Model save

README.md

@@ -20,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.3544
+- Loss: 1.3529
 
 ## Model description
 
@@ -56,9 +56,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.2658        | 1.0   | 326  | 1.2135          |
-| 0.8841        | 2.0   | 652  | 1.2135          |
-| 0.7012        | 3.0   | 978  | 1.3544          |
+| 1.265         | 1.0   | 326  | 1.2147          |
+| 0.8983        | 2.0   | 652  | 1.2119          |
+| 0.7007        | 3.0   | 978  | 1.3529          |
 
 
 ### Framework versions

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 3.0,
     "total_flos": 3642602029056.0,
-    "train_loss": 1.0360665048314506,
-    "train_runtime": 586.176,
+    "train_loss": 1.0182966589927673,
+    "train_runtime": 588.3296,
     "train_samples": 2023,
-    "train_samples_per_second": 6.674,
-    "train_steps_per_second": 1.668
+    "train_samples_per_second": 6.649,
+    "train_steps_per_second": 1.662
 }

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f03ed4af2ea91e8b3cbbc4638db271c30b32401880813e3fffa2a20ad9a26c9
+oid sha256:60939b5e22709c00ec96520ade09dcf2b14bdabe4696c70cb6682237cc9a2418
 size 4945242264

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3070c72853fac5da48f03f70e9348094fc053d7b892771281116c46eed2ff01
+oid sha256:b4723cbfe53b0efa358675b46c28db4b13ac7b62f1db5b842c3fb64d4c56a47c
 size 67121608

runs/Oct30_00-32-00_tinkywinky.edc.iee.ucsb.edu/events.out.tfevents.1730273532.tinkywinky.edc.iee.ucsb.edu.2174734.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbc6cf7a78668c3b6db42bb5db843fabf5fc63b475186313ebd5e38c8f8d797f
+size 48234

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 3.0,
     "total_flos": 3642602029056.0,
-    "train_loss": 1.0360665048314506,
-    "train_runtime": 586.176,
+    "train_loss": 1.0182966589927673,
+    "train_runtime": 588.3296,
     "train_samples": 2023,
-    "train_samples_per_second": 6.674,
-    "train_steps_per_second": 1.668
+    "train_samples_per_second": 6.649,
+    "train_steps_per_second": 1.662
 }

trainer_state.json

@@ -10,1408 +10,1408 @@
   "log_history": [
     {
       "epoch": 0.003067484662576687,
-      "grad_norm": 177.2131714128933,
+      "grad_norm": 177.21839775031285,
       "learning_rate": 2.0408163265306121e-07,
       "loss": 3.0795,
       "step": 1
     },
     {
       "epoch": 0.015337423312883436,
-      "grad_norm": 241.89792017630023,
+      "grad_norm": 81.30898025729103,
       "learning_rate": 1.0204081632653063e-06,
-      "loss": 2.738,
+      "loss": 2.7402,
       "step": 5
     },
     {
       "epoch": 0.03067484662576687,
-      "grad_norm": 97.9005821739992,
+      "grad_norm": 107.55431230634855,
       "learning_rate": 2.0408163265306125e-06,
-      "loss": 2.573,
+      "loss": 2.5637,
       "step": 10
     },
     {
       "epoch": 0.046012269938650305,
-      "grad_norm": 29.935305969711827,
+      "grad_norm": 17.843083480868863,
       "learning_rate": 3.0612244897959185e-06,
-      "loss": 2.4113,
+      "loss": 2.3819,
       "step": 15
     },
     {
       "epoch": 0.06134969325153374,
-      "grad_norm": 6.7865011448005985,
+      "grad_norm": 5.541698118155843,
       "learning_rate": 4.081632653061225e-06,
-      "loss": 1.9856,
+      "loss": 1.9572,
       "step": 20
     },
     {
       "epoch": 0.07668711656441718,
-      "grad_norm": 8.264677565869711,
+      "grad_norm": 10.409288444104131,
       "learning_rate": 5.1020408163265315e-06,
-      "loss": 1.8726,
+      "loss": 1.8593,
       "step": 25
     },
     {
       "epoch": 0.09202453987730061,
-      "grad_norm": 19.32485419745683,
+      "grad_norm": 5.082002576187776,
       "learning_rate": 6.122448979591837e-06,
-      "loss": 1.7009,
+      "loss": 1.6901,
       "step": 30
     },
     {
       "epoch": 0.10736196319018405,
-      "grad_norm": 10.162135754390622,
+      "grad_norm": 6.765872287855098,
       "learning_rate": 7.1428571428571436e-06,
-      "loss": 1.6191,
+      "loss": 1.6053,
       "step": 35
     },
     {
       "epoch": 0.12269938650306748,
-      "grad_norm": 3.0916095898500555,
+      "grad_norm": 11.409653818105163,
       "learning_rate": 8.16326530612245e-06,
-      "loss": 1.5164,
+      "loss": 1.5093,
       "step": 40
     },
     {
       "epoch": 0.13803680981595093,
-      "grad_norm": 43.413150705371685,
+      "grad_norm": 6.4162354324510895,
       "learning_rate": 9.183673469387756e-06,
-      "loss": 1.5097,
+      "loss": 1.5069,
       "step": 45
     },
     {
       "epoch": 0.15337423312883436,
-      "grad_norm": 49.53965734626817,
+      "grad_norm": 3.7675519341487136,
       "learning_rate": 1.0204081632653063e-05,
-      "loss": 1.5785,
+      "loss": 1.575,
       "step": 50
     },
     {
       "epoch": 0.1687116564417178,
-      "grad_norm": 10.98873172852826,
+      "grad_norm": 4.369929690874893,
       "learning_rate": 1.1224489795918367e-05,
-      "loss": 1.4715,
+      "loss": 1.464,
       "step": 55
     },
     {
       "epoch": 0.18404907975460122,
-      "grad_norm": 2.7742625393808584,
+      "grad_norm": 5.732870449440104,
       "learning_rate": 1.2244897959183674e-05,
-      "loss": 1.4263,
+      "loss": 1.4131,
       "step": 60
     },
     {
       "epoch": 0.19938650306748465,
-      "grad_norm": 23.101404729255652,
+      "grad_norm": 2.9718012621338654,
       "learning_rate": 1.326530612244898e-05,
-      "loss": 1.4043,
+      "loss": 1.3903,
       "step": 65
     },
     {
       "epoch": 0.2147239263803681,
-      "grad_norm": 12.92305440472016,
+      "grad_norm": 2.986883686963562,
       "learning_rate": 1.4285714285714287e-05,
-      "loss": 1.3793,
+      "loss": 1.3699,
       "step": 70
     },
     {
       "epoch": 0.23006134969325154,
-      "grad_norm": 2.882766144097806,
+      "grad_norm": 3.0609005774260116,
       "learning_rate": 1.530612244897959e-05,
-      "loss": 1.4183,
+      "loss": 1.4099,
       "step": 75
     },
     {
       "epoch": 0.24539877300613497,
-      "grad_norm": 11.111007796269925,
+      "grad_norm": 3.391633696280962,
       "learning_rate": 1.63265306122449e-05,
-      "loss": 1.4517,
+      "loss": 1.4246,
       "step": 80
     },
     {
       "epoch": 0.2607361963190184,
-      "grad_norm": 2.897394026367856,
+      "grad_norm": 2.446544689335844,
       "learning_rate": 1.7346938775510206e-05,
-      "loss": 1.3393,
