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_laradjusted8
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # global_batchsize_1024_laradjusted8
 
-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": 0.44577710639249096,
+    "train_runtime": 115174.1649,
+    "train_samples_per_second": 2.968,
+    "train_steps_per_second": 0.003
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9943851768669285,
+    "total_flos": 4607052149424128.0,
+    "train_loss": 0.44577710639249096,
+    "train_runtime": 115174.1649,
+    "train_samples_per_second": 2.968,
+    "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": 5.839039155729955,
+      "learning_rate": 3.327561176470588e-06,
+      "loss": 0.851,
+      "step": 1
+    },
+    {
+      "epoch": 0.017967434025828188,
+      "grad_norm": 5.796538258918326,
+      "learning_rate": 6.655122352941176e-06,
+      "loss": 0.8589,
+      "step": 2
+    },
+    {
+      "epoch": 0.02695115103874228,
+      "grad_norm": 5.348268151193076,
+      "learning_rate": 9.982683529411766e-06,
+      "loss": 0.8477,
+      "step": 3
+    },
+    {
+      "epoch": 0.035934868051656375,
+      "grad_norm": 2.2398177096044813,
+      "learning_rate": 1.3310244705882352e-05,
+      "loss": 0.7674,
+      "step": 4
+    },
+    {
+      "epoch": 0.044918585064570464,
+      "grad_norm": 4.4966504833626315,
+      "learning_rate": 1.6637805882352943e-05,
+      "loss": 0.7793,
+      "step": 5
+    },
+    {
+      "epoch": 0.05390230207748456,
+      "grad_norm": 4.916507365933338,
+      "learning_rate": 1.996536705882353e-05,
+      "loss": 0.7407,
+      "step": 6
+    },
+    {
+      "epoch": 0.06288601909039865,
+      "grad_norm": 5.555268639221758,
+      "learning_rate": 2.3292928235294116e-05,
+      "loss": 0.7462,
+      "step": 7
+    },
+    {
+      "epoch": 0.07186973610331275,
+      "grad_norm": 2.5335144476571614,
+      "learning_rate": 2.6620489411764705e-05,
+      "loss": 0.6982,
+      "step": 8
+    },
+    {
+      "epoch": 0.08085345311622684,
+      "grad_norm": 2.469631940841415,
+      "learning_rate": 2.9948050588235297e-05,
+      "loss": 0.6821,
+      "step": 9
+    },
+    {
+      "epoch": 0.08983717012914093,
+      "grad_norm": 1.6447880426101031,
+      "learning_rate": 3.3275611764705885e-05,
+      "loss": 0.6572,
+      "step": 10
+    },
+    {
+      "epoch": 0.09882088714205503,
+      "grad_norm": 1.7456324190058856,
+      "learning_rate": 3.660317294117647e-05,
+      "loss": 0.6377,
+      "step": 11
+    },
+    {
+      "epoch": 0.10780460415496912,
+      "grad_norm": 1.3930051452388852,
+      "learning_rate": 3.993073411764706e-05,
+      "loss": 0.6226,
+      "step": 12
+    },
+    {
+      "epoch": 0.11678832116788321,
+      "grad_norm": 1.557060887911309,
+      "learning_rate": 4.325829529411765e-05,
+      "loss": 0.6112,
+      "step": 13
+    },
+    {
+      "epoch": 0.1257720381807973,
+      "grad_norm": 1.5904635848649458,
+      "learning_rate": 4.658585647058823e-05,
+      "loss": 0.6087,
+      "step": 14
+    },
+    {
+      "epoch": 0.13475575519371139,
+      "grad_norm": 1.8726010376862894,
+      "learning_rate": 4.9913417647058825e-05,
+      "loss": 0.6077,
+      "step": 15
+    },
+    {
+      "epoch": 0.1437394722066255,
+      "grad_norm": 1.5092772957558627,
+      "learning_rate": 5.324097882352941e-05,
+      "loss": 0.5984,
+      "step": 16
+    },
+    {
+      "epoch": 0.1527231892195396,
+      "grad_norm": 1.5672151024489294,
+      "learning_rate": 5.656854e-05,
+      "loss": 0.5822,
+      "step": 17
+    },
+    {
+      "epoch": 0.16170690623245368,
+      "grad_norm": 1.4358756651793523,
+      "learning_rate": 5.9896101176470594e-05,
+      "loss": 0.5763,
+      "step": 18
+    },
+    {
+      "epoch": 0.17069062324536777,
+      "grad_norm": 1.6595556675185048,
+      "learning_rate": 6.322366235294119e-05,
+      "loss": 0.5756,
+      "step": 19
+    },
+    {
+      "epoch": 0.17967434025828186,
+      "grad_norm": 0.9833836084606898,
+      "learning_rate": 6.655122352941177e-05,
+      "loss": 0.5734,
+      "step": 20
+    },
+    {
+      "epoch": 0.18865805727119594,
+      "grad_norm": 1.7952700352939273,
+      "learning_rate": 6.987878470588236e-05,
+      "loss": 0.5738,
+      "step": 21
+    },
+    {
+      "epoch": 0.19764177428411006,
+      "grad_norm": 1.6999155441996412,
+      "learning_rate": 7.320634588235294e-05,
+      "loss": 0.5751,
+      "step": 22
+    },
+    {
+      "epoch": 0.20662549129702415,
+      "grad_norm": 2.752809363195007,
+      "learning_rate": 7.653390705882354e-05,
+      "loss": 0.5699,
+      "step": 23
+    },
+    {
+      "epoch": 0.21560920830993824,
+      "grad_norm": 1.8184660667298642,
+      "learning_rate": 7.986146823529412e-05,
+      "loss": 0.5675,
+      "step": 24
+    },
+    {
+      "epoch": 0.22459292532285233,
+      "grad_norm": 3.1030289897828607,
+      "learning_rate": 8.318902941176471e-05,
+      "loss": 0.5782,
+      "step": 25
+    },
+    {
+      "epoch": 0.23357664233576642,
+      "grad_norm": 2.185491499693549,
+      "learning_rate": 8.65165905882353e-05,
+      "loss": 0.5753,
+      "step": 26
+    },
+    {
+      "epoch": 0.2425603593486805,
+      "grad_norm": 2.17356984710728,
+      "learning_rate": 8.984415176470588e-05,
+      "loss": 0.5693,
+      "step": 27
+    },
+    {
+      "epoch": 0.2515440763615946,
+      "grad_norm": 1.5059122763121213,
+      "learning_rate": 9.317171294117647e-05,
+      "loss": 0.5519,
+      "step": 28
+    },
+    {
+      "epoch": 0.2605277933745087,
+      "grad_norm": 1.8271847820321647,
+      "learning_rate": 9.649927411764706e-05,
+      "loss": 0.5563,
+      "step": 29
+    },
+    {
+      "epoch": 0.26951151038742277,
+      "grad_norm": 1.1123828217442284,
+      "learning_rate": 9.982683529411765e-05,
+      "loss": 0.5542,
+      "step": 30
+    },
+    {
+      "epoch": 0.2784952274003369,
+      "grad_norm": 1.8487378177291283,
+      "learning_rate": 0.00010315439647058823,
+      "loss": 0.5533,
+      "step": 31
+    },
+    {
+      "epoch": 0.287478944413251,
+      "grad_norm": 1.8607862520510237,
+      "learning_rate": 0.00010648195764705882,
+      "loss": 0.556,
+      "step": 32
+    },
+    {
+      "epoch": 0.29646266142616506,
+      "grad_norm": 1.6040227599875894,
+      "learning_rate": 0.00010980951882352942,
+      "loss": 0.5351,
+      "step": 33
+    },
+    {
+      "epoch": 0.3054463784390792,
+      "grad_norm": 1.404043953546152,
+      "learning_rate": 0.00011313708,
+      "loss": 0.539,
+      "step": 34
+    },
+    {
+      "epoch": 0.31443009545199324,
+      "grad_norm": 2.397356079807102,
+      "learning_rate": 0.0001131339575294736,
+      "loss": 0.5456,
+      "step": 35
+    },
+    {
+      "epoch": 0.32341381246490736,
+      "grad_norm": 1.9800116620234915,
+      "learning_rate": 0.00011312459046260269,
+      "loss": 0.5434,
+      "step": 36
+    },
+    {
+      "epoch": 0.3323975294778215,
+      "grad_norm": 1.2453642497862838,
+      "learning_rate": 0.0001131089798334741,
+      "loss": 0.5324,
+      "step": 37
+    },
+    {
+      "epoch": 0.34138124649073553,
+      "grad_norm": 2.063267541300665,
+      "learning_rate": 0.00011308712736543891,
+      "loss": 0.5411,
+      "step": 38
+    },
+    {
+      "epoch": 0.35036496350364965,
+      "grad_norm": 1.3887178955612751,
+      "learning_rate": 0.00011305903547092245,
+      "loss": 0.5326,
+      "step": 39
+    },
+    {
+      "epoch": 0.3593486805165637,
