      "step": 603
+    },
+    {
+      "epoch": 1.6474424991417782,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.9149177330534614e-05,
+      "loss": 0.4895,
+      "step": 604
+    },
+    {
+      "epoch": 1.6501888087881909,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.9117849998668212e-05,
+      "loss": 0.4553,
+      "step": 605
+    },
+    {
+      "epoch": 1.6529351184346035,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.9086503229610418e-05,
+      "loss": 0.5583,
+      "step": 606
+    },
+    {
+      "epoch": 1.6556814280810161,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.905513717132513e-05,
+      "loss": 0.3757,
+      "step": 607
+    },
+    {
+      "epoch": 1.6584277377274288,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.90237519718673e-05,
+      "loss": 0.5956,
+      "step": 608
+    },
+    {
+      "epoch": 1.6611740473738414,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.899234777938222e-05,
+      "loss": 1.1236,
+      "step": 609
+    },
+    {
+      "epoch": 1.6639203570202539,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.8960924742104856e-05,
+      "loss": 0.5466,
+      "step": 610
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.892948300835911e-05,
+      "loss": 0.4874,
+      "step": 611
+    },
+    {
+      "epoch": 1.6694129763130792,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.889802272655713e-05,
+      "loss": 0.5116,
+      "step": 612
+    },
+    {
+      "epoch": 1.6721592859594918,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.8866544045198634e-05,
+      "loss": 0.587,
+      "step": 613
+    },
+    {
+      "epoch": 1.6749055956059045,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.8835047112870163e-05,
+      "loss": 0.4174,
+      "step": 614
+    },
+    {
+      "epoch": 1.677651905252317,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.880353207824444e-05,
+      "loss": 0.4023,
+      "step": 615
+    },
+    {
+      "epoch": 1.6803982148987298,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.8771999090079613e-05,
+      "loss": 0.5134,
+      "step": 616
+    },
+    {
+      "epoch": 1.6831445245451424,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.8740448297218575e-05,
+      "loss": 0.4694,
+      "step": 617
+    },
+    {
+      "epoch": 1.685890834191555,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8708879848588268e-05,
+      "loss": 0.5185,
+      "step": 618
+    },
+    {
+      "epoch": 1.6886371438379677,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8677293893198976e-05,
+      "loss": 0.5077,
+      "step": 619
+    },
+    {
+      "epoch": 1.6913834534843804,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.864569058014361e-05,
+      "loss": 0.4517,
+      "step": 620
+    },
+    {
+      "epoch": 1.694129763130793,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.8614070058597014e-05,
+      "loss": 0.4703,
+      "step": 621
+    },
+    {
+      "epoch": 1.6968760727772056,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.8582432477815268e-05,
+      "loss": 0.5061,
+      "step": 622
+    },
+    {
+      "epoch": 1.6996223824236183,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.855077798713497e-05,
+      "loss": 0.5413,
+      "step": 623
+    },
+    {
+      "epoch": 1.702368692070031,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 1.8519106735972535e-05,
+      "loss": 0.4586,
+      "step": 624
+    },
+    {
+      "epoch": 1.7051150017164436,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.84874188738235e-05,
+      "loss": 0.5022,
+      "step": 625
+    },
+    {
+      "epoch": 1.7078613113628562,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.8455714550261793e-05,
+      "loss": 0.4945,
+      "step": 626
+    },
+    {
+      "epoch": 1.7106076210092689,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.8423993914939063e-05,
+      "loss": 0.5806,
+      "step": 627
+    },
+    {
+      "epoch": 1.7133539306556815,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.8392257117583944e-05,
+      "loss": 0.462,
+      "step": 628
+    },
+    {
+      "epoch": 1.7161002403020942,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.836050430800135e-05,
+      "loss": 0.4944,
+      "step": 629
+    },
+    {
+      "epoch": 1.7188465499485068,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.83287356360718e-05,
+      "loss": 1.1722,
+      "step": 630
+    },
+    {
+      "epoch": 1.7215928595949195,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.8296951251750667e-05,
+      "loss": 0.3718,
+      "step": 631
+    },
+    {
+      "epoch": 1.7243391692413321,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.8265151305067486e-05,
+      "loss": 0.484,
+      "step": 632
+    },
+    {
+      "epoch": 1.7270854788877446,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.8233335946125275e-05,
+      "loss": 0.4783,
+      "step": 633
+    },
+    {
+      "epoch": 1.7298317885341572,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.8201505325099782e-05,
+      "loss": 0.5684,
+      "step": 634
+    },
+    {
+      "epoch": 1.7325780981805698,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.8169659592238797e-05,
+      "loss": 0.4518,
+      "step": 635
+    },
+    {
+      "epoch": 1.7353244078269825,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.813779889786144e-05,
+      "loss": 0.4535,
+      "step": 636
+    },
+    {
+      "epoch": 1.7380707174733951,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8105923392357464e-05,
+      "loss": 0.503,
+      "step": 637
+    },
+    {
+      "epoch": 1.7380707174733951,
+      "eval_loss": 0.5051947832107544,
+      "eval_runtime": 630.1537,
+      "eval_samples_per_second": 14.546,
+      "eval_steps_per_second": 14.546,
+      "step": 637
+    },
+    {
+      "epoch": 1.7408170271198078,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.807403322618653e-05,
+      "loss": 0.4961,
+      "step": 638
+    },
+    {
+      "epoch": 1.7435633367662204,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.8042128549877483e-05,
+      "loss": 0.519,
+      "step": 639
+    },
+    {
+      "epoch": 1.7463096464126329,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.8010209514027687e-05,
+      "loss": 0.4011,
+      "step": 640
+    },
+    {
+      "epoch": 1.7490559560590455,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.7978276269302275e-05,
+      "loss": 0.3935,
+      "step": 641
+    },
+    {
+      "epoch": 1.7518022657054582,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.794632896643343e-05,
+      "loss": 0.6534,
+      "step": 642
+    },
+    {
+      "epoch": 1.7545485753518708,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.7914367756219725e-05,
+      "loss": 0.5715,
+      "step": 643
+    },
+    {
+      "epoch": 1.7572948849982835,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.7882392789525358e-05,
+      "loss": 0.5439,
+      "step": 644
+    },
+    {
+      "epoch": 1.760041194644696,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.7850404217279467e-05,
+      "loss": 0.5277,
+      "step": 645
+    },
+    {
+      "epoch": 1.7627875042911088,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.781840219047541e-05,
+      "loss": 0.586,
+      "step": 646
+    },
+    {
+      "epoch": 1.7655338139375214,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.7786386860170054e-05,
+      "loss": 0.5291,
+      "step": 647
+    },
+    {
+      "epoch": 1.768280123583934,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.775435837748306e-05,
+      "loss": 0.3863,
+      "step": 648
+    },
+    {
+      "epoch": 1.7710264332303467,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7722316893596176e-05,
+      "loss": 0.5247,
+      "step": 649
+    },
+    {
+      "epoch": 1.7737727428767593,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.7690262559752516e-05,
+      "loss": 0.4046,
+      "step": 650
+    },
+    {
+      "epoch": 1.776519052523172,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.7658195527255847e-05,
+      "loss": 0.4744,
+      "step": 651
+    },
+    {
+      "epoch": 1.7792653621695846,
+      "grad_norm": 0.0625,
+      "learning_rate": 1.7626115947469877e-05,
+      "loss": 0.424,
+      "step": 652
+    },
+    {
+      "epoch": 1.7820116718159973,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.759402397181754e-05,
+      "loss": 0.4644,
+      "step": 653
+    },
+    {
+      "epoch": 1.78475798146241,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.7561919751780278e-05,
+      "loss": 0.6509,
+      "step": 654
+    },
+    {
+      "epoch": 1.7875042911088226,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.7529803438897346e-05,
+      "loss": 0.4544,
+      "step": 655
+    },
+    {
+      "epoch": 1.7902506007552352,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.7497675184765064e-05,
+      "loss": 0.4991,
+      "step": 656
+    },
+    {
+      "epoch": 1.7929969104016479,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.746553514103611e-05,
+      "loss": 0.5494,
+      "step": 657
+    },
+    {
+      "epoch": 1.7957432200480605,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.743338345941883e-05,
+      "loss": 0.4772,
+      "step": 658
+    },
+    {
+      "epoch": 1.7984895296944732,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.74012202916765e-05,
+      "loss": 0.5995,
+      "step": 659
+    },
+    {
+      "epoch": 1.8012358393408858,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.7369045789626603e-05,
+      "loss": 0.5156,
+      "step": 660
+    },
+    {
+      "epoch": 1.8039821489872985,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.7336860105140134e-05,
+      "loss": 0.3329,
+      "step": 661
+    },
+    {
+      "epoch": 1.806728458633711,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.730466339014086e-05,
+      "loss": 0.4797,
+      "step": 662
+    },
+    {
+      "epoch": 1.8094747682801235,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.7272455796604622e-05,
+      "loss": 0.4494,
+      "step": 663
+    },
+    {
+      "epoch": 1.8122210779265362,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.7240237476558615e-05,
+      "loss": 0.5881,
+      "step": 664
+    },
+    {
+      "epoch": 1.8149673875729488,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.7208008582080652e-05,
+      "loss": 0.451,
+      "step": 665
+    },
+    {
+      "epoch": 1.8177136972193615,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.7175769265298472e-05,
+      "loss": 0.3846,
+      "step": 666
+    },
+    {
+      "epoch": 1.8204600068657741,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.7143519678389004e-05,
+      "loss": 0.4766,
+      "step": 667
+    },
+    {
+      "epoch": 1.8232063165121868,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.7111259973577655e-05,
+      "loss": 0.4932,
+      "step": 668
+    },
+    {
+      "epoch": 1.8259526261585994,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.7078990303137584e-05,
+      "loss": 0.4978,
+      "step": 669
+    },
+    {
+      "epoch": 1.8286989358050119,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.7046710819389012e-05,
+      "loss": 0.5164,
+      "step": 670
+    },
+    {
+      "epoch": 1.8314452454514245,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.7014421674698458e-05,
+      "loss": 0.5542,
+      "step": 671
+    },
+    {
+      "epoch": 1.8341915550978372,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.6982123021478046e-05,
+      "loss": 0.3729,
+      "step": 672
+    },
+    {
+      "epoch": 1.8369378647442498,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.6949815012184795e-05,
+      "loss": 0.4723,
+      "step": 673
+    },
+    {
+      "epoch": 1.8396841743906625,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.6917497799319876e-05,
+      "loss": 0.5643,
+      "step": 674
+    },
+    {
+      "epoch": 1.842430484037075,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.6885171535427913e-05,
+      "loss": 0.4695,
+      "step": 675
+    },
+    {
+      "epoch": 1.8451767936834877,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.685283637309623e-05,
+      "loss": 0.4316,
+      "step": 676
+    },
+    {
+      "epoch": 1.8479231033299004,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.6820492464954187e-05,
+      "loss": 0.4624,
+      "step": 677
+    },
+    {
+      "epoch": 1.850669412976313,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.67881399636724e-05,
+      "loss": 0.4515,
+      "step": 678
+    },
+    {
+      "epoch": 1.8534157226227257,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.6755779021962056e-05,
+      "loss": 0.5498,
+      "step": 679
+    },
+    {
+      "epoch": 1.8561620322691383,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.6723409792574185e-05,
+      "loss": 0.4184,
+      "step": 680
+    },
+    {
+      "epoch": 1.858908341915551,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.6691032428298934e-05,
+      "loss": 0.437,
+      "step": 681
+    },
+    {
+      "epoch": 1.8616546515619636,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.665864708196485e-05,
+      "loss": 0.5498,
+      "step": 682
+    },
+    {
+      "epoch": 1.8644009612083763,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.6626253906438148e-05,
+      "loss": 0.4403,
+      "step": 683
+    },
+    {
+      "epoch": 1.867147270854789,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.6593853054622016e-05,
+      "loss": 0.5116,
+      "step": 684
+    },
+    {
+      "epoch": 1.8698935805012016,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.6561444679455858e-05,
+      "loss": 0.4179,
+      "step": 685
+    },
+    {
+      "epoch": 1.8726398901476142,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.6529028933914604e-05,
+      "loss": 0.4291,
+      "step": 686
+    },
+    {
+      "epoch": 1.8753861997940269,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.649660597100797e-05,
+      "loss": 0.4856,
+      "step": 687
+    },
+    {
+      "epoch": 1.8781325094404395,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.646417594377973e-05,