+      "loss": 1.3225,
       "step": 85
     },
     {
       "epoch": 0.27607361963190186,
-      "grad_norm": 2.9177477230316566,
+      "grad_norm": 2.892905300954704,
       "learning_rate": 1.836734693877551e-05,
-      "loss": 1.428,
+      "loss": 1.4131,
       "step": 90
     },
     {
       "epoch": 0.29141104294478526,
-      "grad_norm": 2.912560429885746,
+      "grad_norm": 2.7276895053980588,
       "learning_rate": 1.9387755102040817e-05,
-      "loss": 1.304,
+      "loss": 1.2981,
       "step": 95
     },
     {
       "epoch": 0.3067484662576687,
-      "grad_norm": 2.5615070520092704,
+      "grad_norm": 3.1096897823921212,
       "learning_rate": 1.9999745104274995e-05,
-      "loss": 1.2997,
+      "loss": 1.2867,
       "step": 100
     },
     {
       "epoch": 0.3220858895705521,
-      "grad_norm": 2.9480036513492336,
+      "grad_norm": 2.6017423936904605,
       "learning_rate": 1.9996877676598733e-05,
-      "loss": 1.4086,
+      "loss": 1.3923,
       "step": 105
     },
     {
       "epoch": 0.3374233128834356,
-      "grad_norm": 2.5407276803281333,
+      "grad_norm": 2.357782746709104,
       "learning_rate": 1.9990825118233958e-05,
-      "loss": 1.3028,
+      "loss": 1.2889,
       "step": 110
     },
     {
       "epoch": 0.35276073619631904,
-      "grad_norm": 3.239233594777726,
+      "grad_norm": 4.4128065982273155,
       "learning_rate": 1.9981589357601727e-05,
-      "loss": 1.4185,
+      "loss": 1.4107,
       "step": 115
     },
     {
       "epoch": 0.36809815950920244,
-      "grad_norm": 2.771896819307909,
+      "grad_norm": 3.425193387107373,
       "learning_rate": 1.9969173337331283e-05,
-      "loss": 1.3523,
+      "loss": 1.3417,
       "step": 120
     },
     {
       "epoch": 0.3834355828220859,
-      "grad_norm": 2.7643269799604724,
+      "grad_norm": 2.8732197303122518,
       "learning_rate": 1.9953581013322503e-05,
-      "loss": 1.3175,
+      "loss": 1.3121,
       "step": 125
     },
     {
       "epoch": 0.3987730061349693,
-      "grad_norm": 3.508281226717504,
+      "grad_norm": 3.121331048192887,
       "learning_rate": 1.99348173534855e-05,
-      "loss": 1.2981,
+      "loss": 1.2892,
       "step": 130
     },
     {
       "epoch": 0.41411042944785276,
-      "grad_norm": 2.4302828555210025,
+      "grad_norm": 2.340124072432182,
       "learning_rate": 1.9912888336157793e-05,
-      "loss": 1.3109,
+      "loss": 1.3019,
       "step": 135
     },
     {
       "epoch": 0.4294478527607362,
-      "grad_norm": 2.3273613141747975,
+      "grad_norm": 2.5077068756467393,
       "learning_rate": 1.9887800948199496e-05,
-      "loss": 1.2887,
+      "loss": 1.2821,
       "step": 140
     },
     {
       "epoch": 0.4447852760736196,
-      "grad_norm": 2.502367541877984,
+      "grad_norm": 5.649859070977053,
       "learning_rate": 1.9859563182767268e-05,
-      "loss": 1.2735,
+      "loss": 1.2652,
       "step": 145
     },
     {
       "epoch": 0.4601226993865031,
-      "grad_norm": 2.8584207562233113,
+      "grad_norm": 2.852076057226547,
       "learning_rate": 1.9828184036767556e-05,
-      "loss": 1.2271,
+      "loss": 1.2215,
       "step": 150
     },
     {
       "epoch": 0.4754601226993865,
-      "grad_norm": 2.190347531165954,
+      "grad_norm": 2.3152692963670853,
       "learning_rate": 1.9793673507990086e-05,
-      "loss": 1.2896,
+      "loss": 1.2821,
       "step": 155
     },
     {
       "epoch": 0.49079754601226994,
-      "grad_norm": 2.2761435625476545,
+      "grad_norm": 2.9501006805800682,
       "learning_rate": 1.9756042591922436e-05,
-      "loss": 1.3294,
+      "loss": 1.3246,
       "step": 160
     },
     {
       "epoch": 0.5061349693251533,
-      "grad_norm": 2.220138706256322,
+      "grad_norm": 2.208339912352479,
       "learning_rate": 1.9715303278246724e-05,
-      "loss": 1.3188,
+      "loss": 1.3105,
       "step": 165
     },
     {
       "epoch": 0.5214723926380368,
-      "grad_norm": 2.294310010674355,
+      "grad_norm": 2.377033071477205,
       "learning_rate": 1.9671468547019575e-05,
-      "loss": 1.2199,
+      "loss": 1.2158,
       "step": 170
     },
     {
       "epoch": 0.5368098159509203,
-      "grad_norm": 2.513372192362633,
+      "grad_norm": 2.2769169863909706,
       "learning_rate": 1.9624552364536472e-05,
-      "loss": 1.3106,
+      "loss": 1.3063,
       "step": 175
     },
     {
       "epoch": 0.5521472392638037,
-      "grad_norm": 2.510468468716461,
+      "grad_norm": 3.488802301164089,
       "learning_rate": 1.9574569678881965e-05,
-      "loss": 1.3708,
+      "loss": 1.3641,
       "step": 180
     },
     {
       "epoch": 0.5674846625766872,
-      "grad_norm": 2.041921641107936,
+      "grad_norm": 3.2567021860207923,
       "learning_rate": 1.952153641516698e-05,
-      "loss": 1.1784,
+      "loss": 1.1764,
       "step": 185
     },
     {
       "epoch": 0.5828220858895705,
-      "grad_norm": 2.27820722339295,
+      "grad_norm": 2.5611122809445117,
       "learning_rate": 1.94654694704549e-05,
-      "loss": 1.1902,
+      "loss": 1.187,
       "step": 190
     },
     {
       "epoch": 0.598159509202454,
-      "grad_norm": 2.328841014102968,
+      "grad_norm": 3.2553565189736906,
       "learning_rate": 1.9406386708377956e-05,
-      "loss": 1.316,
+      "loss": 1.3129,
       "step": 195
     },
     {
       "epoch": 0.6134969325153374,
-      "grad_norm": 2.2926147060214066,
+      "grad_norm": 2.866192118065629,
       "learning_rate": 1.9344306953445632e-05,
-      "loss": 1.3843,
+      "loss": 1.3816,
       "step": 200
     },
     {
       "epoch": 0.6288343558282209,
-      "grad_norm": 3.113673918623077,
+      "grad_norm": 2.729878771251503,
       "learning_rate": 1.9279249985046948e-05,
-      "loss": 1.3199,
+      "loss": 1.3208,
       "step": 205
     },
     {
       "epoch": 0.6441717791411042,
-      "grad_norm": 2.105413236829464,
+      "grad_norm": 2.1585127050514115,
       "learning_rate": 1.92112365311485e-05,
-      "loss": 1.2593,
+      "loss": 1.2483,
       "step": 210
     },
     {
       "epoch": 0.6595092024539877,
-      "grad_norm": 2.2971282506758204,
+      "grad_norm": 2.884411746638962,
       "learning_rate": 1.9140288261690278e-05,
-      "loss": 1.2316,
+      "loss": 1.2221,
       "step": 215
     },
     {
       "epoch": 0.6748466257668712,
