End of training

README.md

@@ -4,6 +4,7 @@ license: apache-2.0
 base_model: Qwen/Qwen2.5-7B-Instruct
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: global_batchsize_1024_laradjusted2
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # global_batchsize_1024_laradjusted2
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the open-thoughts/OpenThoughts-114k dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9943851768669285,
+    "total_flos": 4607052149424128.0,
+    "train_loss": 0.5244213240640657,
+    "train_runtime": 115493.2352,
+    "train_samples_per_second": 2.96,
+    "train_steps_per_second": 0.003
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9943851768669285,
+    "total_flos": 4607052149424128.0,
+    "train_loss": 0.5244213240640657,
+    "train_runtime": 115493.2352,
+    "train_samples_per_second": 2.96,
+    "train_steps_per_second": 0.003
+}

trainer_state.json

@@ -0,0 +1,2373 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9943851768669285,
+  "eval_steps": 500,
+  "global_step": 333,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008983717012914094,
+      "grad_norm": 5.839033078230819,
+      "learning_rate": 6.655117647058823e-06,
+      "loss": 0.851,
+      "step": 1
+    },
+    {
+      "epoch": 0.017967434025828188,
+      "grad_norm": 5.796609591178738,
+      "learning_rate": 1.3310235294117646e-05,
+      "loss": 0.8589,
+      "step": 2
+    },
+    {
+      "epoch": 0.02695115103874228,
+      "grad_norm": 4.022069478985531,
+      "learning_rate": 1.996535294117647e-05,
+      "loss": 0.8091,
+      "step": 3
+    },
+    {
+      "epoch": 0.035934868051656375,
+      "grad_norm": 5.021116871190826,
+      "learning_rate": 2.6620470588235293e-05,
+      "loss": 0.8296,
+      "step": 4
+    },
+    {
+      "epoch": 0.044918585064570464,
+      "grad_norm": 6.766543105088873,
+      "learning_rate": 3.3275588235294115e-05,
+      "loss": 0.8406,
+      "step": 5
+    },
+    {
+      "epoch": 0.05390230207748456,
+      "grad_norm": 3.7488525352605198,
+      "learning_rate": 3.993070588235294e-05,
+      "loss": 0.7848,
+      "step": 6
+    },
+    {
+      "epoch": 0.06288601909039865,
+      "grad_norm": 3.232161490369813,
+      "learning_rate": 4.658582352941176e-05,
+      "loss": 0.7269,
+      "step": 7
+    },
+    {
+      "epoch": 0.07186973610331275,
+      "grad_norm": 3.5299607181225734,
+      "learning_rate": 5.3240941176470586e-05,
+      "loss": 0.7032,
+      "step": 8
+    },
+    {
+      "epoch": 0.08085345311622684,
+      "grad_norm": 2.4635657277606833,
+      "learning_rate": 5.989605882352941e-05,
+      "loss": 0.686,
+      "step": 9
+    },
+    {
+      "epoch": 0.08983717012914093,
+      "grad_norm": 3.5184074662375378,
+      "learning_rate": 6.655117647058823e-05,
+      "loss": 0.6731,
+      "step": 10
+    },
+    {
+      "epoch": 0.09882088714205503,
+      "grad_norm": 2.034739373521881,
+      "learning_rate": 7.320629411764706e-05,
+      "loss": 0.6436,
+      "step": 11
+    },
+    {
+      "epoch": 0.10780460415496912,
+      "grad_norm": 3.484093426894694,
+      "learning_rate": 7.986141176470588e-05,
+      "loss": 0.6546,
+      "step": 12
+    },
+    {
+      "epoch": 0.11678832116788321,
+      "grad_norm": 1.7458659318282235,
+      "learning_rate": 8.65165294117647e-05,
+      "loss": 0.6243,
+      "step": 13
+    },
+    {
+      "epoch": 0.1257720381807973,
+      "grad_norm": 2.696049020472868,
+      "learning_rate": 9.317164705882352e-05,
+      "loss": 0.6431,
+      "step": 14
+    },
+    {
+      "epoch": 0.13475575519371139,
+      "grad_norm": 2.7403846964581335,
+      "learning_rate": 9.982676470588235e-05,
+      "loss": 0.6337,
+      "step": 15
+    },
+    {
+      "epoch": 0.1437394722066255,
+      "grad_norm": 1.9984598625107888,
+      "learning_rate": 0.00010648188235294117,
+      "loss": 0.6219,
+      "step": 16
+    },
+    {
+      "epoch": 0.1527231892195396,
+      "grad_norm": 1.8055549976045306,
+      "learning_rate": 0.000113137,
+      "loss": 0.6009,
+      "step": 17
+    },
+    {
+      "epoch": 0.16170690623245368,
+      "grad_norm": 1.5503606010043973,
+      "learning_rate": 0.00011979211764705882,
+      "loss": 0.5964,
+      "step": 18
+    },
+    {
+      "epoch": 0.17069062324536777,
+      "grad_norm": 2.35861768584361,
+      "learning_rate": 0.00012644723529411765,
+      "loss": 0.6148,
+      "step": 19
+    },
+    {
+      "epoch": 0.17967434025828186,
+      "grad_norm": 1.5396308193214678,
+      "learning_rate": 0.00013310235294117646,
+      "loss": 0.6018,
+      "step": 20
+    },
+    {
+      "epoch": 0.18865805727119594,
+      "grad_norm": 1.6143825069098845,
+      "learning_rate": 0.0001397574705882353,
+      "loss": 0.5954,
+      "step": 21
+    },
+    {
+      "epoch": 0.19764177428411006,
+      "grad_norm": 1.6144901661060405,
+      "learning_rate": 0.0001464125882352941,
+      "loss": 0.6004,
+      "step": 22
+    },
+    {
+      "epoch": 0.20662549129702415,
+      "grad_norm": 3.3171185005484856,
+      "learning_rate": 0.00015306770588235295,
+      "loss": 0.6111,
+      "step": 23
+    },
+    {
+      "epoch": 0.21560920830993824,
+      "grad_norm": 2.2411018078854936,
+      "learning_rate": 0.00015972282352941176,
+      "loss": 0.6119,
+      "step": 24
+    },
+    {
+      "epoch": 0.22459292532285233,
+      "grad_norm": 8.161797640058992,
+      "learning_rate": 0.0001663779411764706,
+      "loss": 0.593,
+      "step": 25
+    },
+    {
+      "epoch": 0.23357664233576642,
+      "grad_norm": 846.7701836327712,
+      "learning_rate": 0.0001730330588235294,
+      "loss": 5.673,
+      "step": 26
+    },
+    {
+      "epoch": 0.2425603593486805,
+      "grad_norm": 4.629361945191611,
+      "learning_rate": 0.00017968817647058823,
+      "loss": 0.7795,
+      "step": 27
+    },
+    {
+      "epoch": 0.2515440763615946,
+      "grad_norm": 85.87349809516256,
+      "learning_rate": 0.00018634329411764704,
+      "loss": 0.7672,
+      "step": 28
+    },
+    {
+      "epoch": 0.2605277933745087,
+      "grad_norm": 2.3423299747514212,
+      "learning_rate": 0.00019299841176470588,
+      "loss": 0.6615,
+      "step": 29
+    },
+    {
+      "epoch": 0.26951151038742277,
+      "grad_norm": 2.629813380478519,
+      "learning_rate": 0.0001996535294117647,
+      "loss": 0.6385,
+      "step": 30
+    },
+    {
+      "epoch": 0.2784952274003369,
+      "grad_norm": 1.2741888860891866,
+      "learning_rate": 0.00020630864705882353,
+      "loss": 0.612,
+      "step": 31
+    },
+    {
+      "epoch": 0.287478944413251,
+      "grad_norm": 31.843678922273796,
+      "learning_rate": 0.00021296376470588234,
+      "loss": 0.6641,
+      "step": 32
+    },
+    {
+      "epoch": 0.29646266142616506,
+      "grad_norm": 5.150042714870819,
+      "learning_rate": 0.00021961888235294118,
+      "loss": 0.9157,
+      "step": 33
+    },
+    {
+      "epoch": 0.3054463784390792,
+      "grad_norm": 610.734122596688,
+      "learning_rate": 0.000226274,
+      "loss": 1.7373,
+      "step": 34
+    },
+    {
+      "epoch": 0.31443009545199324,
+      "grad_norm": 46.58948260555198,
+      "learning_rate": 0.00022626775506336304,
+      "loss": 0.8771,
+      "step": 35
+    },
+    {
+      "epoch": 0.32341381246490736,
+      "grad_norm": 6.962976785617738,
+      "learning_rate": 0.00022624902094286824,
+      "loss": 0.9851,
+      "step": 36
+    },
+    {
+      "epoch": 0.3323975294778215,
+      "grad_norm": 21.673571852019826,
+      "learning_rate": 0.00022621779970668783,
+      "loss": 0.9069,
+      "step": 37