+      "loss": 0.5419,
+      "step": 688
+    },
+    {
+      "epoch": 1.8808788190868522,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6431739005307014e-05,
+      "loss": 0.4287,
+      "step": 689
+    },
+    {
+      "epoch": 1.8836251287332648,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.6399295308699572e-05,
+      "loss": 0.4848,
+      "step": 690
+    },
+    {
+      "epoch": 1.8863714383796775,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.636684500709905e-05,
+      "loss": 0.3635,
+      "step": 691
+    },
+    {
+      "epoch": 1.88911774802609,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.6334388253678285e-05,
+      "loss": 0.5319,
+      "step": 692
+    },
+    {
+      "epoch": 1.8918640576725025,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.6301925201640542e-05,
+      "loss": 0.5852,
+      "step": 693
+    },
+    {
+      "epoch": 1.8946103673189152,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.6269456004218844e-05,
+      "loss": 0.5184,
+      "step": 694
+    },
+    {
+      "epoch": 1.8973566769653278,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.6236980814675204e-05,
+      "loss": 0.4528,
+      "step": 695
+    },
+    {
+      "epoch": 1.9001029866117405,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.620449978629993e-05,
+      "loss": 0.4608,
+      "step": 696
+    },
+    {
+      "epoch": 1.9028492962581531,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.617201307241088e-05,
+      "loss": 0.5007,
+      "step": 697
+    },
+    {
+      "epoch": 1.9055956059045658,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.6139520826352765e-05,
+      "loss": 0.5226,
+      "step": 698
+    },
+    {
+      "epoch": 1.9083419155509782,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6107023201496378e-05,
+      "loss": 0.4345,
+      "step": 699
+    },
+    {
+      "epoch": 1.9110882251973909,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.6074520351237947e-05,
+      "loss": 0.4386,
+      "step": 700
+    },
+    {
+      "epoch": 1.9138345348438035,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.6042012428998325e-05,
+      "loss": 0.4791,
+      "step": 701
+    },
+    {
+      "epoch": 1.9165808444902162,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.6009499588222325e-05,
+      "loss": 0.3982,
+      "step": 702
+    },
+    {
+      "epoch": 1.9193271541366288,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.597698198237797e-05,
+      "loss": 0.3487,
+      "step": 703
+    },
+    {
+      "epoch": 1.9220734637830414,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.5944459764955784e-05,
+      "loss": 0.3082,
+      "step": 704
+    },
+    {
+      "epoch": 1.924819773429454,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.5911933089468048e-05,
+      "loss": 0.4835,
+      "step": 705
+    },
+    {
+      "epoch": 1.9275660830758667,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.5879402109448093e-05,
+      "loss": 0.503,
+      "step": 706
+    },
+    {
+      "epoch": 1.9303123927222794,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.584686697844956e-05,
+      "loss": 0.5597,
+      "step": 707
+    },
+    {
+      "epoch": 1.933058702368692,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.5814327850045697e-05,
+      "loss": 0.5074,
+      "step": 708
+    },
+    {
+      "epoch": 1.9358050120151047,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.5781784877828607e-05,
+      "loss": 0.6022,
+      "step": 709
+    },
+    {
+      "epoch": 1.9385513216615173,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.5749238215408548e-05,
+      "loss": 0.5197,
+      "step": 710
+    },
+    {
+      "epoch": 1.94129763130793,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.571668801641319e-05,
+      "loss": 0.5147,
+      "step": 711
+    },
+    {
+      "epoch": 1.9440439409543426,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.5684134434486893e-05,
+      "loss": 0.5506,
+      "step": 712
+    },
+    {
+      "epoch": 1.9467902506007553,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.565157762329e-05,
+      "loss": 0.4598,
+      "step": 713
+    },
+    {
+      "epoch": 1.949536560247168,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.5619017736498076e-05,
+      "loss": 0.5802,
+      "step": 714
+    },
+    {
+      "epoch": 1.9522828698935806,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.5586454927801223e-05,
+      "loss": 0.567,
+      "step": 715
+    },
+    {
+      "epoch": 1.9550291795399932,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.555388935090332e-05,
+      "loss": 0.5956,
+      "step": 716
+    },
+    {
+      "epoch": 1.9577754891864059,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.5521321159521326e-05,
+      "loss": 0.4019,
+      "step": 717
+    },
+    {
+      "epoch": 1.9605217988328185,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.548875050738453e-05,
+      "loss": 0.4996,
+      "step": 718
+    },
+    {
+      "epoch": 1.9632681084792312,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.545617754823384e-05,
+      "loss": 0.4999,
+      "step": 719
+    },
+    {
+      "epoch": 1.9660144181256438,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.5423602435821055e-05,
+      "loss": 0.6049,
+      "step": 720
+    },
+    {
+      "epoch": 1.9687607277720565,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.5391025323908134e-05,
+      "loss": 0.4799,
+      "step": 721
+    },
+    {
+      "epoch": 1.9715070374184689,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.5358446366266483e-05,
+      "loss": 0.4836,
+      "step": 722
+    },
+    {
+      "epoch": 1.9742533470648815,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.532586571667621e-05,
+      "loss": 1.1204,
+      "step": 723
+    },
+    {
+      "epoch": 1.9769996567112942,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.5293283528925412e-05,
+      "loss": 0.4647,
+      "step": 724
+    },
+    {
+      "epoch": 1.9797459663577068,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.5260699956809456e-05,
+      "loss": 0.4984,
+      "step": 725
+    },
+    {
+      "epoch": 1.9824922760041195,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.522811515413023e-05,
+      "loss": 0.4019,
+      "step": 726
+    },
+    {
+      "epoch": 1.9852385856505321,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.5195529274695436e-05,
+      "loss": 0.4028,
+      "step": 727
+    },
+    {
+      "epoch": 1.9879848952969448,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.5162942472317858e-05,
+      "loss": 0.5023,
+      "step": 728
+    },
+    {
+      "epoch": 1.9879848952969448,
+      "eval_loss": 0.5045637488365173,
+      "eval_runtime": 616.8097,
+      "eval_samples_per_second": 14.86,
+      "eval_steps_per_second": 14.86,
+      "step": 728
+    },
+    {
+      "epoch": 1.9907312049433572,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.5130354900814643e-05,
+      "loss": 0.4917,
+      "step": 729
+    },
+    {
+      "epoch": 1.9934775145897699,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.5097766714006553e-05,
+      "loss": 0.4892,
+      "step": 730
+    },
+    {
+      "epoch": 1.9962238242361825,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.5065178065717274e-05,
+      "loss": 0.4261,
+      "step": 731
+    },
+    {
+      "epoch": 1.9989701338825951,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.5032589109772655e-05,
+      "loss": 0.5345,
+      "step": 732
+    },
+    {
+      "epoch": 2.001716443529008,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.5e-05,
+      "loss": 0.5285,
+      "step": 733
+    },
+    {
+      "epoch": 2.0044627531754204,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.4967410890227347e-05,
+      "loss": 0.4216,
+      "step": 734
+    },
+    {
+      "epoch": 2.007209062821833,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.4934821934282728e-05,
+      "loss": 0.5005,
+      "step": 735
+    },
+    {
+      "epoch": 2.0099553724682457,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.4902233285993447e-05,
+      "loss": 0.495,
+      "step": 736
+    },
+    {
+      "epoch": 2.0127016821146584,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.4869645099185361e-05,
+      "loss": 0.4652,
+      "step": 737
+    },
+    {
+      "epoch": 2.0013731548232063,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.4837057527682142e-05,
+      "loss": 0.621,
+      "step": 738
+    },
+    {
+      "epoch": 2.004119464469619,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.4804470725304567e-05,
+      "loss": 0.353,
+      "step": 739
+    },
+    {
+      "epoch": 2.0068657741160316,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.4771884845869772e-05,
+      "loss": 0.431,
+      "step": 740
+    },
+    {
+      "epoch": 2.0096120837624443,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.4739300043190547e-05,
+      "loss": 0.5136,
+      "step": 741
+    },
+    {
+      "epoch": 2.012358393408857,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.470671647107459e-05,
+      "loss": 0.3657,
+      "step": 742
+    },
+    {
+      "epoch": 2.0151047030552696,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.4674134283323792e-05,
+      "loss": 0.5771,
+      "step": 743
+    },
+    {
+      "epoch": 2.017851012701682,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.4641553633733519e-05,
+      "loss": 0.3684,
+      "step": 744
+    },
+    {
+      "epoch": 2.020597322348095,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.460897467609187e-05,
+      "loss": 0.5238,
+      "step": 745
+    },
+    {
+      "epoch": 2.0233436319945075,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 1.4576397564178951e-05,
+      "loss": 0.451,
+      "step": 746
+    },
+    {
+      "epoch": 2.02608994164092,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.4543822451766166e-05,
+      "loss": 0.5708,
+      "step": 747
+    },
+    {
+      "epoch": 2.028836251287333,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.4511249492615477e-05,
+      "loss": 0.6172,
+      "step": 748
+    },
+    {
+      "epoch": 2.0315825609337455,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.447867884047868e-05,
+      "loss": 0.6446,
+      "step": 749
+    },
+    {
+      "epoch": 2.034328870580158,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.4446110649096683e-05,
+      "loss": 0.5011,
+      "step": 750
+    },
+    {
+      "epoch": 2.0370751802265707,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.4413545072198783e-05,
+      "loss": 0.4862,
+      "step": 751
+    },
+    {
+      "epoch": 2.0398214898729834,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.438098226350193e-05,
+      "loss": 0.4955,
+      "step": 752
+    },
+    {
+      "epoch": 2.0425677995193956,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.4348422376710009e-05,
+      "loss": 0.3889,
+      "step": 753
+    },
+    {
+      "epoch": 2.0453141091658082,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.4315865565513111e-05,
+      "loss": 1.1553,
+      "step": 754
+    },
+    {
+      "epoch": 2.048060418812221,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.4283311983586818e-05,
+      "loss": 0.5192,
+      "step": 755
+    },
+    {
+      "epoch": 2.0508067284586335,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.4250761784591451e-05,
+      "loss": 0.4926,
+      "step": 756
+    },
+    {
+      "epoch": 2.053553038105046,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.4218215122171392e-05,
+      "loss": 0.5362,
+      "step": 757
+    },
+    {
+      "epoch": 2.056299347751459,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.4185672149954304e-05,
+      "loss": 0.4926,
+      "step": 758
+    },
+    {
+      "epoch": 2.0590456573978715,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.4153133021550438e-05,
+      "loss": 0.6137,
+      "step": 759
+    },
+    {
+      "epoch": 2.061791967044284,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.4120597890551908e-05,
+      "loss": 0.4648,
+      "step": 760
+    },
+    {
+      "epoch": 2.064538276690697,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.4088066910531951e-05,
+      "loss": 0.6486,
+      "step": 761
+    },
+    {
+      "epoch": 2.0672845863371094,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.4055540235044213e-05,
+      "loss": 0.4291,
+      "step": 762
+    },
+    {
+      "epoch": 2.070030895983522,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.402301801762203e-05,
+      "loss": 0.515,
+      "step": 763
+    },
+    {
+      "epoch": 2.0727772056299347,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.3990500411777677e-05,
+      "loss": 0.6079,
+      "step": 764
+    },
+    {
+      "epoch": 2.0755235152763474,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.3957987571001676e-05,
+      "loss": 0.5589,
+      "step": 765
+    },
+    {
+      "epoch": 2.07826982492276,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.3925479648762055e-05,
+      "loss": 0.6439,
+      "step": 766
+    },
+    {
+      "epoch": 2.0810161345691727,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.3892976798503621e-05,
+      "loss": 0.4723,
+      "step": 767
+    },
+    {
+      "epoch": 2.0837624442155853,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.3860479173647241e-05,
+      "loss": 0.5328,
+      "step": 768
+    },
+    {
+      "epoch": 2.086508753861998,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.3827986927589118e-05,
+      "loss": 0.5182,
+      "step": 769
+    },
+    {
+      "epoch": 2.0892550635084106,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.3795500213700072e-05,
+      "loss": 0.4433,
+      "step": 770
+    },
+    {
+      "epoch": 2.0920013731548233,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.3763019185324797e-05,
+      "loss": 0.483,
+      "step": 771
+    },
+    {
+      "epoch": 2.094747682801236,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.3730543995781158e-05,
+      "loss": 0.4826,