-      "grad_norm": 3.060568400530995,
+      "grad_norm": 2.3400822227419704,
       "learning_rate": 1.9066427781681314e-05,
-      "loss": 1.1986,
+      "loss": 1.1909,
       "step": 220
     },
     {
       "epoch": 0.6901840490797546,
-      "grad_norm": 2.0851577530985663,
+      "grad_norm": 2.2424426077461224,
       "learning_rate": 1.8989678623997506e-05,
-      "loss": 1.2465,
+      "loss": 1.2427,
       "step": 225
     },
     {
       "epoch": 0.7055214723926381,
-      "grad_norm": 2.34599375716676,
+      "grad_norm": 2.3532187548733847,
       "learning_rate": 1.891006524188368e-05,
-      "loss": 1.3111,
+      "loss": 1.3086,
       "step": 230
     },
     {
       "epoch": 0.7208588957055214,
-      "grad_norm": 2.0045103451147352,
+      "grad_norm": 2.1575864354340557,
       "learning_rate": 1.8827613001162534e-05,
-      "loss": 1.2048,
+      "loss": 1.2051,
       "step": 235
     },
     {
       "epoch": 0.7361963190184049,
-      "grad_norm": 2.0865280550933663,
+      "grad_norm": 2.2528114135340807,
       "learning_rate": 1.8742348172152728e-05,
-      "loss": 1.3875,
+      "loss": 1.3883,
       "step": 240
     },
     {
       "epoch": 0.7515337423312883,
-      "grad_norm": 2.1899054440028602,
+      "grad_norm": 2.20955115974882,
       "learning_rate": 1.8654297921298862e-05,
-      "loss": 1.2177,
+      "loss": 1.2169,
       "step": 245
     },
     {
       "epoch": 0.7668711656441718,
-      "grad_norm": 2.146615168451316,
+      "grad_norm": 2.1026008601865933,
       "learning_rate": 1.856349030251589e-05,
-      "loss": 1.2302,
+      "loss": 1.2356,
       "step": 250
     },
     {
       "epoch": 0.7822085889570553,
-      "grad_norm": 5.215245761172734,
+      "grad_norm": 4.721033676079731,
       "learning_rate": 1.846995424825079e-05,
-      "loss": 1.1558,
+      "loss": 1.1575,
       "step": 255
     },
     {
       "epoch": 0.7975460122699386,
-      "grad_norm": 2.1303631206592253,
+      "grad_norm": 2.1852216414908976,
       "learning_rate": 1.837371956026433e-05,
-      "loss": 1.2666,
+      "loss": 1.2681,
       "step": 260
     },
     {
       "epoch": 0.8128834355828221,
-      "grad_norm": 3.324267620281515,
+      "grad_norm": 2.4315513970788682,
       "learning_rate": 1.8274816900135842e-05,
-      "loss": 1.2796,
+      "loss": 1.2805,
       "step": 265
     },
     {
       "epoch": 0.8282208588957055,
-      "grad_norm": 2.0369143733640174,
+      "grad_norm": 1.9970132500864828,
       "learning_rate": 1.817327777949407e-05,
-      "loss": 1.2191,
+      "loss": 1.2204,
       "step": 270
     },
     {
       "epoch": 0.843558282208589,
-      "grad_norm": 2.268965846881571,
+      "grad_norm": 2.438536266127501,
       "learning_rate": 1.806913454997717e-05,
-      "loss": 1.2382,
+      "loss": 1.2388,
       "step": 275
     },
     {
       "epoch": 0.8588957055214724,
-      "grad_norm": 1.9534994546311044,
+      "grad_norm": 1.9096886848188332,
       "learning_rate": 1.7962420392925066e-05,
-      "loss": 1.3029,
+      "loss": 1.3021,
       "step": 280
     },
     {
       "epoch": 0.8742331288343558,
-      "grad_norm": 1.9407859425510332,
+      "grad_norm": 2.0120912351402875,
       "learning_rate": 1.785316930880745e-05,
-      "loss": 1.183,
+      "loss": 1.1835,
       "step": 285
     },
     {
       "epoch": 0.8895705521472392,
-      "grad_norm": 1.9539166093233689,
+      "grad_norm": 1.982796410249312,
       "learning_rate": 1.7741416106390828e-05,
-      "loss": 1.1725,
+      "loss": 1.1753,
       "step": 290
     },
     {
       "epoch": 0.9049079754601227,
-      "grad_norm": 2.2999481127748216,
+      "grad_norm": 2.4058710177193525,
       "learning_rate": 1.7627196391647982e-05,
-      "loss": 1.2114,
+      "loss": 1.2109,
       "step": 295
     },
     {
       "epoch": 0.9202453987730062,
-      "grad_norm": 1.9126415566238704,
+      "grad_norm": 1.9621072259264853,
       "learning_rate": 1.75105465564135e-05,
-      "loss": 1.2206,
+      "loss": 1.2202,
       "step": 300
     },
     {
       "epoch": 0.9355828220858896,
-      "grad_norm": 2.3510054432341887,
+      "grad_norm": 8.348303745211982,
       "learning_rate": 1.739150376678883e-05,
-      "loss": 1.301,
+      "loss": 1.299,
       "step": 305
     },
     {
       "epoch": 0.950920245398773,
-      "grad_norm": 2.5269459534002596,
+      "grad_norm": 2.188760659995286,
       "learning_rate": 1.727010595130074e-05,
-      "loss": 1.3914,
+      "loss": 1.3906,
       "step": 310
     },
     {
       "epoch": 0.9662576687116564,
-      "grad_norm": 2.045224623846914,
+      "grad_norm": 2.2138363998997828,
       "learning_rate": 1.714639178881678e-05,
-      "loss": 1.2722,
+      "loss": 1.2704,
       "step": 315
     },
     {
       "epoch": 0.9815950920245399,
-      "grad_norm": 2.0917960158205107,
+      "grad_norm": 7.207718966765089,
       "learning_rate": 1.7020400696221737e-05,
-      "loss": 1.2093,
+      "loss": 1.227,
       "step": 320
     },
     {
       "epoch": 0.9969325153374233,
-      "grad_norm": 2.017009094857716,
+      "grad_norm": 2.033313094726782,
       "learning_rate": 1.6892172815858896e-05,
-      "loss": 1.2658,
+      "loss": 1.265,
       "step": 325
     },
     {
       "epoch": 1.0,
-      "eval_loss": 1.2135429382324219,
-      "eval_runtime": 6.9064,
-      "eval_samples_per_second": 23.022,
-      "eval_steps_per_second": 5.792,
+      "eval_loss": 1.214716911315918,
+      "eval_runtime": 7.4506,
+      "eval_samples_per_second": 21.341,
+      "eval_steps_per_second": 5.369,
       "step": 326
     },
     {
       "epoch": 1.0122699386503067,
-      "grad_norm": 2.1076821373912433,
+      "grad_norm": 2.246404161643353,
       "learning_rate": 1.6761749002740195e-05,
-      "loss": 1.126,
+      "loss": 1.1231,
       "step": 330
     },
     {
       "epoch": 1.0276073619631902,
-      "grad_norm": 2.532948886360268,
+      "grad_norm": 2.6307191299518595,
       "learning_rate": 1.662917081152932e-05,
-      "loss": 1.0517,
+      "loss": 1.0474,
       "step": 335
     },
     {
       "epoch": 1.0429447852760736,
-      "grad_norm": 2.159362066124978,
+      "grad_norm": 2.2466529263098596,
       "learning_rate": 1.6494480483301836e-05,
-      "loss": 0.9849,
+      "loss": 0.9948,
       "step": 340
     },
     {
       "epoch": 1.058282208588957,