+      "grad_norm": 1.7807218923269645,
+      "learning_rate": 0.00011302470725115774,
+      "loss": 0.5349,
+      "step": 40
+    },
+    {
+      "epoch": 0.36833239752947783,
+      "grad_norm": 0.878665749031585,
+      "learning_rate": 0.00011298414649584322,
+      "loss": 0.5326,
+      "step": 41
+    },
+    {
+      "epoch": 0.3773161145423919,
+      "grad_norm": 1.5918264570623344,
+      "learning_rate": 0.00011293735768272441,
+      "loss": 0.5323,
+      "step": 42
+    },
+    {
+      "epoch": 0.386299831555306,
+      "grad_norm": 1.2803597285407087,
+      "learning_rate": 0.00011288434597709956,
+      "loss": 0.526,
+      "step": 43
+    },
+    {
+      "epoch": 0.3952835485682201,
+      "grad_norm": 1.1274299848013631,
+      "learning_rate": 0.00011282511723124938,
+      "loss": 0.5144,
+      "step": 44
+    },
+    {
+      "epoch": 0.4042672655811342,
+      "grad_norm": 1.6628496159523496,
+      "learning_rate": 0.00011275967798379107,
+      "loss": 0.5354,
+      "step": 45
+    },
+    {
+      "epoch": 0.4132509825940483,
+      "grad_norm": 1.3851908223826852,
+      "learning_rate": 0.00011268803545895634,
+      "loss": 0.5224,
+      "step": 46
+    },
+    {
+      "epoch": 0.42223469960696236,
+      "grad_norm": 1.2685767531050909,
+      "learning_rate": 0.00011261019756579401,
+      "loss": 0.5212,
+      "step": 47
+    },
+    {
+      "epoch": 0.4312184166198765,
+      "grad_norm": 1.2207866310513658,
+      "learning_rate": 0.00011252617289729688,
+      "loss": 0.5113,
+      "step": 48
+    },
+    {
+      "epoch": 0.4402021336327906,
+      "grad_norm": 0.9490873435896563,
+      "learning_rate": 0.00011243597072945297,
+      "loss": 0.5164,
+      "step": 49
+    },
+    {
+      "epoch": 0.44918585064570465,
+      "grad_norm": 1.941201687588716,
+      "learning_rate": 0.00011233960102022164,
+      "loss": 0.509,
+      "step": 50
+    },
+    {
+      "epoch": 0.45816956765861877,
+      "grad_norm": 1.0083083862917877,
+      "learning_rate": 0.00011223707440843418,
+      "loss": 0.5187,
+      "step": 51
+    },
+    {
+      "epoch": 0.46715328467153283,
+      "grad_norm": 1.730383616067843,
+      "learning_rate": 0.00011212840221261932,
+      "loss": 0.5239,
+      "step": 52
+    },
+    {
+      "epoch": 0.47613700168444695,
+      "grad_norm": 1.1626453188818293,
+      "learning_rate": 0.00011201359642975381,
+      "loss": 0.5109,
+      "step": 53
+    },
+    {
+      "epoch": 0.485120718697361,
+      "grad_norm": 1.2420154197254125,
+      "learning_rate": 0.00011189266973393787,
+      "loss": 0.5202,
+      "step": 54
+    },
+    {
+      "epoch": 0.4941044357102751,
+      "grad_norm": 1.0906107384257,
+      "learning_rate": 0.00011176563547499611,
+      "loss": 0.5194,
+      "step": 55
+    },
+    {
+      "epoch": 0.5030881527231892,
+      "grad_norm": 1.1363860383697608,
+      "learning_rate": 0.00011163250767700374,
+      "loss": 0.5084,
+      "step": 56
+    },
+    {
+      "epoch": 0.5120718697361033,
+      "grad_norm": 0.8452490494880119,
+      "learning_rate": 0.00011149330103673833,
+      "loss": 0.5163,
+      "step": 57
+    },
+    {
+      "epoch": 0.5210555867490174,
+      "grad_norm": 0.8074335130399725,
+      "learning_rate": 0.00011134803092205739,
+      "loss": 0.509,
+      "step": 58
+    },
+    {
+      "epoch": 0.5300393037619315,
+      "grad_norm": 1.0175175039841469,
+      "learning_rate": 0.00011119671337020184,
+      "loss": 0.5128,
+      "step": 59
+    },
+    {
+      "epoch": 0.5390230207748455,
+      "grad_norm": 1.4485688423144,
+      "learning_rate": 0.00011103936508602549,
+      "loss": 0.5146,
+      "step": 60
+    },
+    {
+      "epoch": 0.5480067377877597,
+      "grad_norm": 1.0531638689798895,
+      "learning_rate": 0.00011087600344015093,
+      "loss": 0.5058,
+      "step": 61
+    },
+    {
+      "epoch": 0.5569904548006738,
+      "grad_norm": 0.8439958643801734,
+      "learning_rate": 0.0001107066464670519,
+      "loss": 0.4989,
+      "step": 62
+    },
+    {
+      "epoch": 0.5659741718135879,
+      "grad_norm": 1.1158174292200818,
+      "learning_rate": 0.00011053131286306234,
+      "loss": 0.5104,
+      "step": 63
+    },
+    {
+      "epoch": 0.574957888826502,
+      "grad_norm": 1.34506015438112,
+      "learning_rate": 0.00011035002198431236,
+      "loss": 0.4952,
+      "step": 64
+    },
+    {
+      "epoch": 0.583941605839416,
+      "grad_norm": 0.7925867568991634,
+      "learning_rate": 0.00011016279384459147,
+      "loss": 0.4991,
+      "step": 65
+    },
+    {
+      "epoch": 0.5929253228523301,
+      "grad_norm": 1.1528151876727255,
+      "learning_rate": 0.00010996964911313906,
+      "loss": 0.4997,
+      "step": 66
+    },
+    {
+      "epoch": 0.6019090398652442,
+      "grad_norm": 0.9670195638674501,
+      "learning_rate": 0.00010977060911236265,
+      "loss": 0.4937,
+      "step": 67
+    },
+    {
+      "epoch": 0.6108927568781584,
+      "grad_norm": 1.0041541158509282,
+      "learning_rate": 0.00010956569581548399,
+      "loss": 0.5039,
+      "step": 68
+    },
+    {
+      "epoch": 0.6198764738910725,
+      "grad_norm": 1.1699145795561787,
+      "learning_rate": 0.00010935493184411321,
+      "loss": 0.4932,
+      "step": 69
+    },
+    {
+      "epoch": 0.6288601909039865,
+      "grad_norm": 1.5639307567031937,
+      "learning_rate": 0.00010913834046575162,
+      "loss": 0.4954,
+      "step": 70
+    },
+    {
+      "epoch": 0.6378439079169006,
+      "grad_norm": 0.53442675066816,
+      "learning_rate": 0.00010891594559122298,
+      "loss": 0.4916,
+      "step": 71
+    },
+    {
+      "epoch": 0.6468276249298147,
+      "grad_norm": 1.4894433981804676,
+      "learning_rate": 0.00010868777177203387,
+      "loss": 0.4877,
+      "step": 72
+    },
+    {
+      "epoch": 0.6558113419427288,
+      "grad_norm": 0.8751715637414447,
+      "learning_rate": 0.00010845384419766329,
+      "loss": 0.4959,
+      "step": 73
+    },
+    {
+      "epoch": 0.664795058955643,
+      "grad_norm": 1.0266195653044043,
+      "learning_rate": 0.00010821418869278185,
+      "loss": 0.4925,
+      "step": 74
+    },
+    {
+      "epoch": 0.673778775968557,
+      "grad_norm": 0.9407955615022422,
+      "learning_rate": 0.00010796883171440082,
+      "loss": 0.4901,
+      "step": 75
+    },
+    {
+      "epoch": 0.6827624929814711,
+      "grad_norm": 0.9077354205891603,
+      "learning_rate": 0.00010771780034895143,
+      "loss": 0.5053,
+      "step": 76
+    },
+    {
+      "epoch": 0.6917462099943852,
+      "grad_norm": 1.0564576660765181,
+      "learning_rate": 0.00010746112230929458,
+      "loss": 0.4896,
+      "step": 77
+    },
+    {
+      "epoch": 0.7007299270072993,
+      "grad_norm": 0.9140880531387264,
+      "learning_rate": 0.00010719882593166137,
+      "loss": 0.4913,
+      "step": 78
+    },
+    {
+      "epoch": 0.7097136440202133,
+      "grad_norm": 1.0254444567251542,
+      "learning_rate": 0.00010693094017252518,
+      "loss": 0.4809,
+      "step": 79
+    },
+    {
+      "epoch": 0.7186973610331274,
+      "grad_norm": 0.9690806724249464,
+      "learning_rate": 0.00010665749460540463,
+      "loss": 0.4965,
+      "step": 80
+    },
+    {
+      "epoch": 0.7276810780460415,
+      "grad_norm": 0.9212842497405144,
+      "learning_rate": 0.00010637851941759905,
+      "loss": 0.4868,
+      "step": 81
+    },
+    {
+      "epoch": 0.7366647950589557,
+      "grad_norm": 0.9385243969411573,