+      "step": 772
+    },
+    {
+      "epoch": 2.0974939924476486,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.3698074798359458e-05,
+      "loss": 0.5313,
+      "step": 773
+    },
+    {
+      "epoch": 2.100240302094061,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.3665611746321718e-05,
+      "loss": 0.4303,
+      "step": 774
+    },
+    {
+      "epoch": 2.102986611740474,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.363315499290095e-05,
+      "loss": 0.5252,
+      "step": 775
+    },
+    {
+      "epoch": 2.1057329213868865,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.360070469130043e-05,
+      "loss": 0.4501,
+      "step": 776
+    },
+    {
+      "epoch": 2.108479231033299,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.3568260994692988e-05,
+      "loss": 0.4423,
+      "step": 777
+    },
+    {
+      "epoch": 2.111225540679712,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.3535824056220273e-05,
+      "loss": 0.5341,
+      "step": 778
+    },
+    {
+      "epoch": 2.1139718503261244,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.3503394028992032e-05,
+      "loss": 0.4019,
+      "step": 779
+    },
+    {
+      "epoch": 2.116718159972537,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.3470971066085395e-05,
+      "loss": 0.5329,
+      "step": 780
+    },
+    {
+      "epoch": 2.1194644696189497,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.3438555320544143e-05,
+      "loss": 0.5412,
+      "step": 781
+    },
+    {
+      "epoch": 2.1222107792653624,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.3406146945377987e-05,
+      "loss": 0.4902,
+      "step": 782
+    },
+    {
+      "epoch": 2.1249570889117746,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.3373746093561855e-05,
+      "loss": 0.6356,
+      "step": 783
+    },
+    {
+      "epoch": 2.1277033985581872,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.3341352918035156e-05,
+      "loss": 0.4674,
+      "step": 784
+    },
+    {
+      "epoch": 2.1304497082046,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.330896757170107e-05,
+      "loss": 0.5155,
+      "step": 785
+    },
+    {
+      "epoch": 2.1331960178510125,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.327659020742582e-05,
+      "loss": 0.5342,
+      "step": 786
+    },
+    {
+      "epoch": 2.135942327497425,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.3244220978037945e-05,
+      "loss": 0.5219,
+      "step": 787
+    },
+    {
+      "epoch": 2.138688637143838,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.3211860036327604e-05,
+      "loss": 0.5404,
+      "step": 788
+    },
+    {
+      "epoch": 2.1414349467902505,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.3179507535045819e-05,
+      "loss": 0.4683,
+      "step": 789
+    },
+    {
+      "epoch": 2.144181256436663,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.3147163626903774e-05,
+      "loss": 0.4784,
+      "step": 790
+    },
+    {
+      "epoch": 2.1469275660830758,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.3114828464572096e-05,
+      "loss": 0.5399,
+      "step": 791
+    },
+    {
+      "epoch": 2.1496738757294884,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.3082502200680128e-05,
+      "loss": 0.425,
+      "step": 792
+    },
+    {
+      "epoch": 2.152420185375901,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.305018498781521e-05,
+      "loss": 0.4779,
+      "step": 793
+    },
+    {
+      "epoch": 2.1551664950223137,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.301787697852196e-05,
+      "loss": 0.5025,
+      "step": 794
+    },
+    {
+      "epoch": 2.1579128046687264,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.298557832530155e-05,
+      "loss": 0.5144,
+      "step": 795
+    },
+    {
+      "epoch": 2.160659114315139,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.2953289180610994e-05,
+      "loss": 0.4723,
+      "step": 796
+    },
+    {
+      "epoch": 2.1634054239615517,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.2921009696862419e-05,
+      "loss": 0.5309,
+      "step": 797
+    },
+    {
+      "epoch": 2.1661517336079643,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.2888740026422354e-05,
+      "loss": 0.3767,
+      "step": 798
+    },
+    {
+      "epoch": 2.168898043254377,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.2856480321611004e-05,
+      "loss": 0.559,
+      "step": 799
+    },
+    {
+      "epoch": 2.1716443529007896,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.2824230734701535e-05,
+      "loss": 0.565,
+      "step": 800
+    },
+    {
+      "epoch": 2.1743906625472023,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.2791991417919347e-05,
+      "loss": 0.5005,
+      "step": 801
+    },
+    {
+      "epoch": 2.177136972193615,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.2759762523441386e-05,
+      "loss": 0.4312,
+      "step": 802
+    },
+    {
+      "epoch": 2.1798832818400276,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.2727544203395377e-05,
+      "loss": 0.5526,
+      "step": 803
+    },
+    {
+      "epoch": 2.18262959148644,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.269533660985914e-05,
+      "loss": 0.4463,
+      "step": 804
+    },
+    {
+      "epoch": 2.185375901132853,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.2663139894859867e-05,
+      "loss": 0.4219,
+      "step": 805
+    },
+    {
+      "epoch": 2.1881222107792655,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.2630954210373396e-05,
+      "loss": 0.3865,
+      "step": 806
+    },
+    {
+      "epoch": 2.190868520425678,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.2598779708323499e-05,
+      "loss": 0.5792,
+      "step": 807
+    },
+    {
+      "epoch": 2.193614830072091,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.2566616540581168e-05,
+      "loss": 0.462,
+      "step": 808
+    },
+    {
+      "epoch": 2.1963611397185034,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.2534464858963892e-05,
+      "loss": 0.4869,
+      "step": 809
+    },
+    {
+      "epoch": 2.199107449364916,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.2502324815234942e-05,
+      "loss": 0.5559,
+      "step": 810
+    },
+    {
+      "epoch": 2.2018537590113283,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.2470196561102655e-05,
+      "loss": 0.3752,
+      "step": 811
+    },
+    {
+      "epoch": 2.204600068657741,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.2438080248219723e-05,
+      "loss": 0.5054,
+      "step": 812
+    },
+    {
+      "epoch": 2.2073463783041536,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.2405976028182464e-05,
+      "loss": 0.3525,
+      "step": 813
+    },
+    {
+      "epoch": 2.2100926879505662,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.2373884052530127e-05,
+      "loss": 0.4951,
+      "step": 814
+    },
+    {
+      "epoch": 2.212838997596979,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.2341804472744157e-05,
+      "loss": 0.4484,
+      "step": 815
+    },
+    {
+      "epoch": 2.2155853072433915,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.2309737440247486e-05,
+      "loss": 0.5412,
+      "step": 816
+    },
+    {
+      "epoch": 2.218331616889804,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.2277683106403826e-05,
+      "loss": 0.6162,
+      "step": 817
+    },
+    {
+      "epoch": 2.221077926536217,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.2245641622516943e-05,
+      "loss": 0.4606,
+      "step": 818
+    },
+    {
+      "epoch": 2.2238242361826295,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.2213613139829949e-05,
+      "loss": 0.3737,
+      "step": 819
+    },
+    {
+      "epoch": 2.2238242361826295,
+      "eval_loss": 0.504136323928833,
+      "eval_runtime": 615.1173,
+      "eval_samples_per_second": 14.901,
+      "eval_steps_per_second": 14.901,
+      "step": 819
+    },
+    {
+      "epoch": 2.226570545829042,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.2181597809524594e-05,
+      "loss": 0.3953,
+      "step": 820
+    },
+    {
+      "epoch": 2.2293168554754548,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.2149595782720537e-05,
+      "loss": 0.4174,
+      "step": 821
+    },
+    {
+      "epoch": 2.2320631651218674,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.2117607210474645e-05,
+      "loss": 0.5269,
+      "step": 822
+    },
+    {
+      "epoch": 2.23480947476828,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.2085632243780278e-05,
+      "loss": 0.4668,
+      "step": 823
+    },
+    {
+      "epoch": 2.2375557844146927,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.205367103356657e-05,
+      "loss": 0.4565,
+      "step": 824
+    },
+    {
+      "epoch": 2.2403020940611054,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.202172373069773e-05,
+      "loss": 0.4427,
+      "step": 825
+    },
+    {
+      "epoch": 2.243048403707518,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 1.1989790485972312e-05,
+      "loss": 0.4414,
+      "step": 826
+    },
+    {
+      "epoch": 2.2457947133539307,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.1957871450122516e-05,
+      "loss": 0.5547,
+      "step": 827
+    },
+    {
+      "epoch": 2.2485410230003433,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.1925966773813476e-05,
+      "loss": 0.5273,
+      "step": 828
+    },
+    {
+      "epoch": 2.251287332646756,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.1894076607642537e-05,
+      "loss": 0.5066,
+      "step": 829
+    },
+    {
+      "epoch": 2.2540336422931686,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.1862201102138562e-05,
+      "loss": 0.5397,
+      "step": 830
+    },
+    {
+      "epoch": 2.2567799519395813,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.1830340407761207e-05,
+      "loss": 0.4944,
+      "step": 831
+    },
+    {
+      "epoch": 2.259526261585994,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.1798494674900222e-05,
+      "loss": 0.4056,
+      "step": 832
+    },
+    {
+      "epoch": 2.2622725712324065,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.1766664053874726e-05,
+      "loss": 0.5453,
+      "step": 833
+    },
+    {
+      "epoch": 2.265018880878819,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.1734848694932514e-05,
+      "loss": 0.456,
+      "step": 834
+    },
+    {
+      "epoch": 2.267765190525232,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.170304874824934e-05,
+      "loss": 0.5696,
+      "step": 835
+    },
+    {
+      "epoch": 2.2705115001716445,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.1671264363928205e-05,
+      "loss": 0.4873,
+      "step": 836
+    },
+    {
+      "epoch": 2.273257809818057,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.1639495691998653e-05,
+      "loss": 0.5142,
+      "step": 837
+    },
+    {
+      "epoch": 2.27600411946447,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.1607742882416064e-05,
+      "loss": 0.4905,
+      "step": 838
+    },
+    {
+      "epoch": 2.2787504291108824,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.1576006085060941e-05,
+      "loss": 0.4352,
+      "step": 839
+    },
+    {
+      "epoch": 2.281496738757295,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.1544285449738211e-05,
+      "loss": 0.5675,
+      "step": 840
+    },
+    {
+      "epoch": 2.2842430484037077,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.1512581126176508e-05,
+      "loss": 0.4553,
+      "step": 841
+    },
+    {
+      "epoch": 2.2869893580501204,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.1480893264027469e-05,
+      "loss": 0.5391,
+      "step": 842
+    },
+    {
+      "epoch": 2.289735667696533,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.1449222012865037e-05,
+      "loss": 0.5003,
+      "step": 843
+    },
+    {
+      "epoch": 2.2924819773429452,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.1417567522184738e-05,
+      "loss": 0.5302,
+      "step": 844
+    },
+    {
+      "epoch": 2.295228286989358,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.1385929941402993e-05,
+      "loss": 0.571,
+      "step": 845
+    },
+    {
+      "epoch": 2.2979745966357705,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.1354309419856392e-05,
+      "loss": 0.5726,
+      "step": 846
+    },
+    {
+      "epoch": 2.300720906282183,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.1322706106801025e-05,
+      "loss": 0.5884,
+      "step": 847
+    },
+    {
+      "epoch": 2.303467215928596,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.1291120151411731e-05,
+      "loss": 0.5926,
+      "step": 848
+    },
+    {
+      "epoch": 2.3062135255750085,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.1259551702781426e-05,
+      "loss": 0.4487,
+      "step": 849
+    },
+    {
+      "epoch": 2.308959835221421,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.1228000909920388e-05,
+      "loss": 0.3924,
+      "step": 850
+    },
+    {
+      "epoch": 2.3117061448678338,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.119646792175556e-05,
+      "loss": 0.4218,
+      "step": 851
+    },
+    {
+      "epoch": 2.3144524545142464,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.1164952887129836e-05,
+      "loss": 1.1613,
+      "step": 852
+    },
+    {
+      "epoch": 2.317198764160659,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.1133455954801372e-05,
+      "loss": 0.4224,
+      "step": 853
+    },
+    {
+      "epoch": 2.3199450738070717,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.1101977273442873e-05,
+      "loss": 0.4405,
+      "step": 854
+    },
+    {
+      "epoch": 2.3226913834534844,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.1070516991640894e-05,
+      "loss": 0.4972,
+      "step": 855
+    },
+    {
+      "epoch": 2.325437693099897,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.1039075257895146e-05,
+      "loss": 0.5403,
+      "step": 856
+    },
+    {