-      "grad_norm": 1.9634133830158156,
+      "grad_norm": 2.0590187620332308,
       "learning_rate": 1.635772093208669e-05,
-      "loss": 0.961,
+      "loss": 0.9675,
       "step": 345
     },
     {
       "epoch": 1.0736196319018405,
-      "grad_norm": 2.643597563770428,
+      "grad_norm": 2.5079641466397864,
       "learning_rate": 1.6218935731193223e-05,
-      "loss": 0.929,
+      "loss": 0.9397,
       "step": 350
     },
     {
       "epoch": 1.0889570552147239,
-      "grad_norm": 2.2336296848587986,
+      "grad_norm": 2.2015613521602053,
       "learning_rate": 1.6078169099328196e-05,
-      "loss": 0.9695,
+      "loss": 0.9576,
       "step": 355
     },
     {
       "epoch": 1.1042944785276074,
-      "grad_norm": 2.5120125840182803,
+      "grad_norm": 2.5678723465584303,
       "learning_rate": 1.5935465886507143e-05,
-      "loss": 1.1179,
+      "loss": 1.1066,
       "step": 360
     },
     {
       "epoch": 1.1196319018404908,
-      "grad_norm": 2.0896191655801735,
+      "grad_norm": 2.1152311130980976,
       "learning_rate": 1.579087155976459e-05,
-      "loss": 1.1294,
+      "loss": 1.1337,
       "step": 365
     },
     {
       "epoch": 1.1349693251533743,
-      "grad_norm": 2.325319374896529,
+      "grad_norm": 2.36694508137547,
       "learning_rate": 1.5644432188667695e-05,
-      "loss": 0.9826,
+      "loss": 1.0043,
       "step": 370
     },
     {
       "epoch": 1.1503067484662577,
-      "grad_norm": 2.1728488841214095,
+      "grad_norm": 2.257597203782575,
       "learning_rate": 1.5496194430637903e-05,
-      "loss": 0.9251,
+      "loss": 0.9548,
       "step": 375
     },
     {
       "epoch": 1.165644171779141,
-      "grad_norm": 2.424880159324192,
+      "grad_norm": 2.224173969408768,
       "learning_rate": 1.5346205516085305e-05,
-      "loss": 1.0463,
+      "loss": 1.0267,
       "step": 380
     },
     {
       "epoch": 1.1809815950920246,
-      "grad_norm": 3.026265515790728,
+      "grad_norm": 2.785039079502963,
       "learning_rate": 1.5194513233360439e-05,
-      "loss": 0.9217,
+      "loss": 0.9385,
       "step": 385
     },
     {
       "epoch": 1.196319018404908,
-      "grad_norm": 3.1617293471411654,
+      "grad_norm": 2.8341094388368715,
       "learning_rate": 1.504116591352832e-05,
-      "loss": 1.1437,
+      "loss": 1.1401,
       "step": 390
     },
     {
       "epoch": 1.2116564417177913,
-      "grad_norm": 2.2307237557720003,
+      "grad_norm": 2.5109029462747303,
       "learning_rate": 1.4886212414969551e-05,
-      "loss": 0.967,
+      "loss": 0.9811,
       "step": 395
     },
     {
       "epoch": 1.2269938650306749,
-      "grad_norm": 2.430911245376782,
+      "grad_norm": 2.364253368402221,
       "learning_rate": 1.4729702107813438e-05,
-      "loss": 0.999,
+      "loss": 0.9896,
       "step": 400
     },
     {
       "epoch": 1.2423312883435582,
-      "grad_norm": 2.1612231748678896,
+      "grad_norm": 2.833077039900493,
       "learning_rate": 1.4571684858208045e-05,
-      "loss": 1.004,
+      "loss": 1.0038,
       "step": 405
     },
     {
       "epoch": 1.2576687116564418,
-      "grad_norm": 2.6596296516845563,
+      "grad_norm": 2.986726012938249,
       "learning_rate": 1.4412211012432213e-05,
-      "loss": 1.0454,
+      "loss": 1.0369,
       "step": 410
     },
     {
       "epoch": 1.2730061349693251,
-      "grad_norm": 2.093932365865954,
+      "grad_norm": 2.19472166570541,
       "learning_rate": 1.4251331380854602e-05,
-      "loss": 1.0569,
+      "loss": 1.0644,
       "step": 415
     },
     {
       "epoch": 1.2883435582822087,
-      "grad_norm": 2.72153245614344,
+      "grad_norm": 2.539259944541222,
       "learning_rate": 1.408909722174487e-05,
-      "loss": 0.9593,
+      "loss": 0.9615,
       "step": 420
     },
     {
       "epoch": 1.303680981595092,
-      "grad_norm": 2.6462362091450515,
+      "grad_norm": 2.3203428811857667,
       "learning_rate": 1.3925560224942145e-05,
-      "loss": 0.9271,
+      "loss": 0.933,
       "step": 425
     },
     {
       "epoch": 1.3190184049079754,
-      "grad_norm": 2.267205865998314,
+      "grad_norm": 2.190402503264761,
       "learning_rate": 1.3760772495385998e-05,
-      "loss": 0.9582,
+      "loss": 0.9614,
       "step": 430
     },
     {
       "epoch": 1.334355828220859,
-      "grad_norm": 14.299415996684987,
+      "grad_norm": 2.4516332947183055,
       "learning_rate": 1.3594786536515154e-05,
-      "loss": 0.9659,
+      "loss": 0.9784,
       "step": 435
     },
     {
       "epoch": 1.3496932515337423,
-      "grad_norm": 2.8919289960885157,
+      "grad_norm": 2.1838918342014897,
       "learning_rate": 1.3427655233539227e-05,
-      "loss": 0.922,
+      "loss": 0.9477,
       "step": 440
     },
     {
       "epoch": 1.3650306748466257,
-      "grad_norm": 2.0424273794301344,
+      "grad_norm": 2.1974830315305867,
       "learning_rate": 1.3259431836588843e-05,
-      "loss": 0.9932,
+      "loss": 0.9802,
       "step": 445
     },
     {
       "epoch": 1.3803680981595092,
-      "grad_norm": 6.610636759407049,
+      "grad_norm": 2.337562201817739,
       "learning_rate": 1.3090169943749475e-05,
-      "loss": 0.9837,
+      "loss": 1.001,
       "step": 450
     },
     {
       "epoch": 1.3957055214723926,
-      "grad_norm": 2.2497868172019357,
+      "grad_norm": 2.657566030569729,
       "learning_rate": 1.2919923483984415e-05,
-      "loss": 0.9149,
+      "loss": 0.9072,
       "step": 455
     },
     {
       "epoch": 1.4110429447852761,
-      "grad_norm": 2.2842331908069125,
+      "grad_norm": 2.5556378147753787,
       "learning_rate": 1.2748746699952338e-05,
-      "loss": 1.0342,
+      "loss": 1.0344,
       "step": 460
     },
     {
       "epoch": 1.4263803680981595,
-      "grad_norm": 2.1102577421193085,
+      "grad_norm": 2.490197731186031,
       "learning_rate": 1.2576694130724905e-05,
-      "loss": 1.1568,
+      "loss": 1.1441,
       "step": 465
     },
     {
       "epoch": 1.441717791411043,
-      "grad_norm": 2.127882920435348,
+      "grad_norm": 2.456053324999455,
       "learning_rate": 1.2403820594409926e-05,
-      "loss": 0.9638,
+      "loss": 0.9665,
       "step": 470
     },
     {
       "epoch": 1.4570552147239264,
-      "grad_norm": 2.584359793762212,
+      "grad_norm": 2.7963958883475586,
       "learning_rate": 1.2230181170685636e-05,