+    },
+    {
+      "epoch": 0.34138124649073553,
+      "grad_norm": 2.892556097499312,
+      "learning_rate": 0.00022617409480152153,
+      "loss": 0.8182,
+      "step": 38
+    },
+    {
+      "epoch": 0.35036496350364965,
+      "grad_norm": 15.60543009302285,
+      "learning_rate": 0.00022611791105221654,
+      "loss": 0.9486,
+      "step": 39
+    },
+    {
+      "epoch": 0.3593486805165637,
+      "grad_norm": 11.538473656686941,
+      "learning_rate": 0.00022604925466123451,
+      "loss": 0.8591,
+      "step": 40
+    },
+    {
+      "epoch": 0.36833239752947783,
+      "grad_norm": 2.5218994606008907,
+      "learning_rate": 0.00022596813320796707,
+      "loss": 0.7781,
+      "step": 41
+    },
+    {
+      "epoch": 0.3773161145423919,
+      "grad_norm": 4.662799540456364,
+      "learning_rate": 0.00022587455564789883,
+      "loss": 0.8526,
+      "step": 42
+    },
+    {
+      "epoch": 0.386299831555306,
+      "grad_norm": 565.0656362499724,
+      "learning_rate": 0.00022576853231161902,
+      "loss": 1.738,
+      "step": 43
+    },
+    {
+      "epoch": 0.3952835485682201,
+      "grad_norm": 7.6600755456980325,
+      "learning_rate": 0.00022565007490368076,
+      "loss": 0.8999,
+      "step": 44
+    },
+    {
+      "epoch": 0.4042672655811342,
+      "grad_norm": 287.49302909066023,
+      "learning_rate": 0.00022551919650130918,
+      "loss": 7.1473,
+      "step": 45
+    },
+    {
+      "epoch": 0.4132509825940483,
+      "grad_norm": 13.25846612995419,
+      "learning_rate": 0.00022537591155295756,
+      "loss": 1.5197,
+      "step": 46
+    },
+    {
+      "epoch": 0.42223469960696236,
+      "grad_norm": 6.573021527190113,
+      "learning_rate": 0.00022522023587671233,
+      "loss": 1.0434,
+      "step": 47
+    },
+    {
+      "epoch": 0.4312184166198765,
+      "grad_norm": 2.561584699671566,
+      "learning_rate": 0.0002250521866585469,
+      "loss": 0.8122,
+      "step": 48
+    },
+    {
+      "epoch": 0.4402021336327906,
+      "grad_norm": 12.292027687573608,
+      "learning_rate": 0.00022487178245042422,
+      "loss": 0.9547,
+      "step": 49
+    },
+    {
+      "epoch": 0.44918585064570465,
+      "grad_norm": 14.460331630727685,
+      "learning_rate": 0.0002246790431682489,
+      "loss": 1.026,
+      "step": 50
+    },
+    {
+      "epoch": 0.45816956765861877,
+      "grad_norm": 4.017308882255896,
+      "learning_rate": 0.00022447399008966853,
+      "loss": 0.8958,
+      "step": 51
+    },
+    {
+      "epoch": 0.46715328467153283,
+      "grad_norm": 1.8162087381759398,
+      "learning_rate": 0.0002242566458517245,
+      "loss": 0.7584,
+      "step": 52
+    },
+    {
+      "epoch": 0.47613700168444695,
+      "grad_norm": 2.229007766044214,
+      "learning_rate": 0.0002240270344483534,
+      "loss": 0.7519,
+      "step": 53
+    },
+    {
+      "epoch": 0.485120718697361,
+      "grad_norm": 1.275167700368963,
+      "learning_rate": 0.00022378518122773768,
+      "loss": 0.7169,
+      "step": 54
+    },
+    {
+      "epoch": 0.4941044357102751,
+      "grad_norm": 1.2052448804014857,
+      "learning_rate": 0.00022353111288950776,
+      "loss": 0.7069,
+      "step": 55
+    },
+    {
+      "epoch": 0.5030881527231892,
+      "grad_norm": 2.809114390406924,
+      "learning_rate": 0.00022326485748179416,
+      "loss": 0.6934,
+      "step": 56
+    },
+    {
+      "epoch": 0.5120718697361033,
+      "grad_norm": 1.7295819486463446,
+      "learning_rate": 0.00022298644439813125,
+      "loss": 0.6997,
+      "step": 57
+    },
+    {
+      "epoch": 0.5210555867490174,
+      "grad_norm": 1.176047538360094,
+      "learning_rate": 0.00022269590437421234,
+      "loss": 0.6587,
+      "step": 58
+    },
+    {
+      "epoch": 0.5300393037619315,
+      "grad_norm": 0.9770345550489453,
+      "learning_rate": 0.0002223932694844966,
+      "loss": 0.6444,
+      "step": 59
+    },
+    {
+      "epoch": 0.5390230207748455,
+      "grad_norm": 1.496641892403964,
+      "learning_rate": 0.00022207857313866796,
+      "loss": 0.6558,
+      "step": 60
+    },
+    {
+      "epoch": 0.5480067377877597,
+      "grad_norm": 1.7254473728830027,
+      "learning_rate": 0.00022175185007794712,
+      "loss": 0.6647,
+      "step": 61
+    },
+    {
+      "epoch": 0.5569904548006738,
+      "grad_norm": 1.148244331773896,
+      "learning_rate": 0.000221413136371256,
+      "loss": 0.6244,
+      "step": 62
+    },
+    {
+      "epoch": 0.5659741718135879,
+      "grad_norm": 1.082240763317917,
+      "learning_rate": 0.00022106246941123604,
+      "loss": 0.616,
+      "step": 63
+    },
+    {
+      "epoch": 0.574957888826502,
+      "grad_norm": 1.0463382313968703,
+      "learning_rate": 0.0002206998879101201,
+      "loss": 0.6097,
+      "step": 64
+    },
+    {
+      "epoch": 0.583941605839416,
+      "grad_norm": 1.5314621991483217,
+      "learning_rate": 0.00022032543189545893,
+      "loss": 0.6261,
+      "step": 65
+    },
+    {
+      "epoch": 0.5929253228523301,
+      "grad_norm": 0.8440358318436397,
+      "learning_rate": 0.00021993914270570204,
+      "loss": 0.5992,
+      "step": 66
+    },
+    {
+      "epoch": 0.6019090398652442,
+      "grad_norm": 1.1721068603539129,
+      "learning_rate": 0.0002195410629856343,
+      "loss": 0.6003,
+      "step": 67
+    },
+    {
+      "epoch": 0.6108927568781584,
+      "grad_norm": 1.0225694078862206,
+      "learning_rate": 0.00021913123668166815,
+      "loss": 0.5952,
+      "step": 68
+    },
+    {
+      "epoch": 0.6198764738910725,
+      "grad_norm": 1.223954503187751,
+      "learning_rate": 0.00021870970903699184,
+      "loss": 0.5844,
+      "step": 69
+    },
+    {
+      "epoch": 0.6288601909039865,
+      "grad_norm": 0.7517434069071857,
+      "learning_rate": 0.00021827652658657518,
+      "loss": 0.5736,
+      "step": 70
+    },
+    {
+      "epoch": 0.6378439079169006,
+      "grad_norm": 0.542782981372059,
+      "learning_rate": 0.00021783173715203175,
+      "loss": 0.5677,
+      "step": 71
+    },
+    {
+      "epoch": 0.6468276249298147,
+      "grad_norm": 0.8491645074764822,
+      "learning_rate": 0.00021737538983634003,
+      "loss": 0.5567,
+      "step": 72
+    },
+    {
+      "epoch": 0.6558113419427288,
+      "grad_norm": 0.9512722570962481,
+      "learning_rate": 0.00021690753501842243,
+      "loss": 0.5652,
+      "step": 73
+    },
+    {
+      "epoch": 0.664795058955643,
+      "grad_norm": 0.7420499518682178,
+      "learning_rate": 0.00021642822434758365,
+      "loss": 0.5521,
+      "step": 74
+    },
+    {
+      "epoch": 0.673778775968557,
+      "grad_norm": 0.6023655209710193,
+      "learning_rate": 0.0002159375107378088,
+      "loss": 0.5484,
+      "step": 75
+    },
+    {
+      "epoch": 0.6827624929814711,
+      "grad_norm": 0.6704929086374853,
+      "learning_rate": 0.00021543544836192198,
+      "loss": 0.5629,
+      "step": 76
+    },
+    {
+      "epoch": 0.6917462099943852,
+      "grad_norm": 0.8511116323519582,
+      "learning_rate": 0.00021492209264560583,
+      "loss": 0.5477,
+      "step": 77
+    },
+    {
+      "epoch": 0.7007299270072993,
+      "grad_norm": 0.8293067237834507,
+      "learning_rate": 0.00021439750026128255,
+      "loss": 0.5486,
+      "step": 78
+    },
+    {
+      "epoch": 0.7097136440202133,
+      "grad_norm": 0.7951812792060676,
+      "learning_rate": 0.00021386172912185786,
+      "loss": 0.5386,
+      "step": 79
+    },
+    {
+      "epoch": 0.7186973610331274,
+      "grad_norm": 0.5942026077567073,
+      "learning_rate": 0.0002133148383743272,
+      "loss": 0.5526,