+      "epoch": 2.3281840027463097,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.1007652220617778e-05,
+      "loss": 0.5295,
+      "step": 857
+    },
+    {
+      "epoch": 2.3309303123927223,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.0976248028132705e-05,
+      "loss": 0.5899,
+      "step": 858
+    },
+    {
+      "epoch": 2.333676622039135,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.0944862828674872e-05,
+      "loss": 0.4907,
+      "step": 859
+    },
+    {
+      "epoch": 2.3364229316855476,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.0913496770389585e-05,
+      "loss": 0.5142,
+      "step": 860
+    },
+    {
+      "epoch": 2.3391692413319602,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.088215000133179e-05,
+      "loss": 0.5103,
+      "step": 861
+    },
+    {
+      "epoch": 2.341915550978373,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.0850822669465392e-05,
+      "loss": 0.5814,
+      "step": 862
+    },
+    {
+      "epoch": 2.3446618606247855,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.081951492266254e-05,
+      "loss": 0.5544,
+      "step": 863
+    },
+    {
+      "epoch": 2.347408170271198,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.0788226908702945e-05,
+      "loss": 1.1435,
+      "step": 864
+    },
+    {
+      "epoch": 2.350154479917611,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.0756958775273169e-05,
+      "loss": 0.4895,
+      "step": 865
+    },
+    {
+      "epoch": 2.3529007895640235,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.0725710669965936e-05,
+      "loss": 0.5886,
+      "step": 866
+    },
+    {
+      "epoch": 2.355647099210436,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.0694482740279428e-05,
+      "loss": 0.4469,
+      "step": 867
+    },
+    {
+      "epoch": 2.358393408856849,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.0663275133616603e-05,
+      "loss": 0.4049,
+      "step": 868
+    },
+    {
+      "epoch": 2.3611397185032614,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.063208799728448e-05,
+      "loss": 0.3659,
+      "step": 869
+    },
+    {
+      "epoch": 2.3638860281496736,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.0600921478493455e-05,
+      "loss": 0.5023,
+      "step": 870
+    },
+    {
+      "epoch": 2.3666323377960863,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.0569775724356611e-05,
+      "loss": 0.4065,
+      "step": 871
+    },
+    {
+      "epoch": 2.369378647442499,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.0538650881889013e-05,
+      "loss": 0.4033,
+      "step": 872
+    },
+    {
+      "epoch": 2.3721249570889116,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.0507547098007015e-05,
+      "loss": 0.4139,
+      "step": 873
+    },
+    {
+      "epoch": 2.3748712667353242,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.0476464519527574e-05,
+      "loss": 0.5499,
+      "step": 874
+    },
+    {
+      "epoch": 2.377617576381737,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.0445403293167547e-05,
+      "loss": 0.4546,
+      "step": 875
+    },
+    {
+      "epoch": 2.3803638860281495,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.0414363565543016e-05,
+      "loss": 0.4263,
+      "step": 876
+    },
+    {
+      "epoch": 2.383110195674562,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.0383345483168573e-05,
+      "loss": 0.446,
+      "step": 877
+    },
+    {
+      "epoch": 2.385856505320975,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.0352349192456643e-05,
+      "loss": 0.445,
+      "step": 878
+    },
+    {
+      "epoch": 2.3886028149673875,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.032137483971679e-05,
+      "loss": 0.5788,
+      "step": 879
+    },
+    {
+      "epoch": 2.3913491246138,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.0290422571155024e-05,
+      "loss": 0.4657,
+      "step": 880
+    },
+    {
+      "epoch": 2.3940954342602128,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.0259492532873113e-05,
+      "loss": 0.5917,
+      "step": 881
+    },
+    {
+      "epoch": 2.3968417439066254,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.0228584870867896e-05,
+      "loss": 0.4403,
+      "step": 882
+    },
+    {
+      "epoch": 2.399588053553038,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.0197699731030584e-05,
+      "loss": 0.4274,
+      "step": 883
+    },
+    {
+      "epoch": 2.4023343631994507,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.016683725914609e-05,
+      "loss": 0.5997,
+      "step": 884
+    },
+    {
+      "epoch": 2.4050806728458634,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.0135997600892316e-05,
+      "loss": 0.4612,
+      "step": 885
+    },
+    {
+      "epoch": 2.407826982492276,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.0105180901839487e-05,
+      "loss": 0.4969,
+      "step": 886
+    },
+    {
+      "epoch": 2.4105732921386887,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.0074387307449452e-05,
+      "loss": 0.4989,
+      "step": 887
+    },
+    {
+      "epoch": 2.4133196017851013,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0043616963075001e-05,
+      "loss": 0.4879,
+      "step": 888
+    },
+    {
+      "epoch": 2.416065911431514,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.0012870013959182e-05,
+      "loss": 0.4521,
+      "step": 889
+    },
+    {
+      "epoch": 2.4188122210779266,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.982146605234604e-06,
+      "loss": 0.6295,
+      "step": 890
+    },
+    {
+      "epoch": 2.4215585307243392,
+      "grad_norm": 0.044921875,
+      "learning_rate": 9.95144688192277e-06,
+      "loss": 0.4266,
+      "step": 891
+    },
+    {
+      "epoch": 2.424304840370752,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 9.920770988933366e-06,
+      "loss": 0.3733,
+      "step": 892
+    },
+    {
+      "epoch": 2.4270511500171645,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 9.890119071063624e-06,
+      "loss": 0.5311,
+      "step": 893
+    },
+    {
+      "epoch": 2.429797459663577,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 9.859491272997579e-06,
+      "loss": 0.3879,
+      "step": 894
+    },
+    {
+      "epoch": 2.43254376930999,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.828887739305423e-06,
+      "loss": 0.4734,
+      "step": 895
+    },
+    {
+      "epoch": 2.4352900789564025,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 9.798308614442822e-06,
+      "loss": 0.4965,
+      "step": 896
+    },
+    {
+      "epoch": 2.438036388602815,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 9.767754042750214e-06,
+      "loss": 0.502,
+      "step": 897
+    },
+    {
+      "epoch": 2.4407826982492278,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 9.737224168452154e-06,
+      "loss": 0.4924,
+      "step": 898
+    },
+    {
+      "epoch": 2.4435290078956404,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.706719135656613e-06,
+      "loss": 0.5116,
+      "step": 899
+    },
+    {
+      "epoch": 2.446275317542053,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.676239088354302e-06,
+      "loss": 0.5165,
+      "step": 900
+    },
+    {
+      "epoch": 2.4490216271884657,
+      "grad_norm": 0.04296875,
+      "learning_rate": 9.645784170417996e-06,
+      "loss": 0.4293,
+      "step": 901
+    },
+    {
+      "epoch": 2.4517679368348784,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 9.615354525601859e-06,
+      "loss": 0.5352,
+      "step": 902
+    },
+    {
+      "epoch": 2.4545142464812906,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 9.584950297540759e-06,
+      "loss": 0.474,
+      "step": 903
+    },
+    {
+      "epoch": 2.457260556127703,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 9.554571629749585e-06,
+      "loss": 0.495,
+      "step": 904
+    },
+    {
+      "epoch": 2.460006865774116,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.524218665622578e-06,
+      "loss": 0.4432,
+      "step": 905
+    },
+    {
+      "epoch": 2.4627531754205285,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 9.493891548432654e-06,
+      "loss": 0.4222,
+      "step": 906
+    },
+    {
+      "epoch": 2.465499485066941,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.463590421330727e-06,
+      "loss": 0.5396,
+      "step": 907
+    },
+    {
+      "epoch": 2.468245794713354,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.433315427345028e-06,
+      "loss": 0.4899,
+      "step": 908
+    },
+    {
+      "epoch": 2.4709921043597665,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 9.403066709380432e-06,
+      "loss": 0.6021,
+      "step": 909
+    },
+    {
+      "epoch": 2.473738414006179,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.372844410217792e-06,
+      "loss": 0.505,
+      "step": 910
+    },
+    {
+      "epoch": 2.473738414006179,
+      "eval_loss": 0.5039077997207642,
+      "eval_runtime": 617.6957,
+      "eval_samples_per_second": 14.839,
+      "eval_steps_per_second": 14.839,
+      "step": 910
+    },
+    {
+      "epoch": 2.4764847236525918,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 9.342648672513254e-06,
+      "loss": 0.5927,
+      "step": 911
+    },
+    {
+      "epoch": 2.4792310332990044,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.31247963879759e-06,
+      "loss": 0.4284,
+      "step": 912
+    },
+    {
+      "epoch": 2.481977342945417,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 9.28233745147552e-06,
+      "loss": 0.4401,
+      "step": 913
+    },
+    {
+      "epoch": 2.4847236525918297,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.252222252825043e-06,
+      "loss": 0.5268,
+      "step": 914
+    },
+    {
+      "epoch": 2.4874699622382423,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 9.222134184996769e-06,
+      "loss": 0.5029,
+      "step": 915
+    },
+    {
+      "epoch": 2.490216271884655,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 9.19207339001324e-06,
+      "loss": 0.4947,
+      "step": 916
+    },
+    {
+      "epoch": 2.4929625815310676,
+      "grad_norm": 0.048828125,
+      "learning_rate": 9.16204000976827e-06,
+      "loss": 0.5041,
+      "step": 917
+    },
+    {
+      "epoch": 2.4957088911774803,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 9.13203418602626e-06,
+      "loss": 0.4975,
+      "step": 918
+    },
+    {
+      "epoch": 2.498455200823893,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.102056060421545e-06,
+      "loss": 0.4445,
+      "step": 919
+    },
+    {
+      "epoch": 2.5012015104703056,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.07210577445772e-06,
+      "loss": 0.4441,
+      "step": 920
+    },
+    {
+      "epoch": 2.5039478201167182,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 9.042183469506964e-06,
+      "loss": 0.5294,
+      "step": 921
+    },
+    {
+      "epoch": 2.506694129763131,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.012289286809384e-06,
+      "loss": 0.5546,
+      "step": 922
+    },
+    {
+      "epoch": 2.5094404394095435,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.982423367472344e-06,
+      "loss": 1.2097,
+      "step": 923
+    },
+    {
+      "epoch": 2.512186749055956,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 8.95258585246979e-06,
+      "loss": 0.5349,
+      "step": 924
+    },
+    {
+      "epoch": 2.514933058702369,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.922776882641604e-06,
+      "loss": 0.531,
+      "step": 925
+    },
+    {
+      "epoch": 2.5176793683487815,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 8.892996598692928e-06,
+      "loss": 0.4151,
+      "step": 926
+    },
+    {
+      "epoch": 2.5204256779951937,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 8.863245141193487e-06,
+      "loss": 0.4963,
+      "step": 927
+    },
+    {
+      "epoch": 2.5231719876416063,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 8.833522650576955e-06,
+      "loss": 0.5466,
+      "step": 928
+    },
+    {
+      "epoch": 2.525918297288019,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 8.803829267140263e-06,
+      "loss": 0.4034,
+      "step": 929
+    },
+    {
+      "epoch": 2.5286646069344316,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 8.774165131042957e-06,
+      "loss": 0.4094,
+      "step": 930
+    },
+    {
+      "epoch": 2.5314109165808443,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.744530382306528e-06,
+      "loss": 0.47,
+      "step": 931
+    },
+    {
+      "epoch": 2.534157226227257,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 8.714925160813752e-06,
+      "loss": 0.4783,
+      "step": 932
+    },
+    {
+      "epoch": 2.5369035358736696,
+      "grad_norm": 0.0390625,
+      "learning_rate": 8.68534960630802e-06,
+      "loss": 0.4795,
+      "step": 933
+    },
+    {
+      "epoch": 2.539649845520082,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 8.655803858392707e-06,
+      "loss": 0.5637,
+      "step": 934
+    },
+    {
+      "epoch": 2.542396155166495,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.626288056530474e-06,
+      "loss": 0.5958,
+      "step": 935
+    },
+    {
+      "epoch": 2.5451424648129075,
+      "grad_norm": 0.0546875,
+      "learning_rate": 8.596802340042648e-06,
+      "loss": 0.5443,
+      "step": 936
+    },
+    {
+      "epoch": 2.54788877445932,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 8.567346848108523e-06,
+      "loss": 0.5042,
+      "step": 937
+    },
+    {
+      "epoch": 2.550635084105733,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 8.53792171976476e-06,
+      "loss": 0.3745,
+      "step": 938
+    },
+    {
+      "epoch": 2.5533813937521455,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 8.508527093904663e-06,
+      "loss": 0.4595,
+      "step": 939
+    },
+    {
+      "epoch": 2.556127703398558,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.479163109277583e-06,
+      "loss": 0.5502,
+      "step": 940
+    },
+    {