-      "loss": 0.9673,
+      "loss": 0.9733,
       "step": 475
     },
     {
       "epoch": 1.4723926380368098,
-      "grad_norm": 2.264911154520169,
+      "grad_norm": 2.3556048604718955,
       "learning_rate": 1.2055831183251608e-05,
-      "loss": 1.0154,
+      "loss": 1.0198,
       "step": 480
     },
     {
       "epoch": 1.4877300613496933,
-      "grad_norm": 2.0087330179243335,
+      "grad_norm": 2.7492055859532365,
       "learning_rate": 1.1880826182201926e-05,
-      "loss": 0.8973,
+      "loss": 0.9002,
       "step": 485
     },
     {
       "epoch": 1.5030674846625767,
-      "grad_norm": 1.9321511470889676,
+      "grad_norm": 2.138357362754652,
       "learning_rate": 1.170522192632624e-05,
-      "loss": 0.9371,
+      "loss": 0.9411,
       "step": 490
     },
     {
       "epoch": 1.51840490797546,
-      "grad_norm": 2.121122226050571,
+      "grad_norm": 2.1341509025323644,
       "learning_rate": 1.1529074365344302e-05,
-      "loss": 1.033,
+      "loss": 1.0553,
       "step": 495
     },
     {
       "epoch": 1.5337423312883436,
-      "grad_norm": 2.029011447995027,
+      "grad_norm": 2.1458405128765627,
       "learning_rate": 1.1352439622079689e-05,
-      "loss": 0.9768,
+      "loss": 1.0077,
       "step": 500
     },
     {
       "epoch": 1.5490797546012272,
-      "grad_norm": 2.108744801443365,
+      "grad_norm": 2.1132360957639698,
       "learning_rate": 1.1175373974578378e-05,
-      "loss": 0.9089,
+      "loss": 0.9168,
       "step": 505
     },
     {
       "epoch": 1.5644171779141103,
-      "grad_norm": 2.2204587915399876,
+      "grad_norm": 2.5113223130286686,
       "learning_rate": 1.0997933838177828e-05,
-      "loss": 1.0449,
+      "loss": 1.0312,
       "step": 510
     },
     {
       "epoch": 1.5797546012269938,
-      "grad_norm": 2.1535325181232294,
+      "grad_norm": 2.0666811534447906,
       "learning_rate": 1.0820175747532373e-05,
-      "loss": 1.0172,
+      "loss": 1.0088,
       "step": 515
     },
     {
       "epoch": 1.5950920245398774,
-      "grad_norm": 21.819229878115912,
+      "grad_norm": 2.18910972259404,
       "learning_rate": 1.064215633860055e-05,
-      "loss": 0.9769,
+      "loss": 0.9566,
       "step": 520
     },
     {
       "epoch": 1.6104294478527608,
-      "grad_norm": 2.1296751297142897,
+      "grad_norm": 2.1335643657003764,
       "learning_rate": 1.0463932330600197e-05,
-      "loss": 0.9483,
+      "loss": 0.9537,
       "step": 525
     },
     {
       "epoch": 1.6257668711656441,
-      "grad_norm": 2.1616752600527134,
+      "grad_norm": 2.211695419294786,
       "learning_rate": 1.0285560507936962e-05,
-      "loss": 4.3216,
+      "loss": 0.9519,
       "step": 530
     },
     {
       "epoch": 1.6411042944785277,
-      "grad_norm": 3.2786842755698795,
+      "grad_norm": 1.9927545915886413,
       "learning_rate": 1.010709770211212e-05,
-      "loss": 1.1255,
+      "loss": 1.0909,
       "step": 535
     },
     {
       "epoch": 1.656441717791411,
-      "grad_norm": 2.4763448399092147,
+      "grad_norm": 2.1799108916751266,
       "learning_rate": 9.928600773615306e-06,
-      "loss": 1.0498,
+      "loss": 1.0668,
       "step": 540
     },
     {
       "epoch": 1.6717791411042944,
-      "grad_norm": 2.006782735936947,
+      "grad_norm": 2.0870390082509824,
       "learning_rate": 9.750126593808083e-06,
-      "loss": 0.927,
+      "loss": 0.9425,
       "step": 545
     },
     {
       "epoch": 1.687116564417178,
-      "grad_norm": 2.2437759465461777,
+      "grad_norm": 2.219919825754393,
       "learning_rate": 9.571732026803978e-06,
-      "loss": 0.906,
+      "loss": 0.9223,
       "step": 550
     },
     {
       "epoch": 1.7024539877300615,
-      "grad_norm": 1.9556898555009257,
+      "grad_norm": 2.0158507556929712,
       "learning_rate": 9.393473911350895e-06,
-      "loss": 0.9715,
+      "loss": 0.9779,
       "step": 555
     },
     {
       "epoch": 1.7177914110429446,
-      "grad_norm": 2.082737830638052,
+      "grad_norm": 2.146438648027479,
       "learning_rate": 9.215409042721553e-06,
-      "loss": 1.0,
+      "loss": 0.9929,
       "step": 560
     },
     {
       "epoch": 1.7331288343558282,
-      "grad_norm": 2.3756819460619534,
+      "grad_norm": 2.2820045196023115,
       "learning_rate": 9.037594154617811e-06,
-      "loss": 0.9755,
+      "loss": 0.9746,
       "step": 565
     },
     {
       "epoch": 1.7484662576687118,
-      "grad_norm": 2.7832428710692265,
+      "grad_norm": 2.1811945127821626,
       "learning_rate": 8.860085901094595e-06,
-      "loss": 0.9694,
+      "loss": 0.9463,
       "step": 570
     },
     {
       "epoch": 1.7638036809815951,
-      "grad_norm": 2.299830262573217,
+      "grad_norm": 2.2195282023793976,
       "learning_rate": 8.682940838509206e-06,
-      "loss": 0.8829,
+      "loss": 0.8636,
       "step": 575
     },
     {
       "epoch": 1.7791411042944785,
-      "grad_norm": 1.9164299833335463,
+      "grad_norm": 1.9502264600710417,
       "learning_rate": 8.50621540750175e-06,
-      "loss": 0.9634,
+      "loss": 0.9595,
       "step": 580
     },
     {
       "epoch": 1.794478527607362,
-      "grad_norm": 2.0803428687947023,
+      "grad_norm": 2.1028286450122784,
       "learning_rate": 8.329965915012451e-06,
-      "loss": 1.0047,
+      "loss": 1.0149,
       "step": 585
     },
     {
       "epoch": 1.8098159509202454,
-      "grad_norm": 1.961542308344144,
+      "grad_norm": 1.996134849337679,
       "learning_rate": 8.154248516341547e-06,
-      "loss": 0.9783,
+      "loss": 0.9714,
       "step": 590
     },
     {
       "epoch": 1.8251533742331287,
-      "grad_norm": 1.98951654932812,
+      "grad_norm": 2.0229355706203673,
       "learning_rate": 7.979119197257505e-06,
-      "loss": 1.0009,
+      "loss": 1.0042,
       "step": 595
     },
     {
       "epoch": 1.8404907975460123,
-      "grad_norm": 2.1181463512012653,
+      "grad_norm": 2.198065583918555,
       "learning_rate": 7.804633756159258e-06,
-      "loss": 0.9263,
+      "loss": 0.9354,
       "step": 600
     },
     {
       "epoch": 1.8558282208588959,
-      "grad_norm": 2.017971907316727,
+      "grad_norm": 2.0001193503472248,
       "learning_rate": 7.63084778629813e-06,
-      "loss": 0.9933,
+      "loss": 0.9796,
       "step": 605
     },
     {