+      "step": 80
+    },
+    {
+      "epoch": 0.7276810780460415,
+      "grad_norm": 0.7074314677815655,
+      "learning_rate": 0.00021275688839324653,
+      "loss": 0.5355,
+      "step": 81
+    },
+    {
+      "epoch": 0.7366647950589557,
+      "grad_norm": 1.0064759251882633,
+      "learning_rate": 0.00021218794077406699,
+      "loss": 0.5214,
+      "step": 82
+    },
+    {
+      "epoch": 0.7456485120718698,
+      "grad_norm": 0.5936960428857333,
+      "learning_rate": 0.00021160805832633507,
+      "loss": 0.5334,
+      "step": 83
+    },
+    {
+      "epoch": 0.7546322290847838,
+      "grad_norm": 0.5587606186465214,
+      "learning_rate": 0.00021101730506675872,
+      "loss": 0.5246,
+      "step": 84
+    },
+    {
+      "epoch": 0.7636159460976979,
+      "grad_norm": 0.7563054405639342,
+      "learning_rate": 0.00021041574621214016,
+      "loss": 0.5211,
+      "step": 85
+    },
+    {
+      "epoch": 0.772599663110612,
+      "grad_norm": 0.4650706352210191,
+      "learning_rate": 0.00020980344817217607,
+      "loss": 0.52,
+      "step": 86
+    },
+    {
+      "epoch": 0.7815833801235261,
+      "grad_norm": 0.46292748233999875,
+      "learning_rate": 0.00020918047854212644,
+      "loss": 0.5086,
+      "step": 87
+    },
+    {
+      "epoch": 0.7905670971364402,
+      "grad_norm": 0.45822776585768843,
+      "learning_rate": 0.00020854690609535217,
+      "loss": 0.5205,
+      "step": 88
+    },
+    {
+      "epoch": 0.7995508141493542,
+      "grad_norm": 0.3680018488886641,
+      "learning_rate": 0.0002079028007757228,
+      "loss": 0.4995,
+      "step": 89
+    },
+    {
+      "epoch": 0.8085345311622684,
+      "grad_norm": 0.3651907315130092,
+      "learning_rate": 0.000207248233689895,
+      "loss": 0.5108,
+      "step": 90
+    },
+    {
+      "epoch": 0.8175182481751825,
+      "grad_norm": 0.38739099960496165,
+      "learning_rate": 0.00020658327709946274,
+      "loss": 0.5105,
+      "step": 91
+    },
+    {
+      "epoch": 0.8265019651880966,
+      "grad_norm": 0.37587537297348633,
+      "learning_rate": 0.00020590800441297976,
+      "loss": 0.5043,
+      "step": 92
+    },
+    {
+      "epoch": 0.8354856822010107,
+      "grad_norm": 0.5725177706736702,
+      "learning_rate": 0.00020522249017785566,
+      "loss": 0.5056,
+      "step": 93
+    },
+    {
+      "epoch": 0.8444693992139247,
+      "grad_norm": 1.0428309894412633,
+      "learning_rate": 0.00020452681007212612,
+      "loss": 0.5189,
+      "step": 94
+    },
+    {
+      "epoch": 0.8534531162268388,
+      "grad_norm": 1.1604159616874143,
+      "learning_rate": 0.0002038210408960984,
+      "loss": 0.5167,
+      "step": 95
+    },
+    {
+      "epoch": 0.862436833239753,
+      "grad_norm": 0.5208537644478249,
+      "learning_rate": 0.0002031052605638728,
+      "loss": 0.5145,
+      "step": 96
+    },
+    {
+      "epoch": 0.8714205502526671,
+      "grad_norm": 0.83804291321802,
+      "learning_rate": 0.00020237954809474134,
+      "loss": 0.5158,
+      "step": 97
+    },
+    {
+      "epoch": 0.8804042672655812,
+      "grad_norm": 0.7064273962256518,
+      "learning_rate": 0.00020164398360446436,
+      "loss": 0.4931,
+      "step": 98
+    },
+    {
+      "epoch": 0.8893879842784952,
+      "grad_norm": 0.5057800234712763,
+      "learning_rate": 0.00020089864829642596,
+      "loss": 0.5075,
+      "step": 99
+    },
+    {
+      "epoch": 0.8983717012914093,
+      "grad_norm": 0.5940667934373025,
+      "learning_rate": 0.0002001436244526695,
+      "loss": 0.4907,
+      "step": 100
+    },
+    {
+      "epoch": 0.9073554183043234,
+      "grad_norm": 0.5019525607511636,
+      "learning_rate": 0.00019937899542481408,
+      "loss": 0.5041,
+      "step": 101
+    },
+    {
+      "epoch": 0.9163391353172375,
+      "grad_norm": 0.4491147515500972,
+      "learning_rate": 0.00019860484562485276,
+      "loss": 0.488,
+      "step": 102
+    },
+    {
+      "epoch": 0.9253228523301515,
+      "grad_norm": 0.5412361814676229,
+      "learning_rate": 0.00019782126051583386,
+      "loss": 0.5023,
+      "step": 103
+    },
+    {
+      "epoch": 0.9343065693430657,
+      "grad_norm": 0.3383672378854014,
+      "learning_rate": 0.00019702832660242615,
+      "loss": 0.489,
+      "step": 104
+    },
+    {
+      "epoch": 0.9432902863559798,
+      "grad_norm": 0.36276664892975297,
+      "learning_rate": 0.0001962261314213691,
+      "loss": 0.4996,
+      "step": 105
+    },
+    {
+      "epoch": 0.9522740033688939,
+      "grad_norm": 0.3454995598586516,
+      "learning_rate": 0.00019541476353180916,
+      "loss": 0.4856,
+      "step": 106
+    },
+    {
+      "epoch": 0.961257720381808,
+      "grad_norm": 0.311134046456636,
+      "learning_rate": 0.00019459431250552317,
+      "loss": 0.4855,
+      "step": 107
+    },
+    {
+      "epoch": 0.970241437394722,
+      "grad_norm": 0.3425949299888489,
+      "learning_rate": 0.0001937648689170301,
+      "loss": 0.4762,
+      "step": 108
+    },
+    {
+      "epoch": 0.9792251544076361,
+      "grad_norm": 0.3496537132478162,
+      "learning_rate": 0.00019292652433359177,
+      "loss": 0.4834,
+      "step": 109
+    },
+    {
+      "epoch": 0.9882088714205502,
+      "grad_norm": 0.4269583108093911,
+      "learning_rate": 0.00019207937130510442,
+      "loss": 0.4946,
+      "step": 110
+    },
+    {
+      "epoch": 0.9971925884334644,
+      "grad_norm": 0.4147252598563634,
+      "learning_rate": 0.0001912235033538814,
+      "loss": 0.4846,
+      "step": 111
+    },
+    {
+      "epoch": 1.0075800112296462,
+      "grad_norm": 0.8355251290758435,
+      "learning_rate": 0.00019035901496432887,
+      "loss": 0.8602,
+      "step": 112
+    },
+    {
+      "epoch": 1.0165637282425604,
+      "grad_norm": 1.6156052899406323,
+      "learning_rate": 0.0001894860015725148,
+      "loss": 0.4935,
+      "step": 113
+    },
+    {
+      "epoch": 1.0255474452554745,
+      "grad_norm": 0.46493272524584955,
+      "learning_rate": 0.0001886045595556335,
+      "loss": 0.4655,
+      "step": 114
+    },
+    {
+      "epoch": 1.0345311622683886,
+      "grad_norm": 0.979120652286332,
+      "learning_rate": 0.00018771478622136586,
+      "loss": 0.4778,
+      "step": 115
+    },
+    {
+      "epoch": 1.0435148792813027,
+      "grad_norm": 0.7656350182275576,
+      "learning_rate": 0.0001868167797971369,
+      "loss": 0.4786,
+      "step": 116
+    },
+    {
+      "epoch": 1.0524985962942168,
+      "grad_norm": 0.5200988865180658,
+      "learning_rate": 0.00018591063941927197,
+      "loss": 0.4667,
+      "step": 117
+    },
+    {
+      "epoch": 1.0614823133071307,
+      "grad_norm": 0.5373218112082903,
+      "learning_rate": 0.00018499646512205238,
+      "loss": 0.4698,
+      "step": 118
+    },
+    {
+      "epoch": 1.0704660303200448,
+      "grad_norm": 0.3981583489798308,
+      "learning_rate": 0.00018407435782667207,
+      "loss": 0.4573,
+      "step": 119
+    },
+    {
+      "epoch": 1.079449747332959,
+      "grad_norm": 0.4729814422997812,
+      "learning_rate": 0.0001831444193300964,
+      "loss": 0.4537,
+      "step": 120
+    },
+    {
+      "epoch": 1.088433464345873,
+      "grad_norm": 0.40160292629773175,
+      "learning_rate": 0.00018220675229382407,
+      "loss": 0.4607,
+      "step": 121
+    },
+    {
+      "epoch": 1.0974171813587872,
+      "grad_norm": 0.34191199159101826,
+      "learning_rate": 0.0001812614602325536,
+      "loss": 0.4491,
+      "step": 122
+    },
+    {
+      "epoch": 1.1064008983717013,
+      "grad_norm": 0.38115499724656887,
