Model save

README.md

@@ -2,15 +2,11 @@
 license: gemma
 base_model: google/gemma-2-9b
 tags:
-- alignment-handbook
-- trl
-- sft
-- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- HuggingFaceH4/ultrachat_200k
+- generator
 model-index:
 - name: zephyr-gemma-2-9b-sft
   results: []
@@ -19,12 +15,12 @@ model-index:
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/tanliboy/huggingface/runs/v8t9mu31)
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/tanliboy/huggingface/runs/lk9oo65j)
 # zephyr-gemma-2-9b-sft
 
-This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) on the HuggingFaceH4/ultrachat_200k dataset.
+This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0638
+- Loss: 1.0639
 
 ## Model description
 
@@ -61,12 +57,12 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 1.096         | 0.9995 | 951  | 1.0638          |
+| 1.0962        | 0.9995 | 951  | 1.0639          |
 
 
 ### Framework versions
 
-- Transformers 4.42.4
+- Transformers 4.43.0.dev0
 - Pytorch 2.3.1+cu121
 - Datasets 2.19.1
 - Tokenizers 0.19.1

all_results.json

@@ -1,14 +1,9 @@
 {
     "epoch": 0.999474513925381,
-    "eval_loss": 1.0638214349746704,
-    "eval_runtime": 592.7206,
-    "eval_samples": 23109,
-    "eval_samples_per_second": 22.734,
-    "eval_steps_per_second": 0.712,
     "total_flos": 905758069751808.0,
-    "train_loss": 1.1023603343061092,
-    "train_runtime": 21224.38,
+    "train_loss": 1.102385032064155,
+    "train_runtime": 21864.5552,
     "train_samples": 207864,
-    "train_samples_per_second": 5.738,
-    "train_steps_per_second": 0.045
+    "train_samples_per_second": 5.57,
+    "train_steps_per_second": 0.043
 }

generation_config.json

@@ -4,5 +4,5 @@
   "cache_implementation": "hybrid",
   "eos_token_id": 1,
   "pad_token_id": 0,
-  "transformers_version": "4.42.4"
+  "transformers_version": "4.43.0.dev0"
 }

runs/Jul19_14-58-09_action-graph-trainer/events.out.tfevents.1721401122.action-graph-trainer.704471.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:726b9ae38b843a9e75a36bc294f4c0b0b9927e997a375271d2957174f34b6c0d
-size 45931
+oid sha256:899e651843faea4fc5e62ec965e216057a695c914c9f61bddbdcc8c09817fa44
+size 46556

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 0.999474513925381,
     "total_flos": 905758069751808.0,
-    "train_loss": 1.1023603343061092,
-    "train_runtime": 21224.38,
+    "train_loss": 1.102385032064155,
+    "train_runtime": 21864.5552,
     "train_samples": 207864,
-    "train_samples_per_second": 5.738,
-    "train_steps_per_second": 0.045
+    "train_samples_per_second": 5.57,
+    "train_steps_per_second": 0.043
 }