       "epoch": 1.871165644171779,
-      "grad_norm": 2.1164163856970135,
+      "grad_norm": 2.150238538587084,
       "learning_rate": 7.4578166580651335e-06,
-      "loss": 0.9392,
+      "loss": 0.9495,
       "step": 610
     },
     {
       "epoch": 1.8865030674846626,
-      "grad_norm": 2.118557747327772,
+      "grad_norm": 2.1692276656591543,
       "learning_rate": 7.285595501349259e-06,
-      "loss": 0.9537,
+      "loss": 0.972,
       "step": 615
     },
     {
       "epoch": 1.9018404907975461,
-      "grad_norm": 2.087163676278093,
+      "grad_norm": 2.2030115157082073,
       "learning_rate": 7.114239187972416e-06,
-      "loss": 0.8722,
+      "loss": 0.8763,
       "step": 620
     },
     {
       "epoch": 1.9171779141104295,
-      "grad_norm": 2.222443842922782,
+      "grad_norm": 13.25110262390118,
       "learning_rate": 6.94380231420656e-06,
-      "loss": 0.9518,
+      "loss": 0.9911,
       "step": 625
     },
     {
       "epoch": 1.9325153374233128,
-      "grad_norm": 2.0519691059771503,
+      "grad_norm": 2.128565521517256,
       "learning_rate": 6.774339183378663e-06,
-      "loss": 0.9595,
+      "loss": 0.98,
       "step": 630
     },
     {
       "epoch": 1.9478527607361964,
-      "grad_norm": 2.395088570054642,
+      "grad_norm": 2.446455411382673,
       "learning_rate": 6.605903788568962e-06,
-      "loss": 0.9082,
+      "loss": 0.9163,
       "step": 635
     },
     {
       "epoch": 1.9631901840490797,
-      "grad_norm": 4.89604575465679,
+      "grad_norm": 3.6457472773842055,
       "learning_rate": 6.438549795408107e-06,
-      "loss": 0.9101,
+      "loss": 0.893,
       "step": 640
     },
     {
       "epoch": 1.978527607361963,
-      "grad_norm": 2.1171931357011236,
+      "grad_norm": 2.1918272282922615,
       "learning_rate": 6.272330524978613e-06,
-      "loss": 0.9687,
+      "loss": 0.9862,
       "step": 645
     },
     {
       "epoch": 1.9938650306748467,
-      "grad_norm": 2.0986919385384137,
+      "grad_norm": 2.173238130186426,
       "learning_rate": 6.107298936826086e-06,
-      "loss": 0.8841,
+      "loss": 0.8983,
       "step": 650
     },
     {
       "epoch": 2.0,
-      "eval_loss": 1.2134720087051392,
-      "eval_runtime": 7.1411,
-      "eval_samples_per_second": 22.266,
-      "eval_steps_per_second": 5.601,
+      "eval_loss": 1.211948275566101,
+      "eval_runtime": 6.699,
+      "eval_samples_per_second": 23.735,
+      "eval_steps_per_second": 5.971,
       "step": 652
     },
     {
       "epoch": 2.0092024539877302,
-      "grad_norm": 2.1628044565671556,
+      "grad_norm": 2.2524003786276254,
       "learning_rate": 5.943507612085661e-06,
-      "loss": 0.8251,
+      "loss": 0.8223,
       "step": 655
     },
     {
       "epoch": 2.0245398773006134,
-      "grad_norm": 1.980162767088004,
+      "grad_norm": 1.9974383389908232,
       "learning_rate": 5.781008736728975e-06,
-      "loss": 0.6487,
+      "loss": 0.6493,
       "step": 660
     },
     {
       "epoch": 2.039877300613497,
-      "grad_norm": 2.862228737237209,
+      "grad_norm": 2.9682960559686897,
       "learning_rate": 5.619854084937085e-06,
-      "loss": 0.6706,
+      "loss": 0.6784,
       "step": 665
     },
     {
       "epoch": 2.0552147239263805,
-      "grad_norm": 2.2439249118077136,
+      "grad_norm": 2.2655157887127024,
       "learning_rate": 5.460095002604533e-06,
-      "loss": 0.6652,
+      "loss": 0.6748,
       "step": 670
     },
     {
       "epoch": 2.0705521472392636,
-      "grad_norm": 2.254257809075137,
+      "grad_norm": 2.2085119385865863,
       "learning_rate": 5.3017823909799295e-06,
-      "loss": 0.7293,
+      "loss": 0.7027,
       "step": 675
     },
     {
       "epoch": 2.085889570552147,
-      "grad_norm": 2.223082549239095,
+      "grad_norm": 2.27766004579248,
       "learning_rate": 5.144966690448159e-06,
-      "loss": 0.755,
+      "loss": 0.7278,
       "step": 680
     },
     {
       "epoch": 2.1012269938650308,
-      "grad_norm": 2.0645657168776617,
+      "grad_norm": 2.128219709033872,
       "learning_rate": 4.9896978644594516e-06,
-      "loss": 0.5896,
+      "loss": 0.5989,
       "step": 685
     },
     {
       "epoch": 2.116564417177914,
-      "grad_norm": 2.543974903041173,
+      "grad_norm": 2.4873822979923315,
       "learning_rate": 4.836025383610382e-06,
-      "loss": 0.7457,
+      "loss": 0.7357,
       "step": 690
     },
     {
       "epoch": 2.1319018404907975,
-      "grad_norm": 2.1096988101915524,
+      "grad_norm": 2.1568784749495244,
       "learning_rate": 4.683998209881943e-06,
-      "loss": 0.6847,
+      "loss": 0.714,
       "step": 695
     },
     {
       "epoch": 2.147239263803681,
-      "grad_norm": 2.36099194700885,
+      "grad_norm": 2.3234411443951126,
       "learning_rate": 4.533664781039622e-06,
-      "loss": 0.715,
+      "loss": 0.7281,
       "step": 700
     },
     {
       "epoch": 2.1625766871165646,
-      "grad_norm": 2.633833128566785,
+      "grad_norm": 2.735689720360169,
       "learning_rate": 4.385072995200532e-06,
-      "loss": 0.626,
+      "loss": 0.6257,
       "step": 705
     },
     {
       "epoch": 2.1779141104294477,
-      "grad_norm": 2.506530086487105,
+      "grad_norm": 3.0839192153525934,
       "learning_rate": 4.2382701955724724e-06,
-      "loss": 0.7172,
+      "loss": 0.7162,
       "step": 710
     },
     {
       "epoch": 2.1932515337423313,
-      "grad_norm": 2.3802287660865553,
+      "grad_norm": 2.4589255584801624,
       "learning_rate": 4.093303155369771e-06,
-      "loss": 0.6839,
+      "loss": 0.6726,
       "step": 715
     },
     {
       "epoch": 2.208588957055215,
-      "grad_norm": 2.24220538972355,
+      "grad_norm": 2.3530595892180637,
       "learning_rate": 3.950218062910776e-06,
-      "loss": 0.5827,
+      "loss": 0.6077,
       "step": 720
     },
     {
       "epoch": 2.223926380368098,
-      "grad_norm": 2.2370918425887267,
+      "grad_norm": 2.2994180114078375,
       "learning_rate": 3.8090605069016596e-06,
-      "loss": 0.6599,
+      "loss": 0.6484,
       "step": 725
     },
     {
       "epoch": 2.2392638036809815,
-      "grad_norm": 2.1088652822907448,
+      "grad_norm": 2.1614299270488844,
       "learning_rate": 3.6698754619112974e-06,
-      "loss": 0.7594,
+      "loss": 0.7606,
       "step": 730
     },
     {