+      "learning_rate": 0.00018030864750275598,
+      "loss": 0.4649,
+      "step": 123
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.2939815268385644,
+      "learning_rate": 0.0001793484192911539,
+      "loss": 0.4405,
+      "step": 124
+    },
+    {
+      "epoch": 1.1243683323975295,
+      "grad_norm": 0.38182282790459066,
+      "learning_rate": 0.00017838088160310975,
+      "loss": 0.4673,
+      "step": 125
+    },
+    {
+      "epoch": 1.1333520494104437,
+      "grad_norm": 0.26951517499854655,
+      "learning_rate": 0.00017740614125092276,
+      "loss": 0.4366,
+      "step": 126
+    },
+    {
+      "epoch": 1.1423357664233578,
+      "grad_norm": 0.32986266756899896,
+      "learning_rate": 0.00017642430584203776,
+      "loss": 0.4506,
+      "step": 127
+    },
+    {
+      "epoch": 1.1513194834362717,
+      "grad_norm": 0.29552314624204684,
+      "learning_rate": 0.0001754354837671654,
+      "loss": 0.4409,
+      "step": 128
+    },
+    {
+      "epoch": 1.1603032004491858,
+      "grad_norm": 0.2715172818791699,
+      "learning_rate": 0.0001744397841883164,
+      "loss": 0.455,
+      "step": 129
+    },
+    {
+      "epoch": 1.1692869174621,
+      "grad_norm": 0.30375726886135684,
+      "learning_rate": 0.00017343731702675036,
+      "loss": 0.4453,
+      "step": 130
+    },
+    {
+      "epoch": 1.178270634475014,
+      "grad_norm": 0.20896293317807582,
+      "learning_rate": 0.00017242819295084123,
+      "loss": 0.4447,
+      "step": 131
+    },
+    {
+      "epoch": 1.1872543514879281,
+      "grad_norm": 0.26828433742607094,
+      "learning_rate": 0.0001714125233638595,
+      "loss": 0.444,
+      "step": 132
+    },
+    {
+      "epoch": 1.1962380685008422,
+      "grad_norm": 0.21271544817391416,
+      "learning_rate": 0.00017039042039167416,
+      "loss": 0.4434,
+      "step": 133
+    },
+    {
+      "epoch": 1.2052217855137564,
+      "grad_norm": 0.22231269773708437,
+      "learning_rate": 0.00016936199687037412,
+      "loss": 0.4392,
+      "step": 134
+    },
+    {
+      "epoch": 1.2142055025266705,
+      "grad_norm": 0.21858395274334466,
+      "learning_rate": 0.00016832736633381174,
+      "loss": 0.44,
+      "step": 135
+    },
+    {
+      "epoch": 1.2231892195395846,
+      "grad_norm": 0.2359281892515482,
+      "learning_rate": 0.00016728664300106905,
+      "loss": 0.4395,
+      "step": 136
+    },
+    {
+      "epoch": 1.2321729365524985,
+      "grad_norm": 0.21457493342859105,
+      "learning_rate": 0.0001662399417638485,
+      "loss": 0.442,
+      "step": 137
+    },
+    {
+      "epoch": 1.2411566535654126,
+      "grad_norm": 0.17832258654038546,
+      "learning_rate": 0.00016518737817378936,
+      "loss": 0.4379,
+      "step": 138
+    },
+    {
+      "epoch": 1.2501403705783267,
+      "grad_norm": 0.20100339820701033,
+      "learning_rate": 0.0001641290684297112,
+      "loss": 0.4371,
+      "step": 139
+    },
+    {
+      "epoch": 1.2591240875912408,
+      "grad_norm": 0.23303786372370514,
+      "learning_rate": 0.00016306512936478605,
+      "loss": 0.4353,
+      "step": 140
+    },
+    {
+      "epoch": 1.268107804604155,
+      "grad_norm": 0.25687763850096407,
+      "learning_rate": 0.00016199567843364066,
+      "loss": 0.4379,
+      "step": 141
+    },
+    {
+      "epoch": 1.277091521617069,
+      "grad_norm": 0.23974642432643137,
+      "learning_rate": 0.00016092083369938952,
+      "loss": 0.4364,
+      "step": 142
+    },
+    {
+      "epoch": 1.2860752386299832,
+      "grad_norm": 0.20652369486142858,
+      "learning_rate": 0.00015984071382060174,
+      "loss": 0.4385,
+      "step": 143
+    },
+    {
+      "epoch": 1.2950589556428973,
+      "grad_norm": 0.20585656880887795,
+      "learning_rate": 0.00015875543803820115,
+      "loss": 0.4425,
+      "step": 144
+    },
+    {
+      "epoch": 1.3040426726558114,
+      "grad_norm": 0.22079211853825997,
+      "learning_rate": 0.00015766512616230286,
+      "loss": 0.4251,
+      "step": 145
+    },
+    {
+      "epoch": 1.3130263896687255,
+      "grad_norm": 0.19534999766663985,
+      "learning_rate": 0.00015656989855898653,
+      "loss": 0.4374,
+      "step": 146
+    },
+    {
+      "epoch": 1.3220101066816397,
+      "grad_norm": 0.19699056774469947,
+      "learning_rate": 0.00015546987613700858,
+      "loss": 0.4385,
+      "step": 147
+    },
+    {
+      "epoch": 1.3309938236945535,
+      "grad_norm": 0.25025460629960233,
+      "learning_rate": 0.00015436518033445427,
+      "loss": 0.4414,
+      "step": 148
+    },
+    {
+      "epoch": 1.3399775407074677,
+      "grad_norm": 0.22783106244726478,
+      "learning_rate": 0.00015325593310533135,
+      "loss": 0.4292,
+      "step": 149
+    },
+    {
+      "epoch": 1.3489612577203818,
+      "grad_norm": 0.24800972153045137,
+      "learning_rate": 0.00015214225690610695,
+      "loss": 0.4367,
+      "step": 150
+    },
+    {
+      "epoch": 1.357944974733296,
+      "grad_norm": 0.2755781135491422,
+      "learning_rate": 0.00015102427468218884,
+      "loss": 0.4343,
+      "step": 151
+    },
+    {
+      "epoch": 1.36692869174621,
+      "grad_norm": 0.3302080679814042,
+      "learning_rate": 0.00014990210985435274,
+      "loss": 0.4225,
+      "step": 152
+    },
+    {
+      "epoch": 1.3759124087591241,
+      "grad_norm": 0.3884646633170279,
+      "learning_rate": 0.00014877588630511702,
+      "loss": 0.4291,
+      "step": 153
+    },
+    {
+      "epoch": 1.3848961257720382,
+      "grad_norm": 0.4487151204051441,
+      "learning_rate": 0.0001476457283650669,
+      "loss": 0.4314,
+      "step": 154
+    },
+    {
+      "epoch": 1.3938798427849521,
+      "grad_norm": 0.4142847149164689,
+      "learning_rate": 0.00014651176079912844,
+      "loss": 0.4248,
+      "step": 155
+    },
+    {
+      "epoch": 1.4028635597978663,
+      "grad_norm": 0.3852901369815094,
+      "learning_rate": 0.00014537410879279536,
+      "loss": 0.4233,
+      "step": 156
+    },
+    {
+      "epoch": 1.4118472768107804,
+      "grad_norm": 0.364518339635796,
+      "learning_rate": 0.00014423289793830883,
+      "loss": 0.4362,
+      "step": 157
+    },
+    {
+      "epoch": 1.4208309938236945,
+      "grad_norm": 0.35417387758842084,
+      "learning_rate": 0.00014308825422079265,
+      "loss": 0.4252,
+      "step": 158
+    },
+    {
+      "epoch": 1.4298147108366086,
+      "grad_norm": 0.3557663368174752,
+      "learning_rate": 0.00014194030400434496,
+      "loss": 0.4395,
+      "step": 159
+    },
+    {
+      "epoch": 1.4387984278495227,
+      "grad_norm": 0.34124968887701046,
+      "learning_rate": 0.00014078917401808824,
+      "loss": 0.4355,
+      "step": 160
+    },
+    {
+      "epoch": 1.4477821448624368,
+      "grad_norm": 0.253304289924485,
+      "learning_rate": 0.0001396349913421788,
+      "loss": 0.4355,
+      "step": 161
+    },
+    {
+      "epoch": 1.456765861875351,
+      "grad_norm": 0.2979898627616292,
+      "learning_rate": 0.00013847788339377788,
+      "loss": 0.4411,
+      "step": 162
+    },
+    {
+      "epoch": 1.465749578888265,
+      "grad_norm": 0.2927807074804572,
+      "learning_rate": 0.0001373179779129849,
+      "loss": 0.434,
+      "step": 163
+    },
+    {
+      "epoch": 1.4747332959011792,
+      "grad_norm": 0.23083501058351574,
+      "learning_rate": 0.00013615540294873585,
+      "loss": 0.4296,
+      "step": 164
+    },
+    {
+      "epoch": 1.4837170129140933,
+      "grad_norm": 0.25440061192162694,
+      "learning_rate": 0.00013499028684466692,
+      "loss": 0.4316,
+      "step": 165
+    },
+    {