+      "learning_rate": 0.00010609404540685576,
+      "loss": 0.4804,
+      "step": 82
+    },
+    {
+      "epoch": 0.7456485120718698,
+      "grad_norm": 0.9315727772670367,
+      "learning_rate": 0.00010580410397797024,
+      "loss": 0.4993,
+      "step": 83
+    },
+    {
+      "epoch": 0.7546322290847838,
+      "grad_norm": 0.8277882575497786,
+      "learning_rate": 0.0001055087271393191,
+      "loss": 0.4848,
+      "step": 84
+    },
+    {
+      "epoch": 0.7636159460976979,
+      "grad_norm": 0.5582581268949416,
+      "learning_rate": 0.00010520794749932649,
+      "loss": 0.4835,
+      "step": 85
+    },
+    {
+      "epoch": 0.772599663110612,
+      "grad_norm": 0.9860386326087675,
+      "learning_rate": 0.00010490179826286422,
+      "loss": 0.4858,
+      "step": 86
+    },
+    {
+      "epoch": 0.7815833801235261,
+      "grad_norm": 0.8975461926762575,
+      "learning_rate": 0.00010459031322758622,
+      "loss": 0.477,
+      "step": 87
+    },
+    {
+      "epoch": 0.7905670971364402,
+      "grad_norm": 0.5950312004330175,
+      "learning_rate": 0.00010427352678019722,
+      "loss": 0.4857,
+      "step": 88
+    },
+    {
+      "epoch": 0.7995508141493542,
+      "grad_norm": 0.8621308039815323,
+      "learning_rate": 0.00010395147389265675,
+      "loss": 0.4704,
+      "step": 89
+    },
+    {
+      "epoch": 0.8085345311622684,
+      "grad_norm": 0.6061879218137259,
+      "learning_rate": 0.00010362419011831827,
+      "loss": 0.4795,
+      "step": 90
+    },
+    {
+      "epoch": 0.8175182481751825,
+      "grad_norm": 0.6900535166809252,
+      "learning_rate": 0.0001032917115880043,
+      "loss": 0.4802,
+      "step": 91
+    },
+    {
+      "epoch": 0.8265019651880966,
+      "grad_norm": 0.5872577816133826,
+      "learning_rate": 0.00010295407500601769,
+      "loss": 0.473,
+      "step": 92
+    },
+    {
+      "epoch": 0.8354856822010107,
+      "grad_norm": 0.515096046234384,
+      "learning_rate": 0.00010261131764608956,
+      "loss": 0.4711,
+      "step": 93
+    },
+    {
+      "epoch": 0.8444693992139247,
+      "grad_norm": 0.6579448677863009,
+      "learning_rate": 0.00010226347734726456,
+      "loss": 0.471,
+      "step": 94
+    },
+    {
+      "epoch": 0.8534531162268388,
+      "grad_norm": 0.5555092732467758,
+      "learning_rate": 0.00010191059250972342,
+      "loss": 0.4742,
+      "step": 95
+    },
+    {
+      "epoch": 0.862436833239753,
+      "grad_norm": 0.7623755773246812,
+      "learning_rate": 0.00010155270209054387,
+      "loss": 0.4798,
+      "step": 96
+    },
+    {
+      "epoch": 0.8714205502526671,
+      "grad_norm": 1.052115563519428,
+      "learning_rate": 0.00010118984559939985,
+      "loss": 0.4797,
+      "step": 97
+    },
+    {
+      "epoch": 0.8804042672655812,
+      "grad_norm": 1.2698101123883572,
+      "learning_rate": 0.00010082206309419983,
+      "loss": 0.4623,
+      "step": 98
+    },
+    {
+      "epoch": 0.8893879842784952,
+      "grad_norm": 0.6989330224303171,
+      "learning_rate": 0.00010044939517666461,
+      "loss": 0.4752,
+      "step": 99
+    },
+    {
+      "epoch": 0.8983717012914093,
+      "grad_norm": 0.47484834468390363,
+      "learning_rate": 0.00010007188298784493,
+      "loss": 0.461,
+      "step": 100
+    },
+    {
+      "epoch": 0.9073554183043234,
+      "grad_norm": 0.5244362501719997,
+      "learning_rate": 9.968956820357984e-05,
+      "loss": 0.4743,
+      "step": 101
+    },
+    {
+      "epoch": 0.9163391353172375,
+      "grad_norm": 0.8974297175269308,
+      "learning_rate": 9.93024930298957e-05,
+      "loss": 0.4623,
+      "step": 102
+    },
+    {
+      "epoch": 0.9253228523301515,
+      "grad_norm": 1.1951854808745772,
+      "learning_rate": 9.891070019834687e-05,
+      "loss": 0.4766,
+      "step": 103
+    },
+    {
+      "epoch": 0.9343065693430657,
+      "grad_norm": 0.6837683720949369,
+      "learning_rate": 9.851423296129832e-05,
+      "loss": 0.4636,
+      "step": 104
+    },
+    {
+      "epoch": 0.9432902863559798,
+      "grad_norm": 0.6430304956873865,
+      "learning_rate": 9.811313508715075e-05,
+      "loss": 0.4742,
+      "step": 105
+    },
+    {
+      "epoch": 0.9522740033688939,
+      "grad_norm": 0.6793832364041675,
+      "learning_rate": 9.77074508555087e-05,
+      "loss": 0.4621,
+      "step": 106
+    },
+    {
+      "epoch": 0.961257720381808,
+      "grad_norm": 0.9539121918298344,
+      "learning_rate": 9.729722505229225e-05,
+      "loss": 0.4664,
+      "step": 107
+    },
+    {
+      "epoch": 0.970241437394722,
+      "grad_norm": 1.3186009652346484,
+      "learning_rate": 9.68825029647929e-05,
+      "loss": 0.4606,
+      "step": 108
+    },
+    {
+      "epoch": 0.9792251544076361,
+      "grad_norm": 0.6175258191923642,
+      "learning_rate": 9.646333037667394e-05,
+      "loss": 0.4626,
+      "step": 109
+    },
+    {
+      "epoch": 0.9882088714205502,
+      "grad_norm": 0.6726481106097943,
+      "learning_rate": 9.603975356291622e-05,
+      "loss": 0.4721,
+      "step": 110
+    },
+    {
+      "epoch": 0.9971925884334644,
+      "grad_norm": 1.1397862605061986,
+      "learning_rate": 9.561181928470947e-05,
+      "loss": 0.4678,
+      "step": 111
+    },
+    {
+      "epoch": 1.0075800112296462,
+      "grad_norm": 1.6584649278716128,
+      "learning_rate": 9.517957478429017e-05,
+      "loss": 0.8226,
+      "step": 112
+    },
+    {
+      "epoch": 1.0165637282425604,
+      "grad_norm": 0.694442738541462,
+      "learning_rate": 9.474306777972606e-05,
+      "loss": 0.4199,
+      "step": 113
+    },
+    {
+      "epoch": 1.0255474452554745,
+      "grad_norm": 0.6319509579943051,
+      "learning_rate": 9.430234645964837e-05,
+      "loss": 0.4325,
+      "step": 114
+    },
+    {
+      "epoch": 1.0345311622683886,
+      "grad_norm": 0.9373067626850686,
+      "learning_rate": 9.385745947793193e-05,
+      "loss": 0.4371,
+      "step": 115
+    },
+    {
+      "epoch": 1.0435148792813027,
+      "grad_norm": 1.0372714974698216,
+      "learning_rate": 9.3408455948324e-05,
+      "loss": 0.4413,
+      "step": 116
+    },
+    {
+      "epoch": 1.0524985962942168,
+      "grad_norm": 0.7160514032984542,
+      "learning_rate": 9.295538543902229e-05,
+      "loss": 0.4313,
+      "step": 117
+    },
+    {
+      "epoch": 1.0614823133071307,
+      "grad_norm": 0.7731576151711236,
+      "learning_rate": 9.249829796720283e-05,
+      "loss": 0.4364,
+      "step": 118
+    },
+    {
+      "epoch": 1.0704660303200448,
+      "grad_norm": 0.7817245159389434,
+      "learning_rate": 9.203724399349826e-05,
+      "loss": 0.4269,
+      "step": 119
+    },
+    {
+      "epoch": 1.079449747332959,
+      "grad_norm": 0.6667107465508227,
+      "learning_rate": 9.157227441642727e-05,
+      "loss": 0.4306,
+      "step": 120
+    },
+    {
+      "epoch": 1.088433464345873,
+      "grad_norm": 0.7602370272000606,
+      "learning_rate": 9.11034405667755e-05,
+      "loss": 0.4378,
+      "step": 121
+    },
+    {
+      "epoch": 1.0974171813587872,
+      "grad_norm": 0.6574081216079507,
+      "learning_rate": 9.063079420192878e-05,
+      "loss": 0.4266,
+      "step": 122
+    },
+    {
+      "epoch": 1.1064008983717013,
+      "grad_norm": 0.6824331126919658,
+      "learning_rate": 9.015438750015956e-05,
+      "loss": 0.4443,
+      "step": 123
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.684954693277142,
+      "learning_rate": 8.967427305486632e-05,
+      "loss": 0.4221,