trainer_state.json

@@ -150,1217 +150,1217 @@
     },
     {
       "epoch": 0.10509721492380451,
-      "grad_norm": 0.9462854605403194,
+      "grad_norm": 1.0117170886130586,
       "learning_rate": 2.9998379903275155e-06,
       "loss": 1.1108,
       "step": 100
     },
     {
       "epoch": 0.11035207566999475,
-      "grad_norm": 0.9678742768022235,
+      "grad_norm": 0.906919049524167,
       "learning_rate": 2.9991798860113893e-06,
       "loss": 1.1162,
       "step": 105
     },
     {
       "epoch": 0.11560693641618497,
-      "grad_norm": 0.9200494964131395,
+      "grad_norm": 0.9141362173846742,
       "learning_rate": 2.998015783397426e-06,
       "loss": 1.1091,
       "step": 110
     },
     {
       "epoch": 0.1208617971623752,
-      "grad_norm": 0.9581044551729833,
+      "grad_norm": 0.9619396255470063,
       "learning_rate": 2.9963460753897363e-06,
-      "loss": 1.0963,
+      "loss": 1.0962,
       "step": 115
     },
     {
       "epoch": 0.1261166579085654,
-      "grad_norm": 0.911911983689549,
+      "grad_norm": 1.0465936968149612,
       "learning_rate": 2.994171325542714e-06,
       "loss": 1.0911,
       "step": 120
     },
     {
       "epoch": 0.13137151865475566,
-      "grad_norm": 0.8931755968055785,
+      "grad_norm": 0.9530685775511196,
       "learning_rate": 2.991492267870822e-06,
       "loss": 1.0917,
       "step": 125
     },
     {
       "epoch": 0.13662637940094588,
-      "grad_norm": 0.882022590090018,
+      "grad_norm": 0.9008971542587714,
       "learning_rate": 2.9883098066008556e-06,
-      "loss": 1.088,
+      "loss": 1.0879,
       "step": 130
     },
     {
       "epoch": 0.1418812401471361,
-      "grad_norm": 0.9953724063836157,
+      "grad_norm": 1.010852514785442,
       "learning_rate": 2.984625015866745e-06,
-      "loss": 1.0859,
+      "loss": 1.086,
       "step": 135
     },
     {
       "epoch": 0.14713610089332632,
-      "grad_norm": 1.878615978166485,
+      "grad_norm": 0.8992929518552587,
       "learning_rate": 2.9804391393470235e-06,
-      "loss": 1.1006,
+      "loss": 1.1009,
       "step": 140
     },
     {
       "epoch": 0.15239096163951654,
-      "grad_norm": 0.970035528231665,
+      "grad_norm": 0.9062761866531701,
       "learning_rate": 2.975753589845059e-06,
-      "loss": 1.1022,
+      "loss": 1.1019,
       "step": 145
     },
     {
       "epoch": 0.1576458223857068,
-      "grad_norm": 0.9147530382000614,
+      "grad_norm": 0.8610097516319934,
       "learning_rate": 2.970569948812214e-06,
       "loss": 1.0691,
       "step": 150
     },
     {
       "epoch": 0.162900683131897,
-      "grad_norm": 0.9374168562620145,
+      "grad_norm": 0.8872085969729788,
       "learning_rate": 2.9648899658140767e-06,
-      "loss": 1.0861,
+      "loss": 1.0862,
       "step": 155
     },
     {
       "epoch": 0.16815554387808723,
-      "grad_norm": 0.8832219863831399,
+      "grad_norm": 1.0208916358421982,
       "learning_rate": 2.9587155579399543e-06,
       "loss": 1.0823,
       "step": 160
     },
     {
       "epoch": 0.17341040462427745,
-      "grad_norm": 0.9316518026448157,
+      "grad_norm": 1.000083341373154,
       "learning_rate": 2.9520488091558225e-06,
-      "loss": 1.0772,
+      "loss": 1.0771,
       "step": 165
     },
     {
       "epoch": 0.17866526537046767,
-      "grad_norm": 0.9096920364451422,
+      "grad_norm": 0.9057099612333195,
       "learning_rate": 2.944891969600953e-06,
       "loss": 1.0875,
       "step": 170
     },
     {