+      "epoch": 1.4927007299270074,
+      "grad_norm": 0.3151406421517585,
+      "learning_rate": 0.00013382275822494612,
+      "loss": 0.4268,
+      "step": 166
+    },
+    {
+      "epoch": 1.5016844469399215,
+      "grad_norm": 0.28632589462844177,
+      "learning_rate": 0.00013265294598007347,
+      "loss": 0.4338,
+      "step": 167
+    },
+    {
+      "epoch": 1.5106681639528357,
+      "grad_norm": 0.20398795355439606,
+      "learning_rate": 0.00013148097925265212,
+      "loss": 0.4335,
+      "step": 168
+    },
+    {
+      "epoch": 1.5196518809657495,
+      "grad_norm": 0.3190354573542016,
+      "learning_rate": 0.00013030698742313152,
+      "loss": 0.4206,
+      "step": 169
+    },
+    {
+      "epoch": 1.5286355979786637,
+      "grad_norm": 0.3741573095309751,
+      "learning_rate": 0.00012913110009552428,
+      "loss": 0.4322,
+      "step": 170
+    },
+    {
+      "epoch": 1.5376193149915778,
+      "grad_norm": 0.2617731827572085,
+      "learning_rate": 0.0001279534470830984,
+      "loss": 0.4282,
+      "step": 171
+    },
+    {
+      "epoch": 1.546603032004492,
+      "grad_norm": 0.31625614240914174,
+      "learning_rate": 0.00012677415839404646,
+      "loss": 0.4315,
+      "step": 172
+    },
+    {
+      "epoch": 1.5555867490174058,
+      "grad_norm": 0.3559962927646376,
+      "learning_rate": 0.00012559336421713333,
+      "loss": 0.4255,
+      "step": 173
+    },
+    {
+      "epoch": 1.56457046603032,
+      "grad_norm": 0.23926556769719626,
+      "learning_rate": 0.00012441119490732357,
+      "loss": 0.4197,
+      "step": 174
+    },
+    {
+      "epoch": 1.573554183043234,
+      "grad_norm": 0.19009337870934856,
+      "learning_rate": 0.00012322778097139102,
+      "loss": 0.4333,
+      "step": 175
+    },
+    {
+      "epoch": 1.5825379000561481,
+      "grad_norm": 0.2467283593538461,
+      "learning_rate": 0.00012204325305351117,
+      "loss": 0.4327,
+      "step": 176
+    },
+    {
+      "epoch": 1.5915216170690623,
+      "grad_norm": 0.20155153552210148,
+      "learning_rate": 0.00012085774192083878,
+      "loss": 0.424,
+      "step": 177
+    },
+    {
+      "epoch": 1.6005053340819764,
+      "grad_norm": 0.2055516806210948,
+      "learning_rate": 0.00011967137844907157,
+      "loss": 0.4198,
+      "step": 178
+    },
+    {
+      "epoch": 1.6094890510948905,
+      "grad_norm": 0.21888134195759792,
+      "learning_rate": 0.00011848429360800205,
+      "loss": 0.4279,
+      "step": 179
+    },
+    {
+      "epoch": 1.6184727681078046,
+      "grad_norm": 0.18186305811122733,
+      "learning_rate": 0.00011729661844705912,
+      "loss": 0.4249,
+      "step": 180
+    },
+    {
+      "epoch": 1.6274564851207187,
+      "grad_norm": 0.13875458201438734,
+      "learning_rate": 0.00011610848408084054,
+      "loss": 0.4275,
+      "step": 181
+    },
+    {
+      "epoch": 1.6364402021336328,
+      "grad_norm": 0.1717023475061491,
+      "learning_rate": 0.0001149200216746385,
+      "loss": 0.4225,
+      "step": 182
+    },
+    {
+      "epoch": 1.645423919146547,
+      "grad_norm": 0.15952136716834545,
+      "learning_rate": 0.00011373136242995958,
+      "loss": 0.4153,
+      "step": 183
+    },
+    {
+      "epoch": 1.654407636159461,
+      "grad_norm": 0.15537668646212147,
+      "learning_rate": 0.00011254263757004044,
+      "loss": 0.4164,
+      "step": 184
+    },
+    {
+      "epoch": 1.6633913531723752,
+      "grad_norm": 0.160282534660321,
+      "learning_rate": 0.00011135397832536151,
+      "loss": 0.4253,
+      "step": 185
+    },
+    {
+      "epoch": 1.6723750701852893,
+      "grad_norm": 0.16123326071336702,
+      "learning_rate": 0.00011016551591915949,
+      "loss": 0.4239,
+      "step": 186
+    },
+    {
+      "epoch": 1.6813587871982034,
+      "grad_norm": 0.148970328863511,
+      "learning_rate": 0.0001089773815529409,
+      "loss": 0.4245,
+      "step": 187
+    },
+    {
+      "epoch": 1.6903425042111173,
+      "grad_norm": 0.16045971509449883,
+      "learning_rate": 0.00010778970639199795,
+      "loss": 0.4198,
+      "step": 188
+    },
+    {
+      "epoch": 1.6993262212240314,
+      "grad_norm": 0.16138201760304008,
+      "learning_rate": 0.00010660262155092845,
+      "loss": 0.4275,
+      "step": 189
+    },
+    {
+      "epoch": 1.7083099382369455,
+      "grad_norm": 0.16720529564166428,
+      "learning_rate": 0.00010541625807916123,
+      "loss": 0.4175,
+      "step": 190
+    },
+    {
+      "epoch": 1.7172936552498597,
+      "grad_norm": 0.16645688605969786,
+      "learning_rate": 0.00010423074694648884,
+      "loss": 0.4272,
+      "step": 191
+    },
+    {
+      "epoch": 1.7262773722627736,
+      "grad_norm": 0.23842852394974162,
+      "learning_rate": 0.00010304621902860899,
+      "loss": 0.4274,
+      "step": 192
+    },
+    {
+      "epoch": 1.7352610892756877,
+      "grad_norm": 0.23233366002775901,
+      "learning_rate": 0.00010186280509267643,
+      "loss": 0.4275,
+      "step": 193
+    },
+    {
+      "epoch": 1.7442448062886018,
+      "grad_norm": 0.16088479076799017,
+      "learning_rate": 0.00010068063578286667,
+      "loss": 0.4172,
+      "step": 194
+    },
+    {
+      "epoch": 1.753228523301516,
+      "grad_norm": 0.1990526955294185,
+      "learning_rate": 9.949984160595352e-05,
+      "loss": 0.4279,
+      "step": 195
+    },
+    {
+      "epoch": 1.76221224031443,
+      "grad_norm": 0.2080695351190495,
+      "learning_rate": 9.832055291690164e-05,
+      "loss": 0.4315,
+      "step": 196
+    },
+    {
+      "epoch": 1.7711959573273441,
+      "grad_norm": 0.1890634090210766,
+      "learning_rate": 9.714289990447578e-05,
+      "loss": 0.4182,
+      "step": 197
+    },
+    {
+      "epoch": 1.7801796743402583,
+      "grad_norm": 0.20016634340121803,
+      "learning_rate": 9.596701257686851e-05,
+      "loss": 0.4119,
+      "step": 198
+    },
+    {
+      "epoch": 1.7891633913531724,
+      "grad_norm": 0.23107873036017035,
+      "learning_rate": 9.479302074734792e-05,
+      "loss": 0.4162,
+      "step": 199
+    },
+    {
+      "epoch": 1.7981471083660865,
+      "grad_norm": 0.1830895281774673,
+      "learning_rate": 9.362105401992656e-05,
+      "loss": 0.4177,
+      "step": 200
+    },
+    {
+      "epoch": 1.8071308253790006,
+      "grad_norm": 0.16770209709045605,
+      "learning_rate": 9.245124177505392e-05,
+      "loss": 0.421,
+      "step": 201
+    },
+    {
+      "epoch": 1.8161145423919147,
+      "grad_norm": 0.2297597515397976,
+      "learning_rate": 9.12837131553331e-05,
+      "loss": 0.434,
+      "step": 202
+    },
+    {
+      "epoch": 1.8250982594048288,
+      "grad_norm": 0.16824933783531398,
+      "learning_rate": 9.011859705126419e-05,
+      "loss": 0.418,
+      "step": 203
+    },
+    {
+      "epoch": 1.834081976417743,
+      "grad_norm": 0.16377185436749506,
+      "learning_rate": 8.895602208701511e-05,
+      "loss": 0.4113,
+      "step": 204
+    },
+    {
+      "epoch": 1.843065693430657,
+      "grad_norm": 0.22604619484876676,
+      "learning_rate": 8.779611660622215e-05,
+      "loss": 0.4239,
+      "step": 205
+    },
+    {
+      "epoch": 1.8520494104435712,
+      "grad_norm": 0.2215922382832308,
+      "learning_rate": 8.66390086578212e-05,
+      "loss": 0.4121,
+      "step": 206
+    },
+    {
+      "epoch": 1.861033127456485,
+      "grad_norm": 0.20334095034573255,
+      "learning_rate": 8.548482598191181e-05,
+      "loss": 0.4193,
+      "step": 207
+    },
+    {
+      "epoch": 1.8700168444693992,
+      "grad_norm": 0.1829539632348045,