+      "step": 124
+    },
+    {
+      "epoch": 1.1243683323975295,
+      "grad_norm": 0.6299336818589878,
+      "learning_rate": 8.919050386876776e-05,
+      "loss": 0.4487,
+      "step": 125
+    },
+    {
+      "epoch": 1.1333520494104437,
+      "grad_norm": 0.6761515113142978,
+      "learning_rate": 8.870313334805125e-05,
+      "loss": 0.4204,
+      "step": 126
+    },
+    {
+      "epoch": 1.1423357664233578,
+      "grad_norm": 0.577699447179662,
+      "learning_rate": 8.821221529647726e-05,
+      "loss": 0.4344,
+      "step": 127
+    },
+    {
+      "epoch": 1.1513194834362717,
+      "grad_norm": 0.46584576078965556,
+      "learning_rate": 8.771780390943941e-05,
+      "loss": 0.4248,
+      "step": 128
+    },
+    {
+      "epoch": 1.1603032004491858,
+      "grad_norm": 0.5507277615907759,
+      "learning_rate": 8.721995376798167e-05,
+      "loss": 0.4379,
+      "step": 129
+    },
+    {
+      "epoch": 1.1692869174621,
+      "grad_norm": 0.6852046510797101,
+      "learning_rate": 8.671871983277274e-05,
+      "loss": 0.429,
+      "step": 130
+    },
+    {
+      "epoch": 1.178270634475014,
+      "grad_norm": 0.7770840376090264,
+      "learning_rate": 8.621415743803866e-05,
+      "loss": 0.4311,
+      "step": 131
+    },
+    {
+      "epoch": 1.1872543514879281,
+      "grad_norm": 0.5549943849940224,
+      "learning_rate": 8.570632228545411e-05,
+      "loss": 0.4279,
+      "step": 132
+    },
+    {
+      "epoch": 1.1962380685008422,
+      "grad_norm": 0.34816812187637597,
+      "learning_rate": 8.51952704379932e-05,
+      "loss": 0.4292,
+      "step": 133
+    },
+    {
+      "epoch": 1.2052217855137564,
+      "grad_norm": 0.5569753817528215,
+      "learning_rate": 8.468105831374028e-05,
+      "loss": 0.4247,
+      "step": 134
+    },
+    {
+      "epoch": 1.2142055025266705,
+      "grad_norm": 0.626503583825079,
+      "learning_rate": 8.416374267966168e-05,
+      "loss": 0.426,
+      "step": 135
+    },
+    {
+      "epoch": 1.2231892195395846,
+      "grad_norm": 0.43390517757652813,
+      "learning_rate": 8.36433806453388e-05,
+      "loss": 0.4255,
+      "step": 136
+    },
+    {
+      "epoch": 1.2321729365524985,
+      "grad_norm": 0.49443491978455834,
+      "learning_rate": 8.312002965666348e-05,
+      "loss": 0.4292,
+      "step": 137
+    },
+    {
+      "epoch": 1.2411566535654126,
+      "grad_norm": 0.5612713801192418,
+      "learning_rate": 8.259374748949619e-05,
+      "loss": 0.4248,
+      "step": 138
+    },
+    {
+      "epoch": 1.2501403705783267,
+      "grad_norm": 0.5067527019362641,
+      "learning_rate": 8.206459224328782e-05,
+      "loss": 0.4249,
+      "step": 139
+    },
+    {
+      "epoch": 1.2591240875912408,
+      "grad_norm": 0.6550752736673114,
+      "learning_rate": 8.153262233466573e-05,
+      "loss": 0.4229,
+      "step": 140
+    },
+    {
+      "epoch": 1.268107804604155,
+      "grad_norm": 0.6931302684850563,
+      "learning_rate": 8.099789649098473e-05,
+      "loss": 0.4277,
+      "step": 141
+    },
+    {
+      "epoch": 1.277091521617069,
+      "grad_norm": 0.4395328137723023,
+      "learning_rate": 8.046047374384388e-05,
+      "loss": 0.4246,
+      "step": 142
+    },
+    {
+      "epoch": 1.2860752386299832,
+      "grad_norm": 0.445103660693262,
+      "learning_rate": 7.992041342256966e-05,
+      "loss": 0.427,
+      "step": 143
+    },
+    {
+      "epoch": 1.2950589556428973,
+      "grad_norm": 0.5198361274059841,
+      "learning_rate": 7.937777514766613e-05,
+      "loss": 0.4326,
+      "step": 144
+    },
+    {
+      "epoch": 1.3040426726558114,
+      "grad_norm": 0.5037148228597795,
+      "learning_rate": 7.883261882423324e-05,
+      "loss": 0.4139,
+      "step": 145
+    },
+    {
+      "epoch": 1.3130263896687255,
+      "grad_norm": 0.5127867551211793,
+      "learning_rate": 7.828500463535335e-05,
+      "loss": 0.4259,
+      "step": 146
+    },
+    {
+      "epoch": 1.3220101066816397,
+      "grad_norm": 0.5456520135174062,
+      "learning_rate": 7.773499303544744e-05,
+      "loss": 0.4273,
+      "step": 147
+    },
+    {
+      "epoch": 1.3309938236945535,
+      "grad_norm": 0.4740377116963152,
+      "learning_rate": 7.718264474360103e-05,
+      "loss": 0.43,
+      "step": 148
+    },
+    {
+      "epoch": 1.3399775407074677,
+      "grad_norm": 0.3128198663722557,
+      "learning_rate": 7.662802073686116e-05,
+      "loss": 0.4181,
+      "step": 149
+    },
+    {
+      "epoch": 1.3489612577203818,
+      "grad_norm": 0.28499225118171256,
+      "learning_rate": 7.607118224350467e-05,
+      "loss": 0.4261,
+      "step": 150
+    },
+    {
+      "epoch": 1.357944974733296,
+      "grad_norm": 0.32298989135952116,
+      "learning_rate": 7.551219073627891e-05,
+      "loss": 0.4241,
+      "step": 151
+    },
+    {
+      "epoch": 1.36692869174621,
+      "grad_norm": 0.3419594328654335,
+      "learning_rate": 7.495110792561538e-05,
+      "loss": 0.4117,
+      "step": 152
+    },
+    {
+      "epoch": 1.3759124087591241,
+      "grad_norm": 0.286502571897559,
+      "learning_rate": 7.438799575281708e-05,
+      "loss": 0.4177,
+      "step": 153
+    },
+    {
+      "epoch": 1.3848961257720382,
+      "grad_norm": 0.3018779321311216,
+      "learning_rate": 7.382291638322053e-05,
+      "loss": 0.4194,
+      "step": 154
+    },
+    {
+      "epoch": 1.3938798427849521,
+      "grad_norm": 0.31999883299201165,
+      "learning_rate": 7.325593219933293e-05,
+      "loss": 0.4148,
+      "step": 155
+    },
+    {
+      "epoch": 1.4028635597978663,
+      "grad_norm": 0.4325078603559389,
+      "learning_rate": 7.268710579394537e-05,
+      "loss": 0.4121,
+      "step": 156
+    },
+    {
+      "epoch": 1.4118472768107804,
+      "grad_norm": 0.5275900891251157,
+      "learning_rate": 7.211649996322283e-05,
+      "loss": 0.4253,
+      "step": 157
+    },
+    {
+      "epoch": 1.4208309938236945,
+      "grad_norm": 0.5429726876656492,
+      "learning_rate": 7.154417769977177e-05,
+      "loss": 0.4147,
+      "step": 158
+    },
+    {
+      "epoch": 1.4298147108366086,
+      "grad_norm": 0.5037551180498079,
+      "learning_rate": 7.097020218568592e-05,
+      "loss": 0.4291,
+      "step": 159
+    },
+    {
+      "epoch": 1.4387984278495227,
+      "grad_norm": 0.5548276589841541,
+      "learning_rate": 7.039463678557134e-05,
+      "loss": 0.4239,
+      "step": 160
+    },
+    {
+      "epoch": 1.4477821448624368,
+      "grad_norm": 0.5552593632507392,
+      "learning_rate": 6.981754503955113e-05,
+      "loss": 0.4251,
+      "step": 161
+    },
+    {
+      "epoch": 1.456765861875351,
+      "grad_norm": 0.4270838870926948,
+      "learning_rate": 6.923899065625092e-05,
+      "loss": 0.4305,
+      "step": 162
+    },
+    {
+      "epoch": 1.465749578888265,
+      "grad_norm": 0.3489509965022947,
+      "learning_rate": 6.86590375057656e-05,
+      "loss": 0.4237,
+      "step": 163
+    },
+    {
+      "epoch": 1.4747332959011792,
+      "grad_norm": 0.26758069640190574,
+      "learning_rate": 6.807774961260844e-05,
+      "loss": 0.4202,
+      "step": 164
+    },
+    {
+      "epoch": 1.4837170129140933,
+      "grad_norm": 0.27528849391636534,
+      "learning_rate": 6.749519114864292e-05,
+      "loss": 0.4224,
+      "step": 165
+    },
+    {
+      "epoch": 1.4927007299270074,
+      "grad_norm": 0.20982060979992484,
+      "learning_rate": 6.691142642599853e-05,
+      "loss": 0.4177,
+      "step": 166
+    },
+    {
+      "epoch": 1.5016844469399215,