       "epoch": 0.1839201261166579,
-      "grad_norm": 0.9007871965996922,
+      "grad_norm": 0.9155112430806412,
       "learning_rate": 2.9372474548284537e-06,
-      "loss": 1.0949,
+      "loss": 1.095,
       "step": 175
     },
     {
       "epoch": 0.18917498686284814,
-      "grad_norm": 0.876698627967954,
+      "grad_norm": 0.8983015781319096,
       "learning_rate": 2.9291178449899786e-06,
       "loss": 1.0934,
       "step": 180
     },
     {
       "epoch": 0.19442984760903836,
-      "grad_norm": 0.9059798194790952,
+      "grad_norm": 0.9013247242472799,
       "learning_rate": 2.920505883964884e-06,
-      "loss": 1.0918,
+      "loss": 1.0917,
       "step": 185
     },
     {
       "epoch": 0.19968470835522859,
-      "grad_norm": 0.8939533236216217,
+      "grad_norm": 0.892018911723705,
       "learning_rate": 2.9114144784341226e-06,
       "loss": 1.0854,
       "step": 190
     },
     {
       "epoch": 0.2049395691014188,
-      "grad_norm": 0.9262446830111234,
+      "grad_norm": 0.9231553972162896,
       "learning_rate": 2.9018466968991914e-06,
       "loss": 1.0749,
       "step": 195
     },
     {
       "epoch": 0.21019442984760903,
-      "grad_norm": 0.8549091150362632,
+      "grad_norm": 0.8456793004463601,
       "learning_rate": 2.8918057686464587e-06,
-      "loss": 1.0825,
+      "loss": 1.0823,
       "step": 200
     },
     {
       "epoch": 0.21544929059379928,
-      "grad_norm": 0.8724281083548375,
+      "grad_norm": 0.8761446392793291,
       "learning_rate": 2.881295082657229e-06,
       "loss": 1.0769,
       "step": 205
     },
     {
       "epoch": 0.2207041513399895,
-      "grad_norm": 0.8895456534687597,
+      "grad_norm": 0.8860527749104554,
       "learning_rate": 2.8703181864639013e-06,
-      "loss": 1.0926,
+      "loss": 1.0925,
       "step": 210
     },
     {
       "epoch": 0.22595901208617972,
-      "grad_norm": 0.9539096387018893,
+      "grad_norm": 0.8708280345626856,
       "learning_rate": 2.8588787849526228e-06,
-      "loss": 1.0733,
+      "loss": 1.0731,
       "step": 215
     },
     {
       "epoch": 0.23121387283236994,
-      "grad_norm": 0.8703158811213173,
+      "grad_norm": 0.8688347790794807,
       "learning_rate": 2.846980739112822e-06,
       "loss": 1.0749,
       "step": 220
     },
     {
       "epoch": 0.23646873357856016,
-      "grad_norm": 0.8785853904575431,
+      "grad_norm": 0.8729839981153089,
       "learning_rate": 2.834628064734065e-06,
-      "loss": 1.0698,
+      "loss": 1.0697,
       "step": 225
     },
     {
       "epoch": 0.2417235943247504,
-      "grad_norm": 0.889877361216805,
+      "grad_norm": 0.8993877107760438,
       "learning_rate": 2.821824931050655e-06,
       "loss": 1.0751,
       "step": 230
     },
     {
       "epoch": 0.24697845507094063,
-      "grad_norm": 0.8959992372183687,
+      "grad_norm": 0.8452099155218541,
       "learning_rate": 2.8085756593344505e-06,
       "loss": 1.07,
       "step": 235
     },
     {
       "epoch": 0.2522333158171308,
-      "grad_norm": 0.8657778595116195,
+      "grad_norm": 0.8798731633285007,
       "learning_rate": 2.794884721436361e-06,
-      "loss": 1.0653,
+      "loss": 1.0655,
       "step": 240
     },
     {
       "epoch": 0.25748817656332107,
-      "grad_norm": 0.8631791797614975,
+      "grad_norm": 0.895358087891481,
       "learning_rate": 2.780756738277021e-06,
-      "loss": 1.0979,
+      "loss": 1.0983,
       "step": 245
     },
     {
       "epoch": 0.2627430373095113,