+      "epoch": 2.5588740130449708,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 8.449829904488216e-06,
+      "loss": 0.4784,
+      "step": 941
+    },
+    {
+      "epoch": 2.5616203226913834,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 8.42052761799599e-06,
+      "loss": 0.5084,
+      "step": 942
+    },
+    {
+      "epoch": 2.564366632337796,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 8.391256388114367e-06,
+      "loss": 0.4844,
+      "step": 943
+    },
+    {
+      "epoch": 2.5671129419842087,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.362016353010248e-06,
+      "loss": 0.5863,
+      "step": 944
+    },
+    {
+      "epoch": 2.5698592516306213,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.332807650703255e-06,
+      "loss": 0.453,
+      "step": 945
+    },
+    {
+      "epoch": 2.572605561277034,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 8.303630419065136e-06,
+      "loss": 0.6364,
+      "step": 946
+    },
+    {
+      "epoch": 2.5753518709234466,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 8.274484795819068e-06,
+      "loss": 0.521,
+      "step": 947
+    },
+    {
+      "epoch": 2.5780981805698593,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 8.245370918539057e-06,
+      "loss": 0.47,
+      "step": 948
+    },
+    {
+      "epoch": 2.580844490216272,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.216288924649233e-06,
+      "loss": 0.516,
+      "step": 949
+    },
+    {
+      "epoch": 2.5835907998626846,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 8.187238951423254e-06,
+      "loss": 0.4951,
+      "step": 950
+    },
+    {
+      "epoch": 2.5863371095090972,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.158221135983606e-06,
+      "loss": 0.4366,
+      "step": 951
+    },
+    {
+      "epoch": 2.58908341915551,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.129235615301012e-06,
+      "loss": 0.5727,
+      "step": 952
+    },
+    {
+      "epoch": 2.5918297288019225,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.10028252619373e-06,
+      "loss": 0.601,
+      "step": 953
+    },
+    {
+      "epoch": 2.594576038448335,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 8.07136200532695e-06,
+      "loss": 0.474,
+      "step": 954
+    },
+    {
+      "epoch": 2.597322348094748,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 8.042474189212133e-06,
+      "loss": 0.3888,
+      "step": 955
+    },
+    {
+      "epoch": 2.6000686577411605,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 8.013619214206353e-06,
+      "loss": 0.4508,
+      "step": 956
+    },
+    {
+      "epoch": 2.602814967387573,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.984797216511686e-06,
+      "loss": 0.45,
+      "step": 957
+    },
+    {
+      "epoch": 2.6055612770339858,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 7.956008332174523e-06,
+      "loss": 0.4348,
+      "step": 958
+    },
+    {
+      "epoch": 2.6083075866803984,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 7.927252697084976e-06,
+      "loss": 0.4279,
+      "step": 959
+    },
+    {
+      "epoch": 2.611053896326811,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.898530446976194e-06,
+      "loss": 0.5555,
+      "step": 960
+    },
+    {
+      "epoch": 2.6138002059732237,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 7.86984171742376e-06,
+      "loss": 0.5695,
+      "step": 961
+    },
+    {
+      "epoch": 2.6165465156196364,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.841186643845009e-06,
+      "loss": 0.4705,
+      "step": 962
+    },
+    {
+      "epoch": 2.619292825266049,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.81256536149844e-06,
+      "loss": 0.486,
+      "step": 963
+    },
+    {
+      "epoch": 2.6220391349124617,
+      "grad_norm": 0.0546875,
+      "learning_rate": 7.783978005483024e-06,
+      "loss": 0.5018,
+      "step": 964
+    },
+    {
+      "epoch": 2.624785444558874,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 7.75542471073761e-06,
+      "loss": 0.4491,
+      "step": 965
+    },
+    {
+      "epoch": 2.6275317542052865,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 7.726905612040257e-06,
+      "loss": 0.4566,
+      "step": 966
+    },
+    {
+      "epoch": 2.630278063851699,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.698420844007624e-06,
+      "loss": 0.5227,
+      "step": 967
+    },
+    {
+      "epoch": 2.633024373498112,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 7.669970541094304e-06,
+      "loss": 0.4866,
+      "step": 968
+    },
+    {
+      "epoch": 2.6357706831445245,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 7.64155483759223e-06,
+      "loss": 0.4499,
+      "step": 969
+    },
+    {
+      "epoch": 2.638516992790937,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.613173867629991e-06,
+      "loss": 0.9577,
+      "step": 970
+    },
+    {
+      "epoch": 2.6412633024373497,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 7.584827765172254e-06,
+      "loss": 0.51,
+      "step": 971
+    },
+    {
+      "epoch": 2.6440096120837624,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 7.5565166640190784e-06,
+      "loss": 0.4697,
+      "step": 972
+    },
+    {
+      "epoch": 2.646755921730175,
+      "grad_norm": 0.046875,
+      "learning_rate": 7.528240697805321e-06,
+      "loss": 0.4789,
+      "step": 973
+    },
+    {
+      "epoch": 2.6495022313765877,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.500000000000004e-06,
+      "loss": 0.5087,
+      "step": 974
+    },
+    {
+      "epoch": 2.6522485410230003,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.471794703905647e-06,
+      "loss": 0.5238,
+      "step": 975
+    },
+    {
+      "epoch": 2.654994850669413,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.443624942657698e-06,
+      "loss": 0.5521,
+      "step": 976
+    },
+    {
+      "epoch": 2.6577411603158256,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 7.415490849223844e-06,
+      "loss": 0.4471,
+      "step": 977
+    },
+    {
+      "epoch": 2.6604874699622383,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.387392556403433e-06,
+      "loss": 0.4795,
+      "step": 978
+    },
+    {
+      "epoch": 2.663233779608651,
+      "grad_norm": 0.044921875,
+      "learning_rate": 7.359330196826808e-06,
+      "loss": 0.4769,
+      "step": 979
+    },
+    {
+      "epoch": 2.6659800892550636,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 7.33130390295472e-06,
+      "loss": 0.3953,
+      "step": 980
+    },
+    {
+      "epoch": 2.6687263989014762,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 7.303313807077658e-06,
+      "loss": 0.5334,
+      "step": 981
+    },
+    {
+      "epoch": 2.671472708547889,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.275360041315263e-06,
+      "loss": 0.512,
+      "step": 982
+    },
+    {
+      "epoch": 2.6742190181943015,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 7.24744273761569e-06,
+      "loss": 0.4317,
+      "step": 983
+    },
+    {
+      "epoch": 2.676965327840714,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 7.219562027754985e-06,
+      "loss": 0.5105,
+      "step": 984
+    },
+    {
+      "epoch": 2.679711637487127,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 7.191718043336447e-06,
+      "loss": 0.4319,
+      "step": 985
+    },
+    {
+      "epoch": 2.6824579471335395,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 7.163910915790047e-06,
+      "loss": 0.4596,
+      "step": 986
+    },
+    {
+      "epoch": 2.6852042567799517,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 7.13614077637176e-06,
+      "loss": 0.4915,
+      "step": 987
+    },
+    {
+      "epoch": 2.6879505664263643,
+      "grad_norm": 0.04296875,
+      "learning_rate": 7.108407756162988e-06,
+      "loss": 0.4317,
+      "step": 988
+    },
+    {
+      "epoch": 2.690696876072777,
+      "grad_norm": 0.04296875,
+      "learning_rate": 7.080711986069905e-06,
+      "loss": 0.5411,
+      "step": 989
+    },
+    {
+      "epoch": 2.6934431857191896,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 7.053053596822872e-06,
+      "loss": 0.3315,
+      "step": 990
+    },
+    {
+      "epoch": 2.6961894953656023,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.025432718975787e-06,
+      "loss": 0.417,
+      "step": 991
+    },
+    {
+      "epoch": 2.698935805012015,
+      "grad_norm": 0.048828125,
+      "learning_rate": 6.997849482905506e-06,
+      "loss": 0.5751,
+      "step": 992
+    },
+    {
+      "epoch": 2.7016821146584276,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 6.970304018811183e-06,
+      "loss": 0.5515,
+      "step": 993
+    },
+    {
+      "epoch": 2.70442842430484,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 6.942796456713706e-06,
+      "loss": 0.553,
+      "step": 994
+    },
+    {
+      "epoch": 2.707174733951253,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.915326926455029e-06,
+      "loss": 0.4753,
+      "step": 995
+    },
+    {
+      "epoch": 2.7099210435976655,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 6.887895557697614e-06,
+      "loss": 0.4289,
+      "step": 996
+    },
+    {
+      "epoch": 2.712667353244078,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 6.860502479923769e-06,
+      "loss": 0.4171,
+      "step": 997
+    },
+    {
+      "epoch": 2.715413662890491,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 6.833147822435075e-06,
+      "loss": 0.4769,
+      "step": 998
+    },
+    {
+      "epoch": 2.7181599725369034,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 6.8058317143517615e-06,
+      "loss": 0.4042,
+      "step": 999
+    },
+    {
+      "epoch": 2.720906282183316,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 6.778554284612078e-06,
+      "loss": 0.5019,
+      "step": 1000
+    },
+    {
+      "epoch": 2.7236525918297287,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.751315661971731e-06,
+      "loss": 0.4833,
+      "step": 1001
+    },
+    {
+      "epoch": 2.7236525918297287,
+      "eval_loss": 0.5037957429885864,
+      "eval_runtime": 619.9243,
+      "eval_samples_per_second": 14.786,
+      "eval_steps_per_second": 14.786,
+      "step": 1001
+    },
+    {
+      "epoch": 2.7263989014761414,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 6.724115975003217e-06,
+      "loss": 0.4036,
+      "step": 1002
+    },
+    {
+      "epoch": 2.729145211122554,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 6.696955352095277e-06,
+      "loss": 0.4995,
+      "step": 1003
+    },
+    {
+      "epoch": 2.7318915207689667,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 6.6698339214522374e-06,
+      "loss": 0.39,
+      "step": 1004
+    },
+    {
+      "epoch": 2.7346378304153793,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 6.642751811093446e-06,
+      "loss": 0.4771,
+      "step": 1005
+    },
+    {
+      "epoch": 2.737384140061792,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 6.6157091488526324e-06,
+      "loss": 0.4343,
+      "step": 1006
+    },
+    {
+      "epoch": 2.7401304497082046,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 6.588706062377344e-06,
+      "loss": 0.4141,
+      "step": 1007
+    },
+    {
+      "epoch": 2.7428767593546173,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 6.561742679128296e-06,
+      "loss": 0.4756,
+      "step": 1008
+    },
+    {
+      "epoch": 2.74562306900103,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.534819126378821e-06,
+      "loss": 0.6022,
+      "step": 1009
+    },
+    {
+      "epoch": 2.7483693786474426,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 6.507935531214218e-06,
+      "loss": 0.5495,
+      "step": 1010
+    },
+    {
+      "epoch": 2.751115688293855,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 6.4810920205312006e-06,
+      "loss": 0.4997,
+      "step": 1011
+    },
+    {
+      "epoch": 2.753861997940268,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 6.454288721037252e-06,
+      "loss": 0.438,
+      "step": 1012
+    },
+    {
+      "epoch": 2.7566083075866805,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 6.427525759250071e-06,
+      "loss": 0.5343,
+      "step": 1013
+    },
+    {
+      "epoch": 2.759354617233093,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.400803261496933e-06,
+      "loss": 1.0934,
+      "step": 1014
+    },
+    {
+      "epoch": 2.762100926879506,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 6.374121353914132e-06,
+      "loss": 0.4902,
+      "step": 1015
+    },
+    {
+      "epoch": 2.7648472365259185,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 6.347480162446349e-06,
+      "loss": 0.6164,
+      "step": 1016
+    },
+    {
+      "epoch": 2.767593546172331,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 6.320879812846093e-06,
+      "loss": 0.3764,
+      "step": 1017
+    },
+    {
+      "epoch": 2.7703398558187438,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.294320430673085e-06,
+      "loss": 0.5365,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7730861654651564,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.267802141293657e-06,
+      "loss": 0.4324,
+      "step": 1019
+    },
+    {
+      "epoch": 2.775832475111569,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 6.241325069880198e-06,
+      "loss": 0.367,
+      "step": 1020
+    },
+    {
+      "epoch": 2.7785787847579817,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 6.214889341410512e-06,
+      "loss": 0.4586,
+      "step": 1021
+    },
+    {
+      "epoch": 2.7813250944043943,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 6.188495080667278e-06,
+      "loss": 0.5402,
+      "step": 1022
+    },
+    {
+      "epoch": 2.784071404050807,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 6.162142412237421e-06,
+      "loss": 0.5498,
+      "step": 1023
+    },
+    {
+      "epoch": 2.786817713697219,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 6.135831460511555e-06,
+      "loss": 0.4409,
+      "step": 1024
+    },
+    {