+      "grad_norm": 0.28760448876853106,
+      "learning_rate": 6.632651988997078e-05,
+      "loss": 0.4246,
+      "step": 167
+    },
+    {
+      "epoch": 1.5106681639528357,
+      "grad_norm": 0.2733121631233837,
+      "learning_rate": 6.5740536111907e-05,
+      "loss": 0.4241,
+      "step": 168
+    },
+    {
+      "epoch": 1.5196518809657495,
+      "grad_norm": 0.27237830941976604,
+      "learning_rate": 6.515353978207759e-05,
+      "loss": 0.412,
+      "step": 169
+    },
+    {
+      "epoch": 1.5286355979786637,
+      "grad_norm": 0.32354712657807877,
+      "learning_rate": 6.456559570253471e-05,
+      "loss": 0.4231,
+      "step": 170
+    },
+    {
+      "epoch": 1.5376193149915778,
+      "grad_norm": 0.2930435271516087,
+      "learning_rate": 6.397676877995824e-05,
+      "loss": 0.4191,
+      "step": 171
+    },
+    {
+      "epoch": 1.546603032004492,
+      "grad_norm": 0.33104600547793506,
+      "learning_rate": 6.338712401849045e-05,
+      "loss": 0.423,
+      "step": 172
+    },
+    {
+      "epoch": 1.5555867490174058,
+      "grad_norm": 0.26859987007876845,
+      "learning_rate": 6.279672651255979e-05,
+      "loss": 0.4184,
+      "step": 173
+    },
+    {
+      "epoch": 1.56457046603032,
+      "grad_norm": 0.310979899965973,
+      "learning_rate": 6.220564143969462e-05,
+      "loss": 0.4121,
+      "step": 174
+    },
+    {
+      "epoch": 1.573554183043234,
+      "grad_norm": 0.22842334872784711,
+      "learning_rate": 6.161393405332802e-05,
+      "loss": 0.426,
+      "step": 175
+    },
+    {
+      "epoch": 1.5825379000561481,
+      "grad_norm": 0.24407506306322652,
+      "learning_rate": 6.102166967559392e-05,
+      "loss": 0.4237,
+      "step": 176
+    },
+    {
+      "epoch": 1.5915216170690623,
+      "grad_norm": 0.28179593494328176,
+      "learning_rate": 6.042891369011593e-05,
+      "loss": 0.416,
+      "step": 177
+    },
+    {
+      "epoch": 1.6005053340819764,
+      "grad_norm": 0.23134259411330982,
+      "learning_rate": 5.9835731534789185e-05,
+      "loss": 0.4122,
+      "step": 178
+    },
+    {
+      "epoch": 1.6094890510948905,
+      "grad_norm": 0.2048561898994825,
+      "learning_rate": 5.924218869455623e-05,
+      "loss": 0.4211,
+      "step": 179
+    },
+    {
+      "epoch": 1.6184727681078046,
+      "grad_norm": 0.22770874161176846,
+      "learning_rate": 5.864835069417787e-05,
+      "loss": 0.4173,
+      "step": 180
+    },
+    {
+      "epoch": 1.6274564851207187,
+      "grad_norm": 0.27395747757143973,
+      "learning_rate": 5.8054283090999336e-05,
+      "loss": 0.4209,
+      "step": 181
+    },
+    {
+      "epoch": 1.6364402021336328,
+      "grad_norm": 0.2034665304138163,
+      "learning_rate": 5.746005146771308e-05,
+      "loss": 0.4137,
+      "step": 182
+    },
+    {
+      "epoch": 1.645423919146547,
+      "grad_norm": 0.22021522284044273,
+      "learning_rate": 5.68657214251188e-05,
+      "loss": 0.4088,
+      "step": 183
+    },
+    {
+      "epoch": 1.654407636159461,
+      "grad_norm": 0.20355936161816776,
+      "learning_rate": 5.627135857488122e-05,
+      "loss": 0.4094,
+      "step": 184
+    },
+    {
+      "epoch": 1.6633913531723752,
+      "grad_norm": 0.20055758377638253,
+      "learning_rate": 5.567702853228693e-05,
+      "loss": 0.4181,
+      "step": 185
+    },
+    {
+      "epoch": 1.6723750701852893,
+      "grad_norm": 0.24619886173115546,
+      "learning_rate": 5.508279690900068e-05,
+      "loss": 0.4181,
+      "step": 186
+    },
+    {
+      "epoch": 1.6813587871982034,
+      "grad_norm": 0.24468762370671132,
+      "learning_rate": 5.4488729305822145e-05,
+      "loss": 0.4176,
+      "step": 187
+    },
+    {
+      "epoch": 1.6903425042111173,
+      "grad_norm": 0.2354293474425362,
+      "learning_rate": 5.3894891305443774e-05,
+      "loss": 0.4137,
+      "step": 188
+    },
+    {
+      "epoch": 1.6993262212240314,
+      "grad_norm": 0.2095121629140633,
+      "learning_rate": 5.330134846521083e-05,
+      "loss": 0.421,
+      "step": 189
+    },
+    {
+      "epoch": 1.7083099382369455,
+      "grad_norm": 0.2592307937152872,
+      "learning_rate": 5.270816630988408e-05,
+      "loss": 0.4112,
+      "step": 190
+    },
+    {
+      "epoch": 1.7172936552498597,
+      "grad_norm": 0.2827931651894095,
+      "learning_rate": 5.211541032440609e-05,
+      "loss": 0.4208,
+      "step": 191
+    },
+    {
+      "epoch": 1.7262773722627736,
+      "grad_norm": 0.21965025238307143,
+      "learning_rate": 5.152314594667199e-05,
+      "loss": 0.4208,
+      "step": 192
+    },
+    {
+      "epoch": 1.7352610892756877,
+      "grad_norm": 0.2313849401201638,
+      "learning_rate": 5.0931438560305385e-05,
+      "loss": 0.4204,
+      "step": 193
+    },
+    {
+      "epoch": 1.7442448062886018,
+      "grad_norm": 0.25551123350164007,
+      "learning_rate": 5.034035348744023e-05,
+      "loss": 0.4123,
+      "step": 194
+    },
+    {
+      "epoch": 1.753228523301516,
+      "grad_norm": 0.20212742283098753,
+      "learning_rate": 4.974995598150956e-05,
+      "loss": 0.4217,
+      "step": 195
+    },
+    {
+      "epoch": 1.76221224031443,
+      "grad_norm": 0.23248176800502304,
+      "learning_rate": 4.9160311220041786e-05,
+      "loss": 0.4261,
+      "step": 196
+    },
+    {
+      "epoch": 1.7711959573273441,
+      "grad_norm": 0.21992166698763968,
+      "learning_rate": 4.8571484297465323e-05,
+      "loss": 0.4129,
+      "step": 197
+    },
+    {
+      "epoch": 1.7801796743402583,
+      "grad_norm": 0.2294476301587014,
+      "learning_rate": 4.798354021792242e-05,
+      "loss": 0.4056,
+      "step": 198
+    },
+    {
+      "epoch": 1.7891633913531724,
+      "grad_norm": 0.217331454755954,
+      "learning_rate": 4.739654388809303e-05,
+      "loss": 0.4107,
+      "step": 199
+    },
+    {
+      "epoch": 1.7981471083660865,
+      "grad_norm": 0.1962463954699762,
+      "learning_rate": 4.681056011002923e-05,
+      "loss": 0.4115,
+      "step": 200
+    },
+    {
+      "epoch": 1.8071308253790006,
+      "grad_norm": 0.2458933471747963,
+      "learning_rate": 4.622565357400151e-05,
+      "loss": 0.4153,
+      "step": 201
+    },
+    {
+      "epoch": 1.8161145423919147,
+      "grad_norm": 0.2763883279004427,
+      "learning_rate": 4.5641888851357086e-05,
+      "loss": 0.4281,
+      "step": 202
+    },
+    {
+      "epoch": 1.8250982594048288,
+      "grad_norm": 0.22230612055534762,
+      "learning_rate": 4.5059330387391575e-05,
+      "loss": 0.413,
+      "step": 203
+    },
+    {
+      "epoch": 1.834081976417743,
+      "grad_norm": 0.21800362362658995,
+      "learning_rate": 4.4478042494234406e-05,
+      "loss": 0.4068,
+      "step": 204
+    },
+    {
+      "epoch": 1.843065693430657,
+      "grad_norm": 0.24196596106387433,
+      "learning_rate": 4.3898089343749106e-05,
+      "loss": 0.4186,
+      "step": 205
+    },
+    {
+      "epoch": 1.8520494104435712,
+      "grad_norm": 0.23379412917214937,
+      "learning_rate": 4.331953496044888e-05,
+      "loss": 0.4069,
+      "step": 206
+    },
+    {
+      "epoch": 1.861033127456485,
+      "grad_norm": 0.195954211470671,
+      "learning_rate": 4.2742443214428685e-05,
+      "loss": 0.4143,
+      "step": 207
+    },
+    {
+      "epoch": 1.8700168444693992,
+      "grad_norm": 0.2044100499268781,
+      "learning_rate": 4.21668778143141e-05,
+      "loss": 0.4233,
+      "step": 208
+    },
+    {
+      "epoch": 1.8790005614823133,