       "epoch": 2.254601226993865,
-      "grad_norm": 2.083592158309122,
+      "grad_norm": 2.148628125479035,
       "learning_rate": 3.53270727404179e-06,
-      "loss": 0.6702,
+      "loss": 0.6735,
       "step": 735
     },
     {
       "epoch": 2.2699386503067487,
-      "grad_norm": 2.567792989862267,
+      "grad_norm": 2.791988335379792,
       "learning_rate": 3.3975996467992557e-06,
-      "loss": 0.6419,
+      "loss": 0.6685,
       "step": 740
     },
     {
       "epoch": 2.285276073619632,
-      "grad_norm": 2.334913734729685,
+      "grad_norm": 2.4166794649978036,
       "learning_rate": 3.2645956271693257e-06,
-      "loss": 0.6634,
+      "loss": 0.6727,
       "step": 745
     },
     {
       "epoch": 2.3006134969325154,
-      "grad_norm": 2.3429523817504716,
+      "grad_norm": 2.270680126335755,
       "learning_rate": 3.133737591901864e-06,
-      "loss": 0.6939,
+      "loss": 0.6946,
       "step": 750
     },
     {
       "epoch": 2.315950920245399,
-      "grad_norm": 2.2619584916199615,
+      "grad_norm": 2.2773152491745043,
       "learning_rate": 3.0050672340091723e-06,
-      "loss": 0.6678,
+      "loss": 0.6436,
       "step": 755
     },
     {
       "epoch": 2.331288343558282,
-      "grad_norm": 2.6296479257995777,
+      "grad_norm": 2.351791962083414,
       "learning_rate": 2.878625549482084e-06,
-      "loss": 0.7031,
+      "loss": 0.671,
       "step": 760
     },
     {
       "epoch": 2.3466257668711656,
-      "grad_norm": 2.8130762384906864,
+      "grad_norm": 2.5602963650641604,
       "learning_rate": 2.7544528242281323e-06,
-      "loss": 0.5788,
+      "loss": 0.559,
       "step": 765
     },
     {
       "epoch": 2.361963190184049,
-      "grad_norm": 2.150686530040835,
+      "grad_norm": 2.919982636047529,
       "learning_rate": 2.6325886212359496e-06,
-      "loss": 0.6077,
+      "loss": 0.608,
       "step": 770
     },
     {
       "epoch": 2.3773006134969323,
-      "grad_norm": 2.2806678502354476,
+      "grad_norm": 2.283862507375789,
       "learning_rate": 2.51307176797001e-06,
-      "loss": 0.7225,
+      "loss": 0.7129,
       "step": 775
     },
     {
       "epoch": 2.392638036809816,
-      "grad_norm": 2.257367372279919,
+      "grad_norm": 2.5119018741262273,
       "learning_rate": 2.395940343999691e-06,
-      "loss": 0.66,
+      "loss": 0.6617,
       "step": 780
     },
     {
       "epoch": 2.4079754601226995,
-      "grad_norm": 2.105998185034246,
+      "grad_norm": 2.2326958137785997,
       "learning_rate": 2.2812316688666735e-06,
-      "loss": 0.5642,
+      "loss": 0.6001,
       "step": 785
     },
     {
       "epoch": 2.4233128834355826,
-      "grad_norm": 2.6751693267999164,
+      "grad_norm": 2.4717197828958875,
       "learning_rate": 2.1689822901944456e-06,
-      "loss": 0.6999,
+      "loss": 0.6849,
       "step": 790
     },
     {
       "epoch": 2.438650306748466,
-      "grad_norm": 2.337986334515246,
+      "grad_norm": 2.428100475473941,
       "learning_rate": 2.0592279720437856e-06,
-      "loss": 0.6405,
+      "loss": 0.6541,
       "step": 795
     },
     {
       "epoch": 2.4539877300613497,
-      "grad_norm": 2.5090481285811026,
+      "grad_norm": 2.453498212161667,
       "learning_rate": 1.9520036835178667e-06,
-      "loss": 0.7335,
+      "loss": 0.7362,
       "step": 800
     },
     {
       "epoch": 2.4693251533742333,
-      "grad_norm": 2.222070003754594,
+      "grad_norm": 2.202254537031373,
       "learning_rate": 1.8473435876206792e-06,
-      "loss": 0.6446,
+      "loss": 0.6467,
       "step": 805
     },
     {
       "epoch": 2.4846625766871164,
-      "grad_norm": 2.3534318368821467,
+      "grad_norm": 2.5704434258789557,
       "learning_rate": 1.74528103037226e-06,
-      "loss": 0.7604,
+      "loss": 0.7467,
       "step": 810
     },
     {
       "epoch": 2.5,
-      "grad_norm": 2.1957922076650225,
+      "grad_norm": 2.278793792368562,
       "learning_rate": 1.645848530184233e-06,
-      "loss": 0.6602,
+      "loss": 0.6641,
       "step": 815
     },
     {
       "epoch": 2.5153374233128836,
-      "grad_norm": 2.248469614521019,
+      "grad_norm": 2.3206459000474338,
       "learning_rate": 1.5490777674990376e-06,
-      "loss": 0.7518,
+      "loss": 0.7399,
       "step": 820
     },
     {
       "epoch": 2.530674846625767,
-      "grad_norm": 2.358970858496667,
+      "grad_norm": 2.2760200456954105,
       "learning_rate": 1.4549995746961332e-06,
-      "loss": 0.6974,
+      "loss": 0.6896,
       "step": 825
     },
     {
       "epoch": 2.5460122699386503,
-      "grad_norm": 2.409137169378918,
+      "grad_norm": 2.0966897996478626,
       "learning_rate": 1.3636439262684299e-06,
-      "loss": 0.6054,
+      "loss": 0.5792,
       "step": 830
     },
     {
       "epoch": 2.561349693251534,
-      "grad_norm": 2.3008344145258275,
+      "grad_norm": 2.273137273352939,
       "learning_rate": 1.2750399292720284e-06,
-      "loss": 0.6626,
+      "loss": 0.6727,
       "step": 835
     },
     {
       "epoch": 2.5766871165644174,
-      "grad_norm": 2.2498393839801403,
+      "grad_norm": 2.232573385750787,
       "learning_rate": 1.1892158140523546e-06,
-      "loss": 0.6923,
+      "loss": 0.7411,
       "step": 840
     },
     {
       "epoch": 2.5920245398773005,
-      "grad_norm": 2.287309751844316,
+      "grad_norm": 2.2477459570297427,
       "learning_rate": 1.1061989252496053e-06,
-      "loss": 0.6596,
+      "loss": 0.637,
       "step": 845
     },
     {
       "epoch": 2.607361963190184,
-      "grad_norm": 3.2190551280360937,
+      "grad_norm": 2.4503366494149152,
       "learning_rate": 1.0260157130864178e-06,
-      "loss": 0.6771,
+      "loss": 0.6591,
       "step": 850
     },
     {
       "epoch": 2.6226993865030677,
-      "grad_norm": 2.1948501519870782,
+      "grad_norm": 2.2727030330816707,
       "learning_rate": 9.486917249404815e-07,
-      "loss": 0.6886,
+      "loss": 0.6791,
       "step": 855
     },
     {
       "epoch": 2.638036809815951,
-      "grad_norm": 2.1803684507178955,
+      "grad_norm": 2.4331538301144695,
       "learning_rate": 8.742515972048404e-07,
-      "loss": 0.7121,
+      "loss": 0.689,
       "step": 860
     },
     {
       "epoch": 2.6533742331288344,
-      "grad_norm": 2.089857540368277,
+      "grad_norm": 2.171632535993997,
       "learning_rate": 8.027190474384127e-07,