+      "learning_rate": 8.433369599565508e-05,
+      "loss": 0.4296,
+      "step": 208
+    },
+    {
+      "epoch": 1.8790005614823133,
+      "grad_norm": 0.15716250429615242,
+      "learning_rate": 8.318574577920736e-05,
+      "loss": 0.4164,
+      "step": 209
+    },
+    {
+      "epoch": 1.8879842784952274,
+      "grad_norm": 0.23307948623719757,
+      "learning_rate": 8.204110206169114e-05,
+      "loss": 0.4172,
+      "step": 210
+    },
+    {
+      "epoch": 1.8969679955081415,
+      "grad_norm": 0.1761132573267027,
+      "learning_rate": 8.08998912072046e-05,
+      "loss": 0.4167,
+      "step": 211
+    },
+    {
+      "epoch": 1.9059517125210554,
+      "grad_norm": 0.1690942036116379,
+      "learning_rate": 7.976223920087153e-05,
+      "loss": 0.4247,
+      "step": 212
+    },
+    {
+      "epoch": 1.9149354295339696,
+      "grad_norm": 0.17381964913986805,
+      "learning_rate": 7.86282716349331e-05,
+      "loss": 0.4076,
+      "step": 213
+    },
+    {
+      "epoch": 1.9239191465468837,
+      "grad_norm": 0.1558189761050885,
+      "learning_rate": 7.749811369488296e-05,
+      "loss": 0.425,
+      "step": 214
+    },
+    {
+      "epoch": 1.9329028635597978,
+      "grad_norm": 0.16691148084521792,
+      "learning_rate": 7.637189014564727e-05,
+      "loss": 0.4152,
+      "step": 215
+    },
+    {
+      "epoch": 1.941886580572712,
+      "grad_norm": 0.14084306621184936,
+      "learning_rate": 7.524972531781114e-05,
+      "loss": 0.4201,
+      "step": 216
+    },
+    {
+      "epoch": 1.950870297585626,
+      "grad_norm": 0.1673195421371356,
+      "learning_rate": 7.413174309389303e-05,
+      "loss": 0.4084,
+      "step": 217
+    },
+    {
+      "epoch": 1.9598540145985401,
+      "grad_norm": 0.16765695738584874,
+      "learning_rate": 7.301806689466864e-05,
+      "loss": 0.4275,
+      "step": 218
+    },
+    {
+      "epoch": 1.9688377316114543,
+      "grad_norm": 0.16869078852623143,
+      "learning_rate": 7.190881966554573e-05,
+      "loss": 0.418,
+      "step": 219
+    },
+    {
+      "epoch": 1.9778214486243684,
+      "grad_norm": 0.16808033042759418,
+      "learning_rate": 7.080412386299138e-05,
+      "loss": 0.4141,
+      "step": 220
+    },
+    {
+      "epoch": 1.9868051656372825,
+      "grad_norm": 0.19910528453140844,
+      "learning_rate": 6.970410144101348e-05,
+      "loss": 0.4029,
+      "step": 221
+    },
+    {
+      "epoch": 1.9957888826501966,
+      "grad_norm": 0.20801296591215268,
+      "learning_rate": 6.860887383769717e-05,
+      "loss": 0.4266,
+      "step": 222
+    },
+    {
+      "epoch": 2.0061763054463784,
+      "grad_norm": 0.3587587352425674,
+      "learning_rate": 6.751856196179887e-05,
+      "loss": 0.7214,
+      "step": 223
+    },
+    {
+      "epoch": 2.0151600224592925,
+      "grad_norm": 0.387007013989843,
+      "learning_rate": 6.643328617939827e-05,
+      "loss": 0.3541,
+      "step": 224
+    },
+    {
+      "epoch": 2.0241437394722066,
+      "grad_norm": 0.27250154771422735,
+      "learning_rate": 6.53531663006105e-05,
+      "loss": 0.3741,
+      "step": 225
+    },
+    {
+      "epoch": 2.0331274564851207,
+      "grad_norm": 0.31369626134171097,
+      "learning_rate": 6.427832156635937e-05,
+      "loss": 0.3623,
+      "step": 226
+    },
+    {
+      "epoch": 2.042111173498035,
+      "grad_norm": 0.3899920698764709,
+      "learning_rate": 6.320887063521393e-05,
+      "loss": 0.3666,
+      "step": 227
+    },
+    {
+      "epoch": 2.051094890510949,
+      "grad_norm": 0.22822552402245974,
+      "learning_rate": 6.214493157028881e-05,
+      "loss": 0.3588,
+      "step": 228
+    },
+    {
+      "epoch": 2.060078607523863,
+      "grad_norm": 0.3578216864292053,
+      "learning_rate": 6.108662182621064e-05,
+      "loss": 0.3734,
+      "step": 229
+    },
+    {
+      "epoch": 2.069062324536777,
+      "grad_norm": 0.25875022892169425,
+      "learning_rate": 6.003405823615149e-05,
+      "loss": 0.3595,
+      "step": 230
+    },
+    {
+      "epoch": 2.0780460415496913,
+      "grad_norm": 0.2595984420837872,
+      "learning_rate": 5.898735699893096e-05,
+      "loss": 0.3605,
+      "step": 231
+    },
+    {
+      "epoch": 2.0870297585626054,
+      "grad_norm": 0.298435693379116,
+      "learning_rate": 5.794663366618828e-05,
+      "loss": 0.3632,
+      "step": 232
+    },
+    {
+      "epoch": 2.0960134755755195,
+      "grad_norm": 0.2237606329851956,
+      "learning_rate": 5.691200312962588e-05,
+      "loss": 0.3498,
+      "step": 233
+    },
+    {
+      "epoch": 2.1049971925884337,
+      "grad_norm": 0.26971282475418673,
+      "learning_rate": 5.5883579608325826e-05,
+      "loss": 0.3629,
+      "step": 234
+    },
+    {
+      "epoch": 2.1139809096013478,
+      "grad_norm": 0.2272539853787037,
+      "learning_rate": 5.486147663614048e-05,
+      "loss": 0.3616,
+      "step": 235
+    },
+    {
+      "epoch": 2.1229646266142614,
+      "grad_norm": 0.2772404901897913,
+      "learning_rate": 5.3845807049158815e-05,
+      "loss": 0.3568,
+      "step": 236
+    },
+    {
+      "epoch": 2.1319483436271756,
+      "grad_norm": 0.18005249012493324,
+      "learning_rate": 5.2836682973249665e-05,
+      "loss": 0.3655,
+      "step": 237
+    },
+    {
+      "epoch": 2.1409320606400897,
+      "grad_norm": 0.25329620107734796,
+      "learning_rate": 5.1834215811683654e-05,
+      "loss": 0.3549,
+      "step": 238
+    },
+    {
+      "epoch": 2.149915777653004,
+      "grad_norm": 0.16701885107466424,
+      "learning_rate": 5.0838516232834614e-05,
+      "loss": 0.3566,
+      "step": 239
+    },
+    {
+      "epoch": 2.158899494665918,
+      "grad_norm": 0.2641235688031744,
+      "learning_rate": 4.9849694157962234e-05,
+      "loss": 0.3588,
+      "step": 240
+    },
+    {
+      "epoch": 2.167883211678832,
+      "grad_norm": 0.15619382389891984,
+      "learning_rate": 4.886785874907724e-05,
+      "loss": 0.3488,
+      "step": 241
+    },
+    {
+      "epoch": 2.176866928691746,
+      "grad_norm": 0.2501699730380541,
+      "learning_rate": 4.7893118396890284e-05,
+      "loss": 0.3715,
+      "step": 242
+    },
+    {
+      "epoch": 2.1858506457046603,
+      "grad_norm": 0.16211610093557965,
+      "learning_rate": 4.6925580708846104e-05,
+      "loss": 0.3765,
+      "step": 243
+    },
+    {
+      "epoch": 2.1948343627175744,
+      "grad_norm": 0.21430975519576806,
+      "learning_rate": 4.596535249724404e-05,
+      "loss": 0.3519,
+      "step": 244
+    },
+    {
+      "epoch": 2.2038180797304885,
+      "grad_norm": 0.1482343492785817,
+      "learning_rate": 4.501253976744641e-05,
+      "loss": 0.3561,
+      "step": 245
+    },
+    {
+      "epoch": 2.2128017967434026,
+      "grad_norm": 0.18600390573451364,
+      "learning_rate": 4.406724770617595e-05,
+      "loss": 0.3515,
+      "step": 246
+    },
+    {
+      "epoch": 2.2217855137563167,
+      "grad_norm": 0.13270293221384985,
+      "learning_rate": 4.3129580669903586e-05,
+      "loss": 0.3612,
+      "step": 247
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.15896036552877565,
+      "learning_rate": 4.2199642173327955e-05,
+      "loss": 0.3482,
+      "step": 248
+    },
+    {
+      "epoch": 2.239752947782145,
+      "grad_norm": 0.13144953626310965,
+      "learning_rate": 4.127753487794768e-05,
+      "loss": 0.3559,
+      "step": 249
+    },
+    {
+      "epoch": 2.248736664795059,
+      "grad_norm": 0.14893603771330527,
+      "learning_rate": 4.036336058072806e-05,
+      "loss": 0.356,
+      "step": 250
+    },
+    {
+      "epoch": 2.257720381807973,