+      "grad_norm": 0.17512498856992909,
+      "learning_rate": 4.1592902300228245e-05,
+      "loss": 0.4116,
+      "step": 209
+    },
+    {
+      "epoch": 1.8879842784952274,
+      "grad_norm": 0.19837766670845614,
+      "learning_rate": 4.1020580036777164e-05,
+      "loss": 0.4118,
+      "step": 210
+    },
+    {
+      "epoch": 1.8969679955081415,
+      "grad_norm": 0.21638156320241128,
+      "learning_rate": 4.0449974206054625e-05,
+      "loss": 0.412,
+      "step": 211
+    },
+    {
+      "epoch": 1.9059517125210554,
+      "grad_norm": 0.18508811096243785,
+      "learning_rate": 3.9881147800667064e-05,
+      "loss": 0.4205,
+      "step": 212
+    },
+    {
+      "epoch": 1.9149354295339696,
+      "grad_norm": 0.18896606868228252,
+      "learning_rate": 3.9314163616779464e-05,
+      "loss": 0.4018,
+      "step": 213
+    },
+    {
+      "epoch": 1.9239191465468837,
+      "grad_norm": 0.1839491669816972,
+      "learning_rate": 3.8749084247182924e-05,
+      "loss": 0.4207,
+      "step": 214
+    },
+    {
+      "epoch": 1.9329028635597978,
+      "grad_norm": 0.20537502525219284,
+      "learning_rate": 3.818597207438463e-05,
+      "loss": 0.411,
+      "step": 215
+    },
+    {
+      "epoch": 1.941886580572712,
+      "grad_norm": 0.2184996661258866,
+      "learning_rate": 3.762488926372108e-05,
+      "loss": 0.4148,
+      "step": 216
+    },
+    {
+      "epoch": 1.950870297585626,
+      "grad_norm": 0.1950656383059977,
+      "learning_rate": 3.706589775649533e-05,
+      "loss": 0.4042,
+      "step": 217
+    },
+    {
+      "epoch": 1.9598540145985401,
+      "grad_norm": 0.24300225745853996,
+      "learning_rate": 3.650905926313884e-05,
+      "loss": 0.4228,
+      "step": 218
+    },
+    {
+      "epoch": 1.9688377316114543,
+      "grad_norm": 0.23919079448311986,
+      "learning_rate": 3.595443525639897e-05,
+      "loss": 0.4141,
+      "step": 219
+    },
+    {
+      "epoch": 1.9778214486243684,
+      "grad_norm": 0.1772862881400631,
+      "learning_rate": 3.540208696455256e-05,
+      "loss": 0.4098,
+      "step": 220
+    },
+    {
+      "epoch": 1.9868051656372825,
+      "grad_norm": 0.26195693812853427,
+      "learning_rate": 3.485207536464666e-05,
+      "loss": 0.3992,
+      "step": 221
+    },
+    {
+      "epoch": 1.9957888826501966,
+      "grad_norm": 0.21177609314172854,
+      "learning_rate": 3.4304461175766773e-05,
+      "loss": 0.4227,
+      "step": 222
+    },
+    {
+      "epoch": 2.0061763054463784,
+      "grad_norm": 0.4458598252471374,
+      "learning_rate": 3.3759304852333874e-05,
+      "loss": 0.7234,
+      "step": 223
+    },
+    {
+      "epoch": 2.0151600224592925,
+      "grad_norm": 0.3944909722164142,
+      "learning_rate": 3.321666657743036e-05,
+      "loss": 0.364,
+      "step": 224
+    },
+    {
+      "epoch": 2.0241437394722066,
+      "grad_norm": 0.2662311407451862,
+      "learning_rate": 3.2676606256156134e-05,
+      "loss": 0.3852,
+      "step": 225
+    },
+    {
+      "epoch": 2.0331274564851207,
+      "grad_norm": 0.34100116334938124,
+      "learning_rate": 3.213918350901529e-05,
+      "loss": 0.3734,
+      "step": 226
+    },
+    {
+      "epoch": 2.042111173498035,
+      "grad_norm": 0.2873431472089421,
+      "learning_rate": 3.1604457665334285e-05,
+      "loss": 0.3779,
+      "step": 227
+    },
+    {
+      "epoch": 2.051094890510949,
+      "grad_norm": 0.27849004470855826,
+      "learning_rate": 3.107248775671218e-05,
+      "loss": 0.3698,
+      "step": 228
+    },
+    {
+      "epoch": 2.060078607523863,
+      "grad_norm": 0.3382827261005853,
+      "learning_rate": 3.054333251050381e-05,
+      "loss": 0.3847,
+      "step": 229
+    },
+    {
+      "epoch": 2.069062324536777,
+      "grad_norm": 0.25418057418257334,
+      "learning_rate": 3.0017050343336532e-05,
+      "loss": 0.3708,
+      "step": 230
+    },
+    {
+      "epoch": 2.0780460415496913,
+      "grad_norm": 0.31723965060919906,
+      "learning_rate": 2.949369935466122e-05,
+      "loss": 0.3722,
+      "step": 231
+    },
+    {
+      "epoch": 2.0870297585626054,
+      "grad_norm": 0.2601990103700952,
+      "learning_rate": 2.8973337320338337e-05,
+      "loss": 0.375,
+      "step": 232
+    },
+    {
+      "epoch": 2.0960134755755195,
+      "grad_norm": 0.262982999672935,
+      "learning_rate": 2.845602168625973e-05,
+      "loss": 0.3613,
+      "step": 233
+    },
+    {
+      "epoch": 2.1049971925884337,
+      "grad_norm": 0.26347513902884445,
+      "learning_rate": 2.7941809562006803e-05,
+      "loss": 0.376,
+      "step": 234
+    },
+    {
+      "epoch": 2.1139809096013478,
+      "grad_norm": 0.22982976993903856,
+      "learning_rate": 2.743075771454589e-05,
+      "loss": 0.3741,
+      "step": 235
+    },
+    {
+      "epoch": 2.1229646266142614,
+      "grad_norm": 0.22665105052592302,
+      "learning_rate": 2.6922922561961362e-05,
+      "loss": 0.3702,
+      "step": 236
+    },
+    {
+      "epoch": 2.1319483436271756,
+      "grad_norm": 0.21533239188280737,
+      "learning_rate": 2.6418360167227277e-05,
+      "loss": 0.3776,
+      "step": 237
+    },
+    {
+      "epoch": 2.1409320606400897,
+      "grad_norm": 0.2231373305570788,
+      "learning_rate": 2.5917126232018344e-05,
+      "loss": 0.3675,
+      "step": 238
+    },
+    {
+      "epoch": 2.149915777653004,
+      "grad_norm": 0.21229897521224303,
+      "learning_rate": 2.54192760905606e-05,
+      "loss": 0.3694,
+      "step": 239
+    },
+    {
+      "epoch": 2.158899494665918,
+      "grad_norm": 0.1819831751702882,
+      "learning_rate": 2.4924864703522748e-05,
+      "loss": 0.3717,
+      "step": 240
+    },
+    {
+      "epoch": 2.167883211678832,
+      "grad_norm": 0.21449212046236496,
+      "learning_rate": 2.443394665194875e-05,
+      "loss": 0.3619,
+      "step": 241
+    },
+    {
+      "epoch": 2.176866928691746,
+      "grad_norm": 0.16892264582409286,
+      "learning_rate": 2.3946576131232258e-05,
+      "loss": 0.3842,
+      "step": 242
+    },
+    {
+      "epoch": 2.1858506457046603,
+      "grad_norm": 0.2371411090288409,
+      "learning_rate": 2.3462806945133682e-05,
+      "loss": 0.3897,
+      "step": 243
+    },
+    {
+      "epoch": 2.1948343627175744,
+      "grad_norm": 0.1607253638516768,
+      "learning_rate": 2.2982692499840452e-05,
+      "loss": 0.3652,
+      "step": 244
+    },
+    {
+      "epoch": 2.2038180797304885,
+      "grad_norm": 0.17103269591436154,
+      "learning_rate": 2.2506285798071218e-05,
+      "loss": 0.369,
+      "step": 245
+    },
+    {
+      "epoch": 2.2128017967434026,
+      "grad_norm": 0.16967120460293925,
+      "learning_rate": 2.203363943322452e-05,
+      "loss": 0.3639,
+      "step": 246
+    },
+    {
+      "epoch": 2.2217855137563167,
+      "grad_norm": 0.16154312491100173,
+      "learning_rate": 2.1564805583572732e-05,
+      "loss": 0.3742,
+      "step": 247
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.16774894199509377,
+      "learning_rate": 2.109983600650176e-05,
+      "loss": 0.3607,
+      "step": 248
+    },
+    {
+      "epoch": 2.239752947782145,
+      "grad_norm": 0.16024982566651816,
+      "learning_rate": 2.0638782032797214e-05,
+      "loss": 0.3693,
+      "step": 249
+    },
+    {
+      "epoch": 2.248736664795059,
+      "grad_norm": 0.14084974911793113,
+      "learning_rate": 2.018169456097774e-05,
+      "loss": 0.369,
+      "step": 250
+    },
+    {
+      "epoch": 2.257720381807973,
+      "grad_norm": 0.15128389306058487,