-      "loss": 0.5654,
+      "loss": 0.588,
       "step": 865
     },
     {
       "epoch": 2.668711656441718,
-      "grad_norm": 2.900973015953986,
+      "grad_norm": 2.274712548175332,
       "learning_rate": 7.341168668092857e-07,
-      "loss": 0.5899,
+      "loss": 0.5984,
       "step": 870
     },
     {
       "epoch": 2.684049079754601,
-      "grad_norm": 2.269220220981122,
+      "grad_norm": 2.187602871418788,
       "learning_rate": 6.684669128331655e-07,
-      "loss": 0.7626,
+      "loss": 0.7899,
       "step": 875
     },
     {
       "epoch": 2.6993865030674846,
-      "grad_norm": 2.177246970471886,
+      "grad_norm": 2.172889231318168,
       "learning_rate": 6.057901024092949e-07,
-      "loss": 0.6272,
+      "loss": 0.6107,
       "step": 880
     },
     {
       "epoch": 2.714723926380368,
-      "grad_norm": 2.265583420009415,
+      "grad_norm": 5.181833794013852,
       "learning_rate": 5.461064051560705e-07,
-      "loss": 0.7539,
+      "loss": 0.747,
       "step": 885
     },
     {
       "epoch": 2.7300613496932513,
-      "grad_norm": 2.097337573364251,
+      "grad_norm": 2.1284344076378163,
       "learning_rate": 4.894348370484648e-07,
-      "loss": 0.6404,
+      "loss": 0.6449,
       "step": 890
     },
     {
       "epoch": 2.745398773006135,
-      "grad_norm": 2.6099811023622483,
+      "grad_norm": 2.742405951465162,
       "learning_rate": 4.3579345435930454e-07,
-      "loss": 0.781,
+      "loss": 0.7748,
       "step": 895
     },
     {
       "epoch": 2.7607361963190185,
-      "grad_norm": 2.2217027523814155,
+      "grad_norm": 2.453933407406173,
       "learning_rate": 3.851993479063154e-07,
-      "loss": 0.7015,
+      "loss": 0.7247,
       "step": 900
     },
     {
       "epoch": 2.7760736196319016,
-      "grad_norm": 2.325907671442105,
+      "grad_norm": 2.3349240625236307,
       "learning_rate": 3.3766863760676947e-07,
-      "loss": 0.6844,
+      "loss": 0.6742,
       "step": 905
     },
     {
       "epoch": 2.791411042944785,
-      "grad_norm": 2.264535197628352,
+      "grad_norm": 2.346238889067428,
       "learning_rate": 2.93216467341475e-07,
-      "loss": 0.6508,
+      "loss": 0.6736,
       "step": 910
     },
     {
       "epoch": 2.8067484662576687,
-      "grad_norm": 2.4713808636104533,
+      "grad_norm": 2.682770493121025,
       "learning_rate": 2.5185700012975603e-07,
-      "loss": 0.7762,
+      "loss": 0.7891,
       "step": 915
     },
     {
       "epoch": 2.8220858895705523,
-      "grad_norm": 2.402679896607673,
+      "grad_norm": 2.418319757792411,
       "learning_rate": 2.1360341361692517e-07,
-      "loss": 0.6639,
+      "loss": 0.6781,
       "step": 920
     },
     {
       "epoch": 2.837423312883436,
-      "grad_norm": 2.179790948612016,
+      "grad_norm": 2.1865263805130737,
       "learning_rate": 1.784678958757291e-07,
-      "loss": 0.65,
+      "loss": 0.6458,
       "step": 925
     },
     {
       "epoch": 2.852760736196319,
-      "grad_norm": 2.076797574324857,
+      "grad_norm": 2.15121097086734,
       "learning_rate": 1.464616415230702e-07,
-      "loss": 0.6342,
+      "loss": 0.639,
       "step": 930
     },
     {
       "epoch": 2.8680981595092025,
-      "grad_norm": 2.1047162156910804,
+      "grad_norm": 2.1733157849896547,
       "learning_rate": 1.1759484815326294e-07,
-      "loss": 0.6047,
+      "loss": 0.6231,
       "step": 935
     },
     {
       "epoch": 2.883435582822086,
-      "grad_norm": 2.1660216273789237,
+      "grad_norm": 2.2028794055351235,
       "learning_rate": 9.187671308895418e-08,
-      "loss": 0.6571,
+      "loss": 0.6718,
       "step": 940
     },
     {
       "epoch": 2.8987730061349692,
-      "grad_norm": 2.713364435220778,
+      "grad_norm": 2.7622530458915655,
       "learning_rate": 6.931543045073708e-08,
-      "loss": 0.8338,
+      "loss": 0.8227,
       "step": 945
     },
     {
       "epoch": 2.914110429447853,
-      "grad_norm": 2.6713007123159724,
+      "grad_norm": 2.1282608794755364,
       "learning_rate": 4.991818854640396e-08,
-      "loss": 0.6373,
+      "loss": 0.6131,
       "step": 950
     },
     {
       "epoch": 2.9294478527607364,
-      "grad_norm": 2.1069338794602723,
+      "grad_norm": 2.1900416244520295,
       "learning_rate": 3.369116758066171e-08,
-      "loss": 0.6385,
+      "loss": 0.6493,
       "step": 955
     },
     {
       "epoch": 2.9447852760736195,
-      "grad_norm": 2.312520932069723,
+      "grad_norm": 2.3718267533912125,
       "learning_rate": 2.063953768603799e-08,
-      "loss": 0.7596,
+      "loss": 0.7364,
       "step": 960
     },
     {
       "epoch": 2.960122699386503,
-      "grad_norm": 2.1046770032748077,
+      "grad_norm": 2.2239316758606975,
       "learning_rate": 1.0767457275615567e-08,
-      "loss": 0.634,
+      "loss": 0.6457,
       "step": 965
     },
     {
       "epoch": 2.9754601226993866,
-      "grad_norm": 2.207645312141143,
+      "grad_norm": 2.6083276517224374,
       "learning_rate": 4.0780717181077015e-09,
-      "loss": 0.7555,
+      "loss": 0.7664,
       "step": 970
     },
     {
       "epoch": 2.9907975460122698,
-      "grad_norm": 2.457929371194042,
+      "grad_norm": 2.554211412831243,
       "learning_rate": 5.735123357042405e-10,
-      "loss": 0.7012,
+      "loss": 0.7007,
       "step": 975
     },
     {
       "epoch": 3.0,
-      "eval_loss": 1.3543522357940674,
-      "eval_runtime": 6.5158,
-      "eval_samples_per_second": 24.402,
-      "eval_steps_per_second": 6.139,
+      "eval_loss": 1.3528565168380737,
+      "eval_runtime": 6.5488,
+      "eval_samples_per_second": 24.279,
+      "eval_steps_per_second": 6.108,
       "step": 978
     },
     {
       "epoch": 3.0,
       "step": 978,
       "total_flos": 3642602029056.0,
-      "train_loss": 1.0360665048314506,
-      "train_runtime": 586.176,
-      "train_samples_per_second": 6.674,
-      "train_steps_per_second": 1.668
+      "train_loss": 1.0182966589927673,
+      "train_runtime": 588.3296,
+      "train_samples_per_second": 6.649,
+      "train_steps_per_second": 1.662
     }
   ],
   "logging_steps": 5,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53624d092600031a396068be5c97cc3df1725958186d1ebe3846858ecc969c0e
-size 6968
+oid sha256:eebff96f179673b60d9f5f767d4d743d7be42e01e13c2259d4a1c04b186e03dc
+size 7224