+      "epoch": 2.789564023343632,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.109562349683366e-06,
+      "loss": 0.4341,
+      "step": 1025
+    },
+    {
+      "epoch": 2.7923103329900445,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 6.083335203749059e-06,
+      "loss": 0.6233,
+      "step": 1026
+    },
+    {
+      "epoch": 2.795056642636457,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 6.057150146506732e-06,
+      "loss": 0.5764,
+      "step": 1027
+    },
+    {
+      "epoch": 2.79780295228287,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.031007301555849e-06,
+      "loss": 0.4758,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8005492619292824,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 6.004906792296584e-06,
+      "loss": 0.4903,
+      "step": 1029
+    },
+    {
+      "epoch": 2.803295571575695,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.978848741929308e-06,
+      "loss": 0.5788,
+      "step": 1030
+    },
+    {
+      "epoch": 2.8060418812221077,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 5.952833273453953e-06,
+      "loss": 0.4795,
+      "step": 1031
+    },
+    {
+      "epoch": 2.8087881908685204,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 5.926860509669474e-06,
+      "loss": 0.4128,
+      "step": 1032
+    },
+    {
+      "epoch": 2.811534500514933,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 5.900930573173232e-06,
+      "loss": 0.5129,
+      "step": 1033
+    },
+    {
+      "epoch": 2.8142808101613457,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.8750435863604515e-06,
+      "loss": 0.5751,
+      "step": 1034
+    },
+    {
+      "epoch": 2.8170271198077583,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 5.849199671423609e-06,
+      "loss": 0.4868,
+      "step": 1035
+    },
+    {
+      "epoch": 2.819773429454171,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 5.823398950351886e-06,
+      "loss": 0.5558,
+      "step": 1036
+    },
+    {
+      "epoch": 2.8225197391005836,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 5.797641544930561e-06,
+      "loss": 0.497,
+      "step": 1037
+    },
+    {
+      "epoch": 2.8252660487469963,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 5.771927576740476e-06,
+      "loss": 0.4415,
+      "step": 1038
+    },
+    {
+      "epoch": 2.828012358393409,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 5.746257167157416e-06,
+      "loss": 0.5724,
+      "step": 1039
+    },
+    {
+      "epoch": 2.8307586680398216,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 5.72063043735158e-06,
+      "loss": 0.5275,
+      "step": 1040
+    },
+    {
+      "epoch": 2.833504977686234,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 5.6950475082869685e-06,
+      "loss": 0.4577,
+      "step": 1041
+    },
+    {
+      "epoch": 2.836251287332647,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 5.669508500720849e-06,
+      "loss": 0.5401,
+      "step": 1042
+    },
+    {
+      "epoch": 2.8389975969790595,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 5.6440135352031695e-06,
+      "loss": 0.4133,
+      "step": 1043
+    },
+    {
+      "epoch": 2.841743906625472,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 5.618562732075969e-06,
+      "loss": 0.4756,
+      "step": 1044
+    },
+    {
+      "epoch": 2.844490216271885,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.593156211472861e-06,
+      "loss": 0.5736,
+      "step": 1045
+    },
+    {
+      "epoch": 2.847236525918297,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 5.567794093318403e-06,
+      "loss": 0.5078,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8499828355647097,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.542476497327591e-06,
+      "loss": 0.5637,
+      "step": 1047
+    },
+    {
+      "epoch": 2.8527291452111223,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 5.517203543005242e-06,
+      "loss": 0.4383,
+      "step": 1048
+    },
+    {
+      "epoch": 2.855475454857535,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.491975349645479e-06,
+      "loss": 0.5174,
+      "step": 1049
+    },
+    {
+      "epoch": 2.8582217645039476,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 5.466792036331117e-06,
+      "loss": 0.4554,
+      "step": 1050
+    },
+    {
+      "epoch": 2.8609680741503603,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 5.44165372193315e-06,
+      "loss": 0.537,
+      "step": 1051
+    },
+    {
+      "epoch": 2.863714383796773,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 5.416560525110149e-06,
+      "loss": 0.4111,
+      "step": 1052
+    },
+    {
+      "epoch": 2.8664606934431855,
+      "grad_norm": 0.048828125,
+      "learning_rate": 5.391512564307737e-06,
+      "loss": 0.5282,
+      "step": 1053
+    },
+    {
+      "epoch": 2.869207003089598,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 5.36650995775799e-06,
+      "loss": 0.5688,
+      "step": 1054
+    },
+    {
+      "epoch": 2.871953312736011,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 5.341552823478929e-06,
+      "loss": 0.3545,
+      "step": 1055
+    },
+    {
+      "epoch": 2.8746996223824235,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 5.316641279273909e-06,
+      "loss": 0.3866,
+      "step": 1056
+    },
+    {
+      "epoch": 2.877445932028836,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 5.291775442731112e-06,
+      "loss": 0.4777,
+      "step": 1057
+    },
+    {
+      "epoch": 2.880192241675249,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 5.266955431222949e-06,
+      "loss": 0.498,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8829385513216614,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 5.242181361905548e-06,
+      "loss": 0.4791,
+      "step": 1059
+    },
+    {
+      "epoch": 2.885684860968074,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 5.217453351718155e-06,
+      "loss": 0.435,
+      "step": 1060
+    },
+    {
+      "epoch": 2.8884311706144867,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 5.192771517382627e-06,
+      "loss": 0.4513,
+      "step": 1061
+    },
+    {
+      "epoch": 2.8911774802608994,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 5.168135975402854e-06,
+      "loss": 0.5548,
+      "step": 1062
+    },
+    {
+      "epoch": 2.893923789907312,
+      "grad_norm": 0.04296875,
+      "learning_rate": 5.143546842064209e-06,
+      "loss": 0.4624,
+      "step": 1063
+    },
+    {
+      "epoch": 2.8966700995537247,
+      "grad_norm": 0.046875,
+      "learning_rate": 5.1190042334330185e-06,
+      "loss": 0.5901,
+      "step": 1064
+    },
+    {
+      "epoch": 2.8994164092001373,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 5.094508265355983e-06,
+      "loss": 0.5007,
+      "step": 1065
+    },
+    {
+      "epoch": 2.90216271884655,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 5.070059053459672e-06,
+      "loss": 0.3924,
+      "step": 1066
+    },
+    {
+      "epoch": 2.9049090284929626,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 5.045656713149932e-06,
+      "loss": 0.5346,
+      "step": 1067
+    },
+    {
+      "epoch": 2.9076553381393753,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.021301359611387e-06,
+      "loss": 0.4761,
+      "step": 1068
+    },
+    {
+      "epoch": 2.910401647785788,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.996993107806853e-06,
+      "loss": 0.4432,
+      "step": 1069
+    },
+    {
+      "epoch": 2.9131479574322006,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 4.972732072476831e-06,
+      "loss": 0.4404,
+      "step": 1070
+    },
+    {
+      "epoch": 2.915894267078613,
+      "grad_norm": 0.046875,
+      "learning_rate": 4.948518368138933e-06,
+      "loss": 0.5556,
+      "step": 1071
+    },
+    {
+      "epoch": 2.918640576725026,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 4.9243521090873745e-06,
+      "loss": 0.523,
+      "step": 1072
+    },
+    {
+      "epoch": 2.9213868863714385,
+      "grad_norm": 0.046875,
+      "learning_rate": 4.900233409392409e-06,
+      "loss": 0.5381,
+      "step": 1073
+    },
+    {
+      "epoch": 2.924133196017851,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 4.876162382899809e-06,
+      "loss": 0.5505,
+      "step": 1074
+    },
+    {
+      "epoch": 2.926879505664264,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.852139143230296e-06,
+      "loss": 1.176,
+      "step": 1075
+    },
+    {
+      "epoch": 2.9296258153106765,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 4.828163803779057e-06,
+      "loss": 0.4169,
+      "step": 1076
+    },
+    {
+      "epoch": 2.932372124957089,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.804236477715152e-06,
+      "loss": 0.6101,
+      "step": 1077
+    },
+    {
+      "epoch": 2.9351184346035017,
+      "grad_norm": 0.044921875,
+      "learning_rate": 4.780357277981027e-06,
+      "loss": 0.4059,
+      "step": 1078
+    },
+    {
+      "epoch": 2.9378647442499144,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 4.7565263172919415e-06,
+      "loss": 0.5825,
+      "step": 1079
+    },
+    {
+      "epoch": 2.940611053896327,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 4.732743708135479e-06,
+      "loss": 0.45,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9433573635427397,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.709009562770971e-06,
+      "loss": 0.4906,
+      "step": 1081
+    },
+    {
+      "epoch": 2.9461036731891523,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.685323993229005e-06,
+      "loss": 0.5843,
+      "step": 1082
+    },
+    {
+      "epoch": 2.948849982835565,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.661687111310865e-06,
+      "loss": 0.4679,
+      "step": 1083
+    },
+    {
+      "epoch": 2.951596292481977,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.638099028588034e-06,
+      "loss": 0.5253,
+      "step": 1084
+    },
+    {
+      "epoch": 2.95434260212839,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.614559856401635e-06,
+      "loss": 0.4255,
+      "step": 1085
+    },
+    {
+      "epoch": 2.9570889117748025,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 4.591069705861935e-06,
+      "loss": 0.4591,
+      "step": 1086
+    },
+    {
+      "epoch": 2.959835221421215,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.567628687847808e-06,
+      "loss": 0.4433,
+      "step": 1087
+    },
+    {
+      "epoch": 2.962581531067628,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 4.544236913006199e-06,
+      "loss": 0.4516,
+      "step": 1088
+    },
+    {
+      "epoch": 2.9653278407140404,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.520894491751629e-06,
+      "loss": 0.5292,
+      "step": 1089
+    },
+    {
+      "epoch": 2.968074150360453,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.497601534265641e-06,
+      "loss": 0.5397,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9708204600068657,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 4.4743581504963206e-06,
+      "loss": 0.5584,
+      "step": 1091
+    },
+    {
+      "epoch": 2.9735667696532784,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.451164450157729e-06,
+      "loss": 0.4986,
+      "step": 1092
+    },
+    {
+      "epoch": 2.9735667696532784,
+      "eval_loss": 0.5037021636962891,
+      "eval_runtime": 614.9978,
+      "eval_samples_per_second": 14.904,
+      "eval_steps_per_second": 14.904,
+      "step": 1092
+    },
+    {
+      "epoch": 2.976313079299691,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.428020542729436e-06,
+      "loss": 0.5396,
+      "step": 1093
+    },
+    {
+      "epoch": 2.9790593889461037,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 4.4049265374559536e-06,
+      "loss": 0.4538,
+      "step": 1094
+    },
+    {
+      "epoch": 2.9818056985925163,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.381882543346262e-06,
+      "loss": 0.3633,
+      "step": 1095
+    },
+    {
+      "epoch": 2.984552008238929,
+      "grad_norm": 0.046875,
+      "learning_rate": 4.358888669173264e-06,
+      "loss": 0.5483,
+      "step": 1096
+    },
+    {
+      "epoch": 2.9872983178853416,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.3359450234733e-06,
+      "loss": 0.3848,
+      "step": 1097
+    },
+    {
+      "epoch": 2.9900446275317543,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.3130517145456e-06,
+      "loss": 0.6011,
+      "step": 1098
+    },
+    {
+      "epoch": 2.992790937178167,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 4.29020885045182e-06,
+      "loss": 0.4609,
+      "step": 1099
+    },
+    {
+      "epoch": 2.9955372468245796,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 4.267416539015474e-06,
+      "loss": 0.3615,
+      "step": 1100
+    },
+    {
+      "epoch": 2.998283556470992,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 4.244674887821483e-06,
+      "loss": 0.4688,
+      "step": 1101
+    },
+    {
+      "epoch": 3.0010298661174044,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 4.221984004215623e-06,
+      "loss": 0.3021,
+      "step": 1102
+    },
+    {
+      "epoch": 3.003776175763817,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 4.199343995304044e-06,
+      "loss": 0.3841,
+      "step": 1103
+    },
+    {
+      "epoch": 3.0065224854102297,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 4.176754967952749e-06,
+      "loss": 0.4316,
+      "step": 1104
+    },
+    {
+      "epoch": 3.0092687950566424,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.154217028787101e-06,
+      "loss": 0.5092,
+      "step": 1105
+    },
+    {
+      "epoch": 3.012015104703055,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 4.131730284191321e-06,
+      "loss": 0.4633,
+      "step": 1106
+    },
+    {
+      "epoch": 3.0020597322348093,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.109294840307966e-06,
+      "loss": 0.4454,
+      "step": 1107
+    },
+    {
+      "epoch": 3.004806041881222,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.086910803037453e-06,
+      "loss": 0.4654,