+      "learning_rate": 1.9728624051676018e-05,
+      "loss": 0.3738,
+      "step": 251
+    },
+    {
+      "epoch": 2.2667040988208873,
+      "grad_norm": 0.15091333660825978,
+      "learning_rate": 1.927962052206808e-05,
+      "loss": 0.3798,
+      "step": 252
+    },
+    {
+      "epoch": 2.2756878158338014,
+      "grad_norm": 0.14371641854325717,
+      "learning_rate": 1.8834733540351643e-05,
+      "loss": 0.3666,
+      "step": 253
+    },
+    {
+      "epoch": 2.2846715328467155,
+      "grad_norm": 0.126210618892879,
+      "learning_rate": 1.8394012220273954e-05,
+      "loss": 0.3853,
+      "step": 254
+    },
+    {
+      "epoch": 2.293655249859629,
+      "grad_norm": 0.1336658722457838,
+      "learning_rate": 1.7957505215709848e-05,
+      "loss": 0.3701,
+      "step": 255
+    },
+    {
+      "epoch": 2.3026389668725433,
+      "grad_norm": 0.16030704601719437,
+      "learning_rate": 1.7525260715290535e-05,
+      "loss": 0.3786,
+      "step": 256
+    },
+    {
+      "epoch": 2.3116226838854574,
+      "grad_norm": 0.11861257403655912,
+      "learning_rate": 1.7097326437083794e-05,
+      "loss": 0.3772,
+      "step": 257
+    },
+    {
+      "epoch": 2.3206064008983716,
+      "grad_norm": 0.15802244831319096,
+      "learning_rate": 1.6673749623326064e-05,
+      "loss": 0.3641,
+      "step": 258
+    },
+    {
+      "epoch": 2.3295901179112857,
+      "grad_norm": 0.12362312515039316,
+      "learning_rate": 1.6254577035207113e-05,
+      "loss": 0.3832,
+      "step": 259
+    },
+    {
+      "epoch": 2.3385738349242,
+      "grad_norm": 0.12418525205040346,
+      "learning_rate": 1.5839854947707754e-05,
+      "loss": 0.3587,
+      "step": 260
+    },
+    {
+      "epoch": 2.347557551937114,
+      "grad_norm": 0.13981066977684836,
+      "learning_rate": 1.5429629144491303e-05,
+      "loss": 0.3791,
+      "step": 261
+    },
+    {
+      "epoch": 2.356541268950028,
+      "grad_norm": 0.11943711118891571,
+      "learning_rate": 1.5023944912849248e-05,
+      "loss": 0.3761,
+      "step": 262
+    },
+    {
+      "epoch": 2.365524985962942,
+      "grad_norm": 0.1267710969494925,
+      "learning_rate": 1.462284703870168e-05,
+      "loss": 0.3717,
+      "step": 263
+    },
+    {
+      "epoch": 2.3745087029758563,
+      "grad_norm": 0.12621457698502858,
+      "learning_rate": 1.4226379801653136e-05,
+      "loss": 0.3757,
+      "step": 264
+    },
+    {
+      "epoch": 2.3834924199887704,
+      "grad_norm": 0.13188965610725137,
+      "learning_rate": 1.3834586970104307e-05,
+      "loss": 0.3765,
+      "step": 265
+    },
+    {
+      "epoch": 2.3924761370016845,
+      "grad_norm": 0.129108528771287,
+      "learning_rate": 1.3447511796420162e-05,
+      "loss": 0.3795,
+      "step": 266
+    },
+    {
+      "epoch": 2.4014598540145986,
+      "grad_norm": 0.12702657201457457,
+      "learning_rate": 1.306519701215506e-05,
+      "loss": 0.3659,
+      "step": 267
+    },
+    {
+      "epoch": 2.4104435710275127,
+      "grad_norm": 0.12474862672561715,
+      "learning_rate": 1.2687684823335395e-05,
+      "loss": 0.3778,
+      "step": 268
+    },
+    {
+      "epoch": 2.419427288040427,
+      "grad_norm": 0.13722039305391268,
+      "learning_rate": 1.2315016905800158e-05,
+      "loss": 0.3817,
+      "step": 269
+    },
+    {
+      "epoch": 2.428411005053341,
+      "grad_norm": 0.10994259673231212,
+      "learning_rate": 1.1947234400600164e-05,
+      "loss": 0.361,
+      "step": 270
+    },
+    {
+      "epoch": 2.437394722066255,
+      "grad_norm": 0.11921690061515447,
+      "learning_rate": 1.1584377909456147e-05,
+      "loss": 0.3642,
+      "step": 271
+    },
+    {
+      "epoch": 2.446378439079169,
+      "grad_norm": 0.14125660857982086,
+      "learning_rate": 1.1226487490276591e-05,
+      "loss": 0.3727,
+      "step": 272
+    },
+    {
+      "epoch": 2.4553621560920833,
+      "grad_norm": 0.10666153897619447,
+      "learning_rate": 1.087360265273545e-05,
+      "loss": 0.3647,
+      "step": 273
+    },
+    {
+      "epoch": 2.464345873104997,
+      "grad_norm": 0.11519070228670822,
+      "learning_rate": 1.0525762353910433e-05,
+      "loss": 0.383,
+      "step": 274
+    },
+    {
+      "epoch": 2.473329590117911,
+      "grad_norm": 0.1256473190699589,
+      "learning_rate": 1.018300499398232e-05,
+      "loss": 0.3737,
+      "step": 275
+    },
+    {
+      "epoch": 2.482313307130825,
+      "grad_norm": 0.12764646963019577,
+      "learning_rate": 9.8453684119957e-06,
+      "loss": 0.3781,
+      "step": 276
+    },
+    {
+      "epoch": 2.4912970241437393,
+      "grad_norm": 0.11284786609558817,
+      "learning_rate": 9.512889881681736e-06,
+      "loss": 0.3589,
+      "step": 277
+    },
+    {
+      "epoch": 2.5002807411566534,
+      "grad_norm": 0.10687489300468084,
+      "learning_rate": 9.185606107343261e-06,
+      "loss": 0.36,
+      "step": 278
+    },
+    {
+      "epoch": 2.5092644581695676,
+      "grad_norm": 0.11023304943251037,
+      "learning_rate": 8.863553219802784e-06,
+      "loss": 0.3706,
+      "step": 279
+    },
+    {
+      "epoch": 2.5182481751824817,
+      "grad_norm": 0.11863625673352954,
+      "learning_rate": 8.546766772413786e-06,
+      "loss": 0.3807,
+      "step": 280
+    },
+    {
+      "epoch": 2.527231892195396,
+      "grad_norm": 0.10361273392283875,
+      "learning_rate": 8.23528173713578e-06,
+      "loss": 0.3637,
+      "step": 281
+    },
+    {
+      "epoch": 2.53621560920831,
+      "grad_norm": 0.10388585690401946,
+      "learning_rate": 7.929132500673536e-06,
+      "loss": 0.3667,
+      "step": 282
+    },
+    {
+      "epoch": 2.545199326221224,
+      "grad_norm": 0.11252208805304847,
+      "learning_rate": 7.628352860680906e-06,
+      "loss": 0.3821,
+      "step": 283
+    },
+    {
+      "epoch": 2.554183043234138,
+      "grad_norm": 0.11235953219504816,
+      "learning_rate": 7.332976022029765e-06,
+      "loss": 0.3583,
+      "step": 284
+    },
+    {
+      "epoch": 2.5631667602470523,
+      "grad_norm": 0.11156514177993415,
+      "learning_rate": 7.0430345931442495e-06,
+      "loss": 0.387,
+      "step": 285
+    },
+    {
+      "epoch": 2.5721504772599664,
+      "grad_norm": 0.11165023009158465,
+      "learning_rate": 6.758560582400965e-06,
+      "loss": 0.3747,
+      "step": 286
+    },
+    {
+      "epoch": 2.5811341942728805,
+      "grad_norm": 0.11083982189825406,
+      "learning_rate": 6.479585394595375e-06,
+      "loss": 0.3757,
+      "step": 287
+    },
+    {
+      "epoch": 2.5901179112857946,
+      "grad_norm": 0.09734475490326093,
+      "learning_rate": 6.206139827474829e-06,
+      "loss": 0.3641,
+      "step": 288
+    },
+    {
+      "epoch": 2.5991016282987087,
+      "grad_norm": 0.0969288972848367,
+      "learning_rate": 5.938254068338624e-06,
+      "loss": 0.3646,
+      "step": 289
+    },
+    {
+      "epoch": 2.608085345311623,
+      "grad_norm": 0.10426983457210305,
+      "learning_rate": 5.675957690705446e-06,
+      "loss": 0.381,
+      "step": 290
+    },
+    {
+      "epoch": 2.6170690623245365,
+      "grad_norm": 0.10116862032974816,
+      "learning_rate": 5.419279651048566e-06,
+      "loss": 0.3748,
+      "step": 291
+    },
+    {
+      "epoch": 2.626052779337451,
+      "grad_norm": 0.09343396040555912,
+      "learning_rate": 5.168248285599177e-06,
+      "loss": 0.3605,
+      "step": 292
+    },
+    {
+      "epoch": 2.6350364963503647,
+      "grad_norm": 0.09927514498945143,