+      "grad_norm": 0.13694686666323366,
+      "learning_rate": 3.9457220202863104e-05,
+      "loss": 0.3606,
+      "step": 251
+    },
+    {
+      "epoch": 2.2667040988208873,
+      "grad_norm": 0.14835982621693597,
+      "learning_rate": 3.855921377863414e-05,
+      "loss": 0.3667,
+      "step": 252
+    },
+    {
+      "epoch": 2.2756878158338014,
+      "grad_norm": 0.13903851563190525,
+      "learning_rate": 3.766944044436649e-05,
+      "loss": 0.3548,
+      "step": 253
+    },
+    {
+      "epoch": 2.2846715328467155,
+      "grad_norm": 0.13419589581861444,
+      "learning_rate": 3.678799842748521e-05,
+      "loss": 0.3716,
+      "step": 254
+    },
+    {
+      "epoch": 2.293655249859629,
+      "grad_norm": 0.14685606142797603,
+      "learning_rate": 3.5914985035671156e-05,
+      "loss": 0.3571,
+      "step": 255
+    },
+    {
+      "epoch": 2.3026389668725433,
+      "grad_norm": 0.1304570595276937,
+      "learning_rate": 3.5050496646118584e-05,
+      "loss": 0.3655,
+      "step": 256
+    },
+    {
+      "epoch": 2.3116226838854574,
+      "grad_norm": 0.12826171716327225,
+      "learning_rate": 3.4194628694895594e-05,
+      "loss": 0.364,
+      "step": 257
+    },
+    {
+      "epoch": 2.3206064008983716,
+      "grad_norm": 0.11075309460103316,
+      "learning_rate": 3.334747566640824e-05,
+      "loss": 0.351,
+      "step": 258
+    },
+    {
+      "epoch": 2.3295901179112857,
+      "grad_norm": 0.1312845965125087,
+      "learning_rate": 3.2509131082969915e-05,
+      "loss": 0.3687,
+      "step": 259
+    },
+    {
+      "epoch": 2.3385738349242,
+      "grad_norm": 0.11983646002909432,
+      "learning_rate": 3.167968749447683e-05,
+      "loss": 0.3454,
+      "step": 260
+    },
+    {
+      "epoch": 2.347557551937114,
+      "grad_norm": 0.11479008167106183,
+      "learning_rate": 3.0859236468190844e-05,
+      "loss": 0.3654,
+      "step": 261
+    },
+    {
+      "epoch": 2.356541268950028,
+      "grad_norm": 0.11632197031117864,
+      "learning_rate": 3.00478685786309e-05,
+      "loss": 0.3617,
+      "step": 262
+    },
+    {
+      "epoch": 2.365524985962942,
+      "grad_norm": 0.12594569973876463,
+      "learning_rate": 2.9245673397573843e-05,
+      "loss": 0.3582,
+      "step": 263
+    },
+    {
+      "epoch": 2.3745087029758563,
+      "grad_norm": 0.124698262560256,
+      "learning_rate": 2.8452739484166123e-05,
+      "loss": 0.3628,
+      "step": 264
+    },
+    {
+      "epoch": 2.3834924199887704,
+      "grad_norm": 0.11918130322626423,
+      "learning_rate": 2.7669154375147227e-05,
+      "loss": 0.3617,
+      "step": 265
+    },
+    {
+      "epoch": 2.3924761370016845,
+      "grad_norm": 0.11920547285141371,
+      "learning_rate": 2.6895004575185922e-05,
+      "loss": 0.3657,
+      "step": 266
+    },
+    {
+      "epoch": 2.4014598540145986,
+      "grad_norm": 0.10937163911147071,
+      "learning_rate": 2.6130375547330496e-05,
+      "loss": 0.3519,
+      "step": 267
+    },
+    {
+      "epoch": 2.4104435710275127,
+      "grad_norm": 0.14631208309450952,
+      "learning_rate": 2.5375351703574044e-05,
+      "loss": 0.3631,
+      "step": 268
+    },
+    {
+      "epoch": 2.419427288040427,
+      "grad_norm": 0.1265003307705914,
+      "learning_rate": 2.4630016395535618e-05,
+      "loss": 0.3679,
+      "step": 269
+    },
+    {
+      "epoch": 2.428411005053341,
+      "grad_norm": 0.11607388860612512,
+      "learning_rate": 2.389445190525866e-05,
+      "loss": 0.3475,
+      "step": 270
+    },
+    {
+      "epoch": 2.437394722066255,
+      "grad_norm": 0.13926049695780746,
+      "learning_rate": 2.316873943612722e-05,
+      "loss": 0.3509,
+      "step": 271
+    },
+    {
+      "epoch": 2.446378439079169,
+      "grad_norm": 0.12235594276612768,
+      "learning_rate": 2.2452959103901613e-05,
+      "loss": 0.3597,
+      "step": 272
+    },
+    {
+      "epoch": 2.4553621560920833,
+      "grad_norm": 0.11752333623641517,
+      "learning_rate": 2.1747189927873877e-05,
+      "loss": 0.3515,
+      "step": 273
+    },
+    {
+      "epoch": 2.464345873104997,
+      "grad_norm": 0.11683542380499232,
+      "learning_rate": 2.1051509822144332e-05,
+      "loss": 0.3695,
+      "step": 274
+    },
+    {
+      "epoch": 2.473329590117911,
+      "grad_norm": 0.12084518265976886,
+      "learning_rate": 2.0365995587020234e-05,
+      "loss": 0.3592,
+      "step": 275
+    },
+    {
+      "epoch": 2.482313307130825,
+      "grad_norm": 0.10536387180642533,
+      "learning_rate": 1.969072290053725e-05,
+      "loss": 0.3634,
+      "step": 276
+    },
+    {
+      "epoch": 2.4912970241437393,
+      "grad_norm": 0.12459678439311449,
+      "learning_rate": 1.902576631010499e-05,
+      "loss": 0.3444,
+      "step": 277
+    },
+    {
+      "epoch": 2.5002807411566534,
+      "grad_norm": 0.10423725455797198,
+      "learning_rate": 1.8371199224277213e-05,
+      "loss": 0.3446,
+      "step": 278
+    },
+    {
+      "epoch": 2.5092644581695676,
+      "grad_norm": 0.09677902356707294,
+      "learning_rate": 1.772709390464784e-05,
+      "loss": 0.3565,
+      "step": 279
+    },
+    {
+      "epoch": 2.5182481751824817,
+      "grad_norm": 0.1056308438167975,
+      "learning_rate": 1.7093521457873555e-05,
+      "loss": 0.3667,
+      "step": 280
+    },
+    {
+      "epoch": 2.527231892195396,
+      "grad_norm": 0.09014603021736677,
+      "learning_rate": 1.647055182782392e-05,
+      "loss": 0.3486,
+      "step": 281
+    },
+    {
+      "epoch": 2.53621560920831,
+      "grad_norm": 0.10287065883652714,
+      "learning_rate": 1.5858253787859857e-05,
+      "loss": 0.3525,
+      "step": 282
+    },
+    {
+      "epoch": 2.545199326221224,
+      "grad_norm": 0.09914950751521003,
+      "learning_rate": 1.5256694933241261e-05,
+      "loss": 0.3675,
+      "step": 283
+    },
+    {
+      "epoch": 2.554183043234138,
+      "grad_norm": 0.09453069458816908,
+      "learning_rate": 1.466594167366493e-05,
+      "loss": 0.3443,
+      "step": 284
+    },
+    {
+      "epoch": 2.5631667602470523,
+      "grad_norm": 0.10390944505822122,
+      "learning_rate": 1.4086059225933016e-05,
+      "loss": 0.3719,
+      "step": 285
+    },
+    {
+      "epoch": 2.5721504772599664,
+      "grad_norm": 0.10438780375230042,
+      "learning_rate": 1.3517111606753471e-05,
+      "loss": 0.3605,
+      "step": 286
+    },
+    {
+      "epoch": 2.5811341942728805,
+      "grad_norm": 0.09859350856353293,
+      "learning_rate": 1.2959161625672802e-05,
+      "loss": 0.3616,
+      "step": 287
+    },
+    {
+      "epoch": 2.5901179112857946,
+      "grad_norm": 0.09622498419041688,
+      "learning_rate": 1.2412270878142156e-05,
+      "loss": 0.3501,
+      "step": 288
+    },
+    {
+      "epoch": 2.5991016282987087,
+      "grad_norm": 0.09113584600196091,
+      "learning_rate": 1.1876499738717436e-05,
+      "loss": 0.3503,
+      "step": 289
+    },
+    {
+      "epoch": 2.608085345311623,
+      "grad_norm": 0.09741692104467439,
+      "learning_rate": 1.1351907354394194e-05,
+      "loss": 0.3657,
+      "step": 290
+    },
+    {
+      "epoch": 2.6170690623245365,
+      "grad_norm": 0.09206757777787074,
+      "learning_rate": 1.0838551638078013e-05,
+      "loss": 0.3603,
+      "step": 291
+    },
+    {
+      "epoch": 2.626052779337451,
+      "grad_norm": 0.08188402631450145,
+      "learning_rate": 1.0336489262191212e-05,
+      "loss": 0.3462,
+      "step": 292
+    },
+    {
+      "epoch": 2.6350364963503647,
+      "grad_norm": 0.08808466701218914,
+      "learning_rate": 9.845775652416357e-06,