+      "step": 1108
+    },
+    {
+      "epoch": 3.0075523515276346,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.064578278037542e-06,
+      "loss": 0.4323,
+      "step": 1109
+    },
+    {
+      "epoch": 3.010298661174047,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 4.042297370722851e-06,
+      "loss": 0.4796,
+      "step": 1110
+    },
+    {
+      "epoch": 3.01304497082046,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.0200681862643355e-06,
+      "loss": 0.4253,
+      "step": 1111
+    },
+    {
+      "epoch": 3.0157912804668725,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 3.9978908295888285e-06,
+      "loss": 0.4095,
+      "step": 1112
+    },
+    {
+      "epoch": 3.018537590113285,
+      "grad_norm": 0.044921875,
+      "learning_rate": 3.975765405378502e-06,
+      "loss": 0.4575,
+      "step": 1113
+    },
+    {
+      "epoch": 3.021283899759698,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 3.953692018070417e-06,
+      "loss": 0.4556,
+      "step": 1114
+    },
+    {
+      "epoch": 3.0240302094061104,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.931670771855986e-06,
+      "loss": 0.5403,
+      "step": 1115
+    },
+    {
+      "epoch": 3.026776519052523,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.909701770680524e-06,
+      "loss": 0.4718,
+      "step": 1116
+    },
+    {
+      "epoch": 3.0295228286989357,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 3.887785118242722e-06,
+      "loss": 0.4633,
+      "step": 1117
+    },
+    {
+      "epoch": 3.0322691383453484,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.8659209179941804e-06,
+      "loss": 0.5703,
+      "step": 1118
+    },
+    {
+      "epoch": 3.035015447991761,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.844109273138914e-06,
+      "loss": 0.5709,
+      "step": 1119
+    },
+    {
+      "epoch": 3.0377617576381737,
+      "grad_norm": 0.046875,
+      "learning_rate": 3.822350286632867e-06,
+      "loss": 0.4592,
+      "step": 1120
+    },
+    {
+      "epoch": 3.0405080672845863,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.8006440611834103e-06,
+      "loss": 0.4843,
+      "step": 1121
+    },
+    {
+      "epoch": 3.043254376930999,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.7789906992488875e-06,
+      "loss": 0.4962,
+      "step": 1122
+    },
+    {
+      "epoch": 3.0460006865774116,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.7573903030381003e-06,
+      "loss": 0.5629,
+      "step": 1123
+    },
+    {
+      "epoch": 3.0487469962238243,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.7358429745098525e-06,
+      "loss": 0.5103,
+      "step": 1124
+    },
+    {
+      "epoch": 3.051493305870237,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.7143488153724454e-06,
+      "loss": 0.4677,
+      "step": 1125
+    },
+    {
+      "epoch": 3.0542396155166496,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 3.692907927083217e-06,
+      "loss": 0.423,
+      "step": 1126
+    },
+    {
+      "epoch": 3.0569859251630622,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.6715204108480473e-06,
+      "loss": 0.4903,
+      "step": 1127
+    },
+    {
+      "epoch": 3.059732234809475,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.6501863676208984e-06,
+      "loss": 0.9496,
+      "step": 1128
+    },
+    {
+      "epoch": 3.0624785444558875,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.6289058981033136e-06,
+      "loss": 0.4253,
+      "step": 1129
+    },
+    {
+      "epoch": 3.0652248541023,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 3.607679102743968e-06,
+      "loss": 0.5686,
+      "step": 1130
+    },
+    {
+      "epoch": 3.067971163748713,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.586506081738181e-06,
+      "loss": 0.5278,
+      "step": 1131
+    },
+    {
+      "epoch": 3.0707174733951255,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.5653869350274357e-06,
+      "loss": 0.4348,
+      "step": 1132
+    },
+    {
+      "epoch": 3.073463783041538,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 3.5443217622989294e-06,
+      "loss": 0.5263,
+      "step": 1133
+    },
+    {
+      "epoch": 3.0762100926879508,
+      "grad_norm": 0.046875,
+      "learning_rate": 3.5233106629850736e-06,
+      "loss": 0.5263,
+      "step": 1134
+    },
+    {
+      "epoch": 3.0789564023343634,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.5023537362630605e-06,
+      "loss": 0.4807,
+      "step": 1135
+    },
+    {
+      "epoch": 3.0817027119807756,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.4814510810543553e-06,
+      "loss": 0.6053,
+      "step": 1136
+    },
+    {
+      "epoch": 3.0844490216271883,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 3.46060279602427e-06,
+      "loss": 0.391,
+      "step": 1137
+    },
+    {
+      "epoch": 3.087195331273601,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.439808979581455e-06,
+      "loss": 0.4525,
+      "step": 1138
+    },
+    {
+      "epoch": 3.0899416409200136,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.4190697298774772e-06,
+      "loss": 0.532,
+      "step": 1139
+    },
+    {
+      "epoch": 3.092687950566426,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.398385144806318e-06,
+      "loss": 0.5811,
+      "step": 1140
+    },
+    {
+      "epoch": 3.095434260212839,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.3777553220039455e-06,
+      "loss": 0.5059,
+      "step": 1141
+    },
+    {
+      "epoch": 3.0981805698592515,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.357180358847822e-06,
+      "loss": 0.4974,
+      "step": 1142
+    },
+    {
+      "epoch": 3.100926879505664,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.3366603524564736e-06,
+      "loss": 0.52,
+      "step": 1143
+    },
+    {
+      "epoch": 3.103673189152077,
+      "grad_norm": 0.04296875,
+      "learning_rate": 3.316195399689007e-06,
+      "loss": 0.4295,
+      "step": 1144
+    },
+    {
+      "epoch": 3.1064194987984894,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.2957855971446737e-06,
+      "loss": 0.4381,
+      "step": 1145
+    },
+    {
+      "epoch": 3.109165808444902,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.2754310411623888e-06,
+      "loss": 0.4879,
+      "step": 1146
+    },
+    {
+      "epoch": 3.1119121180913147,
+      "grad_norm": 0.04296875,
+      "learning_rate": 3.255131827820311e-06,
+      "loss": 0.4444,
+      "step": 1147
+    },
+    {
+      "epoch": 3.1146584277377274,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.2348880529353484e-06,
+      "loss": 0.4969,
+      "step": 1148
+    },
+    {
+      "epoch": 3.11740473738414,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.21469981206274e-06,
+      "loss": 0.5399,
+      "step": 1149
+    },
+    {
+      "epoch": 3.1201510470305527,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.194567200495593e-06,
+      "loss": 0.3839,
+      "step": 1150
+    },
+    {
+      "epoch": 3.1228973566769653,
+      "grad_norm": 0.046875,
+      "learning_rate": 3.1744903132644197e-06,
+      "loss": 0.5803,
+      "step": 1151
+    },
+    {
+      "epoch": 3.125643666323378,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.1544692451367147e-06,
+      "loss": 0.5422,
+      "step": 1152
+    },
+    {
+      "epoch": 3.1283899759697906,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.1345040906164787e-06,
+      "loss": 0.4212,
+      "step": 1153
+    },
+    {
+      "epoch": 3.1311362856162033,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.1145949439438054e-06,
+      "loss": 0.4019,
+      "step": 1154
+    },
+    {
+      "epoch": 3.133882595262616,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.094741899094399e-06,
+      "loss": 0.3445,
+      "step": 1155
+    },
+    {
+      "epoch": 3.1366289049090286,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 3.0749450497791693e-06,
+      "loss": 0.518,
+      "step": 1156
+    },
+    {
+      "epoch": 3.139375214555441,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.055204489443753e-06,
+      "loss": 0.4594,
+      "step": 1157
+    },
+    {
+      "epoch": 3.142121524201854,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 3.0355203112681063e-06,
+      "loss": 0.5042,
+      "step": 1158
+    },
+    {
+      "epoch": 3.1448678338482665,
+      "grad_norm": 0.041015625,
+      "learning_rate": 3.0158926081660338e-06,
+      "loss": 0.4187,
+      "step": 1159
+    },
+    {
+      "epoch": 3.147614143494679,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.9963214727847773e-06,
+      "loss": 0.5166,
+      "step": 1160
+    },
+    {
+      "epoch": 3.150360453141092,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.976806997504555e-06,
+      "loss": 0.4656,
+      "step": 1161
+    },
+    {
+      "epoch": 3.1531067627875045,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.9573492744381475e-06,
+      "loss": 0.4555,
+      "step": 1162
+    },
+    {
+      "epoch": 3.155853072433917,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.9379483954304386e-06,
+      "loss": 0.5357,
+      "step": 1163
+    },
+    {
+      "epoch": 3.1585993820803298,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9186044520580145e-06,
+      "loss": 1.2069,
+      "step": 1164
+    },
+    {
+      "epoch": 3.1613456917267424,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.8993175356286934e-06,
+      "loss": 0.5469,
+      "step": 1165
+    },
+    {
+      "epoch": 3.164092001373155,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.8800877371811245e-06,
+      "loss": 0.5138,
+      "step": 1166
+    },
+    {
+      "epoch": 3.1668383110195673,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.8609151474843377e-06,
+      "loss": 0.486,
+      "step": 1167
+    },
+    {
+      "epoch": 3.16958462066598,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.841799857037337e-06,
+      "loss": 0.5253,
+      "step": 1168
+    },
+    {
+      "epoch": 3.1723309303123925,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.822741956068648e-06,
+      "loss": 0.4386,
+      "step": 1169
+    },
+    {
+      "epoch": 3.175077239958805,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.803741534535916e-06,
+      "loss": 0.4587,
+      "step": 1170
+    },
+    {
+      "epoch": 3.177823549605218,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.7847986821254605e-06,
+      "loss": 0.445,
+      "step": 1171
+    },
+    {
+      "epoch": 3.1805698592516305,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.7659134882518715e-06,
+      "loss": 0.532,
+      "step": 1172
+    },
+    {
+      "epoch": 3.183316168898043,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.747086042057566e-06,
+      "loss": 0.5697,
+      "step": 1173
+    },
+    {
+      "epoch": 3.186062478544456,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.7283164324123904e-06,
+      "loss": 0.5351,
+      "step": 1174
+    },
+    {
+      "epoch": 3.1888087881908684,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.7096047479131848e-06,
+      "loss": 0.5825,
+      "step": 1175
+    },
+    {
+      "epoch": 3.191555097837281,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.6909510768833606e-06,
+      "loss": 0.4346,
+      "step": 1176
+    },
+    {
+      "epoch": 3.1943014074836937,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.6723555073725125e-06,
+      "loss": 0.4932,
+      "step": 1177
+    },
+    {
+      "epoch": 3.1970477171301064,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.653818127155959e-06,
+      "loss": 0.5453,
+      "step": 1178
+    },
+    {
+      "epoch": 3.199794026776519,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.635339023734374e-06,
+      "loss": 0.5351,
+      "step": 1179
+    },
+    {
+      "epoch": 3.2025403364229317,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.6169182843333334e-06,
+      "loss": 0.4668,
+      "step": 1180
+    },
+    {
+      "epoch": 3.2052866460693443,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.5985559959029347e-06,
+      "loss": 0.4176,
+      "step": 1181
+    },
+    {
+      "epoch": 3.208032955715757,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.5802522451173627e-06,
+      "loss": 0.4717,
+      "step": 1182
+    },
+    {
+      "epoch": 3.2107792653621696,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.562007118374504e-06,
+      "loss": 0.5227,
+      "step": 1183
+    },
+    {
+      "epoch": 3.2107792653621696,
+      "eval_loss": 0.5037118196487427,
+      "eval_runtime": 639.2982,
+      "eval_samples_per_second": 14.338,
+      "eval_steps_per_second": 14.338,
+      "step": 1183
+    },
+    {
+      "epoch": 3.2135255750085823,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.543820701795511e-06,
+      "loss": 0.4845,
+      "step": 1184
+    },
+    {
+      "epoch": 3.216271884654995,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.5256930812244273e-06,
+      "loss": 1.1749,
+      "step": 1185
+    },
+    {
+      "epoch": 3.2190181943014076,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.507624342227748e-06,
+      "loss": 1.1636,
+      "step": 1186
+    },
+    {
+      "epoch": 3.22176450394782,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.4896145700940524e-06,
+      "loss": 0.508,
+      "step": 1187
+    },
+    {
+      "epoch": 3.224510813594233,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.471663849833567e-06,
+      "loss": 1.1039,
+      "step": 1188
+    },
+    {
+      "epoch": 3.2272571232406455,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.453772266177791e-06,
+      "loss": 0.5217,
+      "step": 1189
+    },
+    {
+      "epoch": 3.230003432887058,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.435939903579075e-06,
+      "loss": 0.5168,
+      "step": 1190
+    },
+    {
+      "epoch": 3.232749742533471,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 2.4181668462102478e-06,
+      "loss": 0.6517,
+      "step": 1191
+    },
+    {
+      "epoch": 3.2354960521798835,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.4004531779641835e-06,