+      "learning_rate": 4.922891307218158e-06,
+      "loss": 0.3778,
+      "step": 293
+    },
+    {
+      "epoch": 2.6440202133632793,
+      "grad_norm": 0.0970025854343804,
+      "learning_rate": 4.6832358023367316e-06,
+      "loss": 0.3748,
+      "step": 294
+    },
+    {
+      "epoch": 2.653003930376193,
+      "grad_norm": 0.0974844589022167,
+      "learning_rate": 4.449308227966143e-06,
+      "loss": 0.366,
+      "step": 295
+    },
+    {
+      "epoch": 2.661987647389107,
+      "grad_norm": 0.09814487372296889,
+      "learning_rate": 4.221134408777029e-06,
+      "loss": 0.3757,
+      "step": 296
+    },
+    {
+      "epoch": 2.670971364402021,
+      "grad_norm": 0.09888940980952132,
+      "learning_rate": 3.998739534248389e-06,
+      "loss": 0.3727,
+      "step": 297
+    },
+    {
+      "epoch": 2.6799550814149353,
+      "grad_norm": 0.10378096684924529,
+      "learning_rate": 3.782148155886801e-06,
+      "loss": 0.3719,
+      "step": 298
+    },
+    {
+      "epoch": 2.6889387984278494,
+      "grad_norm": 0.09900578604392633,
+      "learning_rate": 3.5713841845160227e-06,
+      "loss": 0.3649,
+      "step": 299
+    },
+    {
+      "epoch": 2.6979225154407636,
+      "grad_norm": 0.09639119313849992,
+      "learning_rate": 3.366470887637346e-06,
+      "loss": 0.3638,
+      "step": 300
+    },
+    {
+      "epoch": 2.7069062324536777,
+      "grad_norm": 0.0918907017548552,
+      "learning_rate": 3.1674308868609408e-06,
+      "loss": 0.3732,
+      "step": 301
+    },
+    {
+      "epoch": 2.715889949466592,
+      "grad_norm": 0.10289981723072597,
+      "learning_rate": 2.9742861554085396e-06,
+      "loss": 0.3798,
+      "step": 302
+    },
+    {
+      "epoch": 2.724873666479506,
+      "grad_norm": 0.09970818874140677,
+      "learning_rate": 2.7870580156876457e-06,
+      "loss": 0.3666,
+      "step": 303
+    },
+    {
+      "epoch": 2.73385738349242,
+      "grad_norm": 0.09359734297487839,
+      "learning_rate": 2.6057671369376755e-06,
+      "loss": 0.3622,
+      "step": 304
+    },
+    {
+      "epoch": 2.742841100505334,
+      "grad_norm": 0.09291658961787892,
+      "learning_rate": 2.4304335329481117e-06,
+      "loss": 0.3715,
+      "step": 305
+    },
+    {
+      "epoch": 2.7518248175182483,
+      "grad_norm": 0.09173829150247041,
+      "learning_rate": 2.2610765598490875e-06,
+      "loss": 0.3631,
+      "step": 306
+    },
+    {
+      "epoch": 2.7608085345311624,
+      "grad_norm": 0.08660766388118828,
+      "learning_rate": 2.097714913974531e-06,
+      "loss": 0.3733,
+      "step": 307
+    },
+    {
+      "epoch": 2.7697922515440765,
+      "grad_norm": 0.09369793793404498,
+      "learning_rate": 1.940366629798168e-06,
+      "loss": 0.3563,
+      "step": 308
+    },
+    {
+      "epoch": 2.7787759685569906,
+      "grad_norm": 0.09134341611654641,
+      "learning_rate": 1.7890490779426058e-06,
+      "loss": 0.388,
+      "step": 309
+    },
+    {
+      "epoch": 2.7877596855699043,
+      "grad_norm": 0.09306090597773946,
+      "learning_rate": 1.6437789632616797e-06,
+      "loss": 0.3562,
+      "step": 310
+    },
+    {
+      "epoch": 2.796743402582819,
+      "grad_norm": 0.09113239444716176,
+      "learning_rate": 1.504572322996265e-06,
+      "loss": 0.3765,
+      "step": 311
+    },
+    {
+      "epoch": 2.8057271195957325,
+      "grad_norm": 0.09201136109364715,
+      "learning_rate": 1.371444525003882e-06,
+      "loss": 0.3744,
+      "step": 312
+    },
+    {
+      "epoch": 2.814710836608647,
+      "grad_norm": 0.09271086633176956,
+      "learning_rate": 1.2444102660621238e-06,
+      "loss": 0.367,
+      "step": 313
+    },
+    {
+      "epoch": 2.8236945536215607,
+      "grad_norm": 0.09022777999963022,
+      "learning_rate": 1.1234835702461861e-06,
+      "loss": 0.3731,
+      "step": 314
+    },
+    {
+      "epoch": 2.832678270634475,
+      "grad_norm": 0.08999529135727959,
+      "learning_rate": 1.0086777873806807e-06,
+      "loss": 0.3676,
+      "step": 315
+    },
+    {
+      "epoch": 2.841661987647389,
+      "grad_norm": 0.0868157397412178,
+      "learning_rate": 9.00005591565833e-07,
+      "loss": 0.3709,
+      "step": 316
+    },
+    {
+      "epoch": 2.850645704660303,
+      "grad_norm": 0.0846034105807787,
+      "learning_rate": 7.974789797783608e-07,
+      "loss": 0.3776,
+      "step": 317
+    },
+    {
+      "epoch": 2.859629421673217,
+      "grad_norm": 0.08485636835150868,
+      "learning_rate": 7.01109270547026e-07,
+      "loss": 0.3717,
+      "step": 318
+    },
+    {
+      "epoch": 2.8686131386861313,
+      "grad_norm": 0.0945293909871086,
+      "learning_rate": 6.109071027031233e-07,
+      "loss": 0.3679,
+      "step": 319
+    },
+    {
+      "epoch": 2.8775968556990454,
+      "grad_norm": 0.09014386290340982,
+      "learning_rate": 5.268824342059822e-07,
+      "loss": 0.3753,
+      "step": 320
+    },
+    {
+      "epoch": 2.8865805727119596,
+      "grad_norm": 0.08280800725097154,
+      "learning_rate": 4.4904454104367007e-07,
+      "loss": 0.3693,
+      "step": 321
+    },
+    {
+      "epoch": 2.8955642897248737,
+      "grad_norm": 0.08375874995326571,
+      "learning_rate": 3.7740201620894e-07,
+      "loss": 0.3683,
+      "step": 322
+    },
+    {
+      "epoch": 2.904548006737788,
+      "grad_norm": 0.0859954787257025,
+      "learning_rate": 3.119627687506124e-07,
+      "loss": 0.3729,
+      "step": 323
+    },
+    {
+      "epoch": 2.913531723750702,
+      "grad_norm": 0.0871794922149879,
+      "learning_rate": 2.527340229004407e-07,
+      "loss": 0.3664,
+      "step": 324
+    },
+    {
+      "epoch": 2.922515440763616,
+      "grad_norm": 0.09003709121214834,
+      "learning_rate": 1.9972231727558538e-07,
+      "loss": 0.3773,
+      "step": 325
+    },
+    {
+      "epoch": 2.93149915777653,
+      "grad_norm": 0.0825913756316724,
+      "learning_rate": 1.529335041567867e-07,
+      "loss": 0.3618,
+      "step": 326
+    },
+    {
+      "epoch": 2.9404828747894443,
+      "grad_norm": 0.0839103146643486,
+      "learning_rate": 1.1237274884227121e-07,
+      "loss": 0.3598,
+      "step": 327
+    },
+    {
+      "epoch": 2.9494665918023584,
+      "grad_norm": 0.08504355567985683,
+      "learning_rate": 7.804452907755089e-08,
+      "loss": 0.3728,
+      "step": 328
+    },
+    {
+      "epoch": 2.958450308815272,
+      "grad_norm": 0.08522801556822512,
+      "learning_rate": 4.9952634561082644e-08,
+      "loss": 0.3682,
+      "step": 329
+    },
+    {
+      "epoch": 2.9674340258281866,
+      "grad_norm": 0.08778820538143113,
+      "learning_rate": 2.810016652591411e-08,
+      "loss": 0.3708,
+      "step": 330
+    },
+    {
+      "epoch": 2.9764177428411003,
+      "grad_norm": 0.09049437080438882,
+      "learning_rate": 1.248953739730298e-08,
+      "loss": 0.3737,
+      "step": 331
+    },
+    {
+      "epoch": 2.985401459854015,
+      "grad_norm": 0.08443128930679393,
+      "learning_rate": 3.1224705263979394e-09,
+      "loss": 0.3671,
+      "step": 332
+    },
+    {
+      "epoch": 2.9943851768669285,
+      "grad_norm": 0.09016237166909961,
+      "learning_rate": 0.0,
+      "loss": 0.3706,
+      "step": 333
+    },
+    {
+      "epoch": 2.9943851768669285,
+      "step": 333,
+      "total_flos": 4607052149424128.0,
+      "train_loss": 0.44577710639249096,
+      "train_runtime": 115174.1649,
+      "train_samples_per_second": 2.968,
+      "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
+}