+      "loss": 0.3632,
+      "step": 293
+    },
+    {
+      "epoch": 2.6440202133632793,
+      "grad_norm": 0.0943618643568796,
+      "learning_rate": 9.366464981577584e-06,
+      "loss": 0.3594,
+      "step": 294
+    },
+    {
+      "epoch": 2.653003930376193,
+      "grad_norm": 0.09593739977292585,
+      "learning_rate": 8.89861016365997e-06,
+      "loss": 0.3515,
+      "step": 295
+    },
+    {
+      "epoch": 2.661987647389107,
+      "grad_norm": 0.08248015664645339,
+      "learning_rate": 8.442262847968263e-06,
+      "loss": 0.3608,
+      "step": 296
+    },
+    {
+      "epoch": 2.670971364402021,
+      "grad_norm": 0.08411868089198224,
+      "learning_rate": 7.997473413424846e-06,
+      "loss": 0.3592,
+      "step": 297
+    },
+    {
+      "epoch": 2.6799550814149353,
+      "grad_norm": 0.09490876979791782,
+      "learning_rate": 7.564290963008149e-06,
+      "loss": 0.358,
+      "step": 298
+    },
+    {
+      "epoch": 2.6889387984278494,
+      "grad_norm": 0.08867586845979594,
+      "learning_rate": 7.142763318331872e-06,
+      "loss": 0.351,
+      "step": 299
+    },
+    {
+      "epoch": 2.6979225154407636,
+      "grad_norm": 0.08779362567358494,
+      "learning_rate": 6.732937014365695e-06,
+      "loss": 0.3494,
+      "step": 300
+    },
+    {
+      "epoch": 2.7069062324536777,
+      "grad_norm": 0.08301900637144422,
+      "learning_rate": 6.3348572942979654e-06,
+      "loss": 0.3587,
+      "step": 301
+    },
+    {
+      "epoch": 2.715889949466592,
+      "grad_norm": 0.08891924464679533,
+      "learning_rate": 5.948568104541074e-06,
+      "loss": 0.3651,
+      "step": 302
+    },
+    {
+      "epoch": 2.724873666479506,
+      "grad_norm": 0.08730964097529133,
+      "learning_rate": 5.574112089879872e-06,
+      "loss": 0.3523,
+      "step": 303
+    },
+    {
+      "epoch": 2.73385738349242,
+      "grad_norm": 0.08450880465007451,
+      "learning_rate": 5.211530588763962e-06,
+      "loss": 0.3482,
+      "step": 304
+    },
+    {
+      "epoch": 2.742841100505334,
+      "grad_norm": 0.08342255207562929,
+      "learning_rate": 4.860863628744007e-06,
+      "loss": 0.3564,
+      "step": 305
+    },
+    {
+      "epoch": 2.7518248175182483,
+      "grad_norm": 0.08975850015832335,
+      "learning_rate": 4.522149922052897e-06,
+      "loss": 0.3491,
+      "step": 306
+    },
+    {
+      "epoch": 2.7608085345311624,
+      "grad_norm": 0.08240771555091773,
+      "learning_rate": 4.195426861332049e-06,
+      "loss": 0.3588,
+      "step": 307
+    },
+    {
+      "epoch": 2.7697922515440765,
+      "grad_norm": 0.08798778487887464,
+      "learning_rate": 3.880730515503412e-06,
+      "loss": 0.3427,
+      "step": 308
+    },
+    {
+      "epoch": 2.7787759685569906,
+      "grad_norm": 0.08113053094229192,
+      "learning_rate": 3.57809562578763e-06,
+      "loss": 0.3728,
+      "step": 309
+    },
+    {
+      "epoch": 2.7877596855699043,
+      "grad_norm": 0.08279617250380948,
+      "learning_rate": 3.2875556018687533e-06,
+      "loss": 0.3418,
+      "step": 310
+    },
+    {
+      "epoch": 2.796743402582819,
+      "grad_norm": 0.07729029098691793,
+      "learning_rate": 3.0091425182058514e-06,
+      "loss": 0.3617,
+      "step": 311
+    },
+    {
+      "epoch": 2.8057271195957325,
+      "grad_norm": 0.08856029522706829,
+      "learning_rate": 2.742887110492231e-06,
+      "loss": 0.3602,
+      "step": 312
+    },
+    {
+      "epoch": 2.814710836608647,
+      "grad_norm": 0.09170437352580232,
+      "learning_rate": 2.4888187722622945e-06,
+      "loss": 0.3532,
+      "step": 313
+    },
+    {
+      "epoch": 2.8236945536215607,
+      "grad_norm": 0.08713999383289339,
+      "learning_rate": 2.2469655516466e-06,
+      "loss": 0.3579,
+      "step": 314
+    },
+    {
+      "epoch": 2.832678270634475,
+      "grad_norm": 0.08158645722200215,
+      "learning_rate": 2.017354148275491e-06,
+      "loss": 0.3527,
+      "step": 315
+    },
+    {
+      "epoch": 2.841661987647389,
+      "grad_norm": 0.0775962718555909,
+      "learning_rate": 1.8000099103314957e-06,
+      "loss": 0.357,
+      "step": 316
+    },
+    {
+      "epoch": 2.850645704660303,
+      "grad_norm": 0.07417615515514962,
+      "learning_rate": 1.5949568317510827e-06,
+      "loss": 0.3633,
+      "step": 317
+    },
+    {
+      "epoch": 2.859629421673217,
+      "grad_norm": 0.07076814193057895,
+      "learning_rate": 1.402217549575769e-06,
+      "loss": 0.357,
+      "step": 318
+    },
+    {
+      "epoch": 2.8686131386861313,
+      "grad_norm": 0.08051085296339765,
+      "learning_rate": 1.2218133414530984e-06,
+      "loss": 0.3528,
+      "step": 319
+    },
+    {
+      "epoch": 2.8775968556990454,
+      "grad_norm": 0.07853549988975052,
+      "learning_rate": 1.0537641232876473e-06,
+      "loss": 0.3614,
+      "step": 320
+    },
+    {
+      "epoch": 2.8865805727119596,
+      "grad_norm": 0.07073622608834755,
+      "learning_rate": 8.980884470424321e-07,
+      "loss": 0.3551,
+      "step": 321
+    },
+    {
+      "epoch": 2.8955642897248737,
+      "grad_norm": 0.07143610759857762,
+      "learning_rate": 7.548034986908066e-07,
+      "loss": 0.3545,
+      "step": 322
+    },
+    {
+      "epoch": 2.904548006737788,
+      "grad_norm": 0.0749903587104429,
+      "learning_rate": 6.239250963192269e-07,
+      "loss": 0.3583,
+      "step": 323
+    },
+    {
+      "epoch": 2.913531723750702,
+      "grad_norm": 0.07746634488254509,
+      "learning_rate": 5.054676883809827e-07,
+      "loss": 0.3533,
+      "step": 324
+    },
+    {
+      "epoch": 2.922515440763616,
+      "grad_norm": 0.07444763137173055,
+      "learning_rate": 3.994443521011485e-07,
+      "loss": 0.3627,
+      "step": 325
+    },
+    {
+      "epoch": 2.93149915777653,
+      "grad_norm": 0.06987734521185171,
+      "learning_rate": 3.058667920329281e-07,
+      "loss": 0.348,
+      "step": 326
+    },
+    {
+      "epoch": 2.9404828747894443,
+      "grad_norm": 0.07171007945104789,
+      "learning_rate": 2.2474533876546995e-07,
+      "loss": 0.3455,
+      "step": 327
+    },
+    {
+      "epoch": 2.9494665918023584,
+      "grad_norm": 0.0727402518849161,
+      "learning_rate": 1.560889477834654e-07,
+      "loss": 0.3579,
+      "step": 328
+    },
+    {
+      "epoch": 2.958450308815272,
+      "grad_norm": 0.07322322016212719,
+      "learning_rate": 9.99051984784689e-08,
+      "loss": 0.3528,
+      "step": 329
+    },
+    {
+      "epoch": 2.9674340258281866,
+      "grad_norm": 0.07567114973675214,
+      "learning_rate": 5.620029331218986e-08,
+      "loss": 0.3564,
+      "step": 330
+    },
+    {
+      "epoch": 2.9764177428411003,
+      "grad_norm": 0.07625451494420618,
+      "learning_rate": 2.4979057131732006e-08,
+      "loss": 0.3594,
+      "step": 331
+    },
+    {
+      "epoch": 2.985401459854015,
+      "grad_norm": 0.07133050968488629,
+      "learning_rate": 6.2449366369555176e-09,
+      "loss": 0.3521,
+      "step": 332
+    },
+    {
+      "epoch": 2.9943851768669285,
+      "grad_norm": 0.07662216533129715,
+      "learning_rate": 0.0,
+      "loss": 0.3558,
+      "step": 333
+    },
+    {
+      "epoch": 2.9943851768669285,
+      "step": 333,
+      "total_flos": 4607052149424128.0,
+      "train_loss": 0.5244213240640657,
+      "train_runtime": 115493.2352,
+      "train_samples_per_second": 2.96,
+      "train_steps_per_second": 0.003
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 333,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4607052149424128.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}