+      "loss": 0.439,
+      "step": 1192
+    },
+    {
+      "epoch": 3.238242361826296,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.382798982453444e-06,
+      "loss": 0.4339,
+      "step": 1193
+    },
+    {
+      "epoch": 3.2409886714727087,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.3652043430098624e-06,
+      "loss": 0.493,
+      "step": 1194
+    },
+    {
+      "epoch": 3.243734981119121,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.3476693426841417e-06,
+      "loss": 0.389,
+      "step": 1195
+    },
+    {
+      "epoch": 3.2464812907655336,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.3301940642454932e-06,
+      "loss": 0.5022,
+      "step": 1196
+    },
+    {
+      "epoch": 3.2492276004119462,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.3127785901812093e-06,
+      "loss": 0.5275,
+      "step": 1197
+    },
+    {
+      "epoch": 3.251973910058359,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.2954230026963092e-06,
+      "loss": 0.5681,
+      "step": 1198
+    },
+    {
+      "epoch": 3.2547202197047715,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.278127383713117e-06,
+      "loss": 0.4153,
+      "step": 1199
+    },
+    {
+      "epoch": 3.257466529351184,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.2608918148709057e-06,
+      "loss": 0.4188,
+      "step": 1200
+    },
+    {
+      "epoch": 3.260212838997597,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.2437163775254863e-06,
+      "loss": 0.4721,
+      "step": 1201
+    },
+    {
+      "epoch": 3.2629591486440095,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.22660115274885e-06,
+      "loss": 0.443,
+      "step": 1202
+    },
+    {
+      "epoch": 3.265705458290422,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.2095462213287526e-06,
+      "loss": 0.4396,
+      "step": 1203
+    },
+    {
+      "epoch": 3.268451767936835,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.19255166376837e-06,
+      "loss": 0.4431,
+      "step": 1204
+    },
+    {
+      "epoch": 3.2711980775832474,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.175617560285883e-06,
+      "loss": 0.5235,
+      "step": 1205
+    },
+    {
+      "epoch": 3.27394438722966,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.158743990814128e-06,
+      "loss": 0.357,
+      "step": 1206
+    },
+    {
+      "epoch": 3.2766906968760727,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.1419310350001997e-06,
+      "loss": 1.1825,
+      "step": 1207
+    },
+    {
+      "epoch": 3.2794370065224854,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.1251787722050854e-06,
+      "loss": 0.5559,
+      "step": 1208
+    },
+    {
+      "epoch": 3.282183316168898,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.1084872815032885e-06,
+      "loss": 0.529,
+      "step": 1209
+    },
+    {
+      "epoch": 3.2849296258153107,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.0918566416824557e-06,
+      "loss": 0.5814,
+      "step": 1210
+    },
+    {
+      "epoch": 3.2876759354617233,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.075286931242995e-06,
+      "loss": 0.4434,
+      "step": 1211
+    },
+    {
+      "epoch": 3.290422245108136,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.058778228397726e-06,
+      "loss": 0.4634,
+      "step": 1212
+    },
+    {
+      "epoch": 3.2931685547545486,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.0423306110714846e-06,
+      "loss": 0.4865,
+      "step": 1213
+    },
+    {
+      "epoch": 3.2959148644009613,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.0259441569007836e-06,
+      "loss": 0.4826,
+      "step": 1214
+    },
+    {
+      "epoch": 3.298661174047374,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.0096189432334194e-06,
+      "loss": 0.399,
+      "step": 1215
+    },
+    {
+      "epoch": 3.3014074836937866,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.9933550471281315e-06,
+      "loss": 0.5276,
+      "step": 1216
+    },
+    {
+      "epoch": 3.304153793340199,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.9771525453542123e-06,
+      "loss": 0.4359,
+      "step": 1217
+    },
+    {
+      "epoch": 3.306900102986612,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.9610115143911696e-06,
+      "loss": 0.4383,
+      "step": 1218
+    },
+    {
+      "epoch": 3.3096464126330245,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9449320304283545e-06,
+      "loss": 0.4844,
+      "step": 1219
+    },
+    {
+      "epoch": 3.312392722279437,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.928914169364595e-06,
+      "loss": 0.4703,
+      "step": 1220
+    },
+    {
+      "epoch": 3.31513903192585,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9129580068078556e-06,
+      "loss": 0.5235,
+      "step": 1221
+    },
+    {
+      "epoch": 3.3178853415722624,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.8970636180748542e-06,
+      "loss": 0.4168,
+      "step": 1222
+    },
+    {
+      "epoch": 3.320631651218675,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.8812310781907416e-06,
+      "loss": 0.6157,
+      "step": 1223
+    },
+    {
+      "epoch": 3.3233779608650877,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.8654604618887095e-06,
+      "loss": 0.5189,
+      "step": 1224
+    },
+    {
+      "epoch": 3.3261242705115004,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.8497518436096727e-06,
+      "loss": 0.5767,
+      "step": 1225
+    },
+    {
+      "epoch": 3.328870580157913,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.8341052975018856e-06,
+      "loss": 0.4845,
+      "step": 1226
+    },
+    {
+      "epoch": 3.3316168898043257,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.8185208974206202e-06,
+      "loss": 0.5266,
+      "step": 1227
+    },
+    {
+      "epoch": 3.334363199450738,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.8029987169277962e-06,
+      "loss": 0.4485,
+      "step": 1228
+    },
+    {
+      "epoch": 3.3371095090971505,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 1.7875388292916516e-06,
+      "loss": 0.5584,
+      "step": 1229
+    },
+    {
+      "epoch": 3.339855818743563,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.7721413074863769e-06,
+      "loss": 0.4716,
+      "step": 1230
+    },
+    {
+      "epoch": 3.342602128389976,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.7568062241917937e-06,
+      "loss": 0.5874,
+      "step": 1231
+    },
+    {
+      "epoch": 3.3453484380363885,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.7415336517929886e-06,
+      "loss": 0.384,
+      "step": 1232
+    },
+    {
+      "epoch": 3.348094747682801,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.726323662379992e-06,
+      "loss": 0.5479,
+      "step": 1233
+    },
+    {
+      "epoch": 3.350841057329214,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.7111763277474179e-06,
+      "loss": 0.4631,
+      "step": 1234
+    },
+    {
+      "epoch": 3.3535873669756264,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6960917193941478e-06,
+      "loss": 0.5318,
+      "step": 1235
+    },
+    {
+      "epoch": 3.356333676622039,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.6810699085229685e-06,
+      "loss": 0.4914,
+      "step": 1236
+    },
+    {
+      "epoch": 3.3590799862684517,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6661109660402563e-06,
+      "loss": 0.5059,
+      "step": 1237
+    },
+    {
+      "epoch": 3.3618262959148644,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.6512149625556371e-06,
+      "loss": 0.3797,
+      "step": 1238
+    },
+    {
+      "epoch": 3.364572605561277,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6363819683816372e-06,
+      "loss": 0.5895,
+      "step": 1239
+    },
+    {
+      "epoch": 3.3673189152076897,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.6216120535333818e-06,
+      "loss": 0.5115,
+      "step": 1240
+    },
+    {
+      "epoch": 3.3700652248541023,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.6069052877282292e-06,
+      "loss": 0.414,
+      "step": 1241
+    },
+    {
+      "epoch": 3.372811534500515,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.5922617403854768e-06,
+      "loss": 0.5862,
+      "step": 1242
+    },
+    {
+      "epoch": 3.3755578441469276,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.5776814806260032e-06,
+      "loss": 0.5591,
+      "step": 1243
+    },
+    {
+      "epoch": 3.3783041537933403,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.563164577271965e-06,
+      "loss": 0.5291,
+      "step": 1244
+    },
+    {
+      "epoch": 3.381050463439753,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.5487110988464526e-06,
+      "loss": 0.5359,
+      "step": 1245
+    },
+    {
+      "epoch": 3.3837967730861656,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.5343211135731894e-06,
+      "loss": 0.4184,
+      "step": 1246
+    },
+    {
+      "epoch": 3.386543082732578,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.5199946893761785e-06,
+      "loss": 0.4923,
+      "step": 1247
+    },
+    {
+      "epoch": 3.389289392378991,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.5057318938794195e-06,
+      "loss": 0.4915,
+      "step": 1248
+    },
+    {
+      "epoch": 3.3920357020254035,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.4915327944065544e-06,
+      "loss": 0.5856,
+      "step": 1249
+    },
+    {
+      "epoch": 3.394782011671816,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.4773974579805783e-06,
+      "loss": 0.5155,
+      "step": 1250
+    },
+    {
+      "epoch": 3.397528321318229,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.463325951323496e-06,
+      "loss": 0.5409,
+      "step": 1251
+    },
+    {
+      "epoch": 3.4002746309646414,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.4493183408560361e-06,
+      "loss": 0.4235,
+      "step": 1252
+    },
+    {
+      "epoch": 3.403020940611054,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.4353746926973077e-06,
+      "loss": 0.4903,
+      "step": 1253
+    },
+    {
+      "epoch": 3.4057672502574663,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.421495072664522e-06,
+      "loss": 0.5838,
+      "step": 1254
+    },
+    {
+      "epoch": 3.408513559903879,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.407679546272641e-06,
+      "loss": 0.4681,
+      "step": 1255
+    },
+    {
+      "epoch": 3.4112598695502916,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.3939281787341102e-06,
+      "loss": 0.5009,
+      "step": 1256
+    },
+    {
+      "epoch": 3.4140061791967042,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.3802410349585153e-06,
+      "loss": 0.5052,
+      "step": 1257
+    },
+    {
+      "epoch": 3.416752488843117,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3666181795523004e-06,
+      "loss": 0.6236,
+      "step": 1258
+    },
+    {
+      "epoch": 3.4194987984895295,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.3530596768184478e-06,
+      "loss": 0.5415,
+      "step": 1259
+    },
+    {
+      "epoch": 3.422245108135942,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.3395655907561871e-06,
+      "loss": 0.4757,
+      "step": 1260
+    },
+    {
+      "epoch": 3.424991417782355,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.3261359850606792e-06,
+      "loss": 0.4161,
+      "step": 1261
+    },
+    {
+      "epoch": 3.4277377274287675,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.3127709231227281e-06,
+      "loss": 0.5019,
+      "step": 1262
+    },
+    {
+      "epoch": 3.43048403707518,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.2994704680284786e-06,
+      "loss": 0.4118,
+      "step": 1263
+    },
+    {
+      "epoch": 3.4332303467215928,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 1.2862346825591075e-06,
+      "loss": 0.495,
+      "step": 1264
+    },
+    {
+      "epoch": 3.4359766563680054,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.2730636291905462e-06,
+      "loss": 0.4988,
+      "step": 1265
+    },
+    {
+      "epoch": 3.438722966014418,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.2599573700931666e-06,
+      "loss": 0.3351,
+      "step": 1266
+    },
+    {
+      "epoch": 3.4414692756608307,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.2469159671315072e-06,
+      "loss": 0.5976,
+      "step": 1267
+    },
+    {
+      "epoch": 3.4442155853072434,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.2339394818639583e-06,
+      "loss": 0.5267,
+      "step": 1268
+    },
+    {
+      "epoch": 3.446961894953656,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.2210279755424981e-06,
+      "loss": 0.4385,
+      "step": 1269
+    },
+    {
+      "epoch": 3.4497082046000687,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.2081815091123755e-06,
+      "loss": 0.4201,
+      "step": 1270
+    },
+    {
+      "epoch": 3.4524545142464813,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.1954001432118482e-06,
+      "loss": 0.5371,
+      "step": 1271
+    },
+    {
+      "epoch": 3.455200823892894,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.1826839381718752e-06,
+      "loss": 0.4782,
+      "step": 1272
+    },
+    {
+      "epoch": 3.4579471335393066,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.1700329540158473e-06,
+      "loss": 0.4626,
+      "step": 1273
+    },
+    {
+      "epoch": 3.4606934431857193,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.157447250459292e-06,
+      "loss": 0.5723,
+      "step": 1274
+    },
+    {
+      "epoch": 3.4606934431857193,
+      "eval_loss": 0.5037119388580322,
+      "eval_runtime": 636.191,
+      "eval_samples_per_second": 14.408,
+      "eval_steps_per_second": 14.408,
+      "step": 1274
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1456,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 91,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.582524910880162e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}