-      "grad_norm": 0.9028283942047837,
+      "grad_norm": 0.8609931821431536,
       "learning_rate": 2.766196478287156e-06,
       "loss": 1.0907,
       "step": 250
     },
     {
       "epoch": 0.2679978980557015,
-      "grad_norm": 1.105737597863472,
+      "grad_norm": 0.8553591777726404,
       "learning_rate": 2.751208855798155e-06,
       "loss": 1.0801,
       "step": 255
     },
     {
       "epoch": 0.27325275880189176,
-      "grad_norm": 0.9224613563982136,
+      "grad_norm": 0.8654715123002859,
       "learning_rate": 2.7357989293834005e-06,
-      "loss": 1.0821,
+      "loss": 1.082,
       "step": 260
     },
     {
       "epoch": 0.27850761954808195,
-      "grad_norm": 0.8624741728874925,
+      "grad_norm": 0.8697051687346281,
       "learning_rate": 2.7199719001509175e-06,
-      "loss": 1.075,
+      "loss": 1.0748,
       "step": 265
     },
     {
       "epoch": 0.2837624802942722,
-      "grad_norm": 0.8715022403371375,
+      "grad_norm": 0.880312866484559,
       "learning_rate": 2.7037331099879117e-06,
-      "loss": 1.088,
+      "loss": 1.0878,
       "step": 270
     },
     {
       "epoch": 0.28901734104046245,
-      "grad_norm": 0.8409580987018042,
+      "grad_norm": 0.8452698137198734,
       "learning_rate": 2.687088039757792e-06,
       "loss": 1.0797,
       "step": 275
     },
     {
       "epoch": 0.29427220178665264,
-      "grad_norm": 0.9670308072371985,
+      "grad_norm": 0.9791182527762319,
       "learning_rate": 2.6700423074502888e-06,
       "loss": 1.0717,
       "step": 280
     },
     {
       "epoch": 0.2995270625328429,
-      "grad_norm": 0.8449151190647378,
+      "grad_norm": 0.8645394739775393,
       "learning_rate": 2.652601666285289e-06,
-      "loss": 1.1141,
+      "loss": 1.114,
       "step": 285
     },
     {
       "epoch": 0.3047819232790331,
-      "grad_norm": 0.911604937769195,
+      "grad_norm": 0.9004251617653912,
       "learning_rate": 2.6347720027710253e-06,
       "loss": 1.067,
       "step": 290
     },
     {
       "epoch": 0.31003678402522333,
-      "grad_norm": 0.8655301197268408,
+      "grad_norm": 0.872604926778432,
       "learning_rate": 2.6165593347172837e-06,
-      "loss": 1.0731,
+      "loss": 1.0732,
       "step": 295
     },
     {
       "epoch": 0.3152916447714136,
-      "grad_norm": 0.880039459937268,
+      "grad_norm": 0.8929460807825601,
       "learning_rate": 2.5979698092042925e-06,
-      "loss": 1.0877,
+      "loss": 1.0876,
       "step": 300
     },
     {
       "epoch": 0.3205465055176038,
-      "grad_norm": 0.9360576412818693,
+      "grad_norm": 0.8588249540629153,
       "learning_rate": 2.5790097005079765e-06,
-      "loss": 1.0899,
+      "loss": 1.0897,
       "step": 305
     },
     {
       "epoch": 0.325801366263794,
-      "grad_norm": 0.8489030162852722,
+      "grad_norm": 0.8727624932776766,
       "learning_rate": 2.559685407982288e-06,
-      "loss": 1.0667,
+      "loss": 1.0666,
       "step": 310
     },
     {
       "epoch": 0.3310562270099842,
-      "grad_norm": 0.8638592230026647,
+      "grad_norm": 0.933192018274695,
       "learning_rate": 2.5400034538993135e-06,
-      "loss": 1.0867,
+      "loss": 1.0868,
       "step": 315
     },
     {
       "epoch": 0.33631108775617446,
-      "grad_norm": 0.9101865357911341,
+      "grad_norm": 0.9302912982362859,
       "learning_rate": 2.519970481247901e-06,
-      "loss": 1.0521,
+      "loss": 1.0518,
       "step": 320
     },
     {
       "epoch": 0.3415659485023647,
-      "grad_norm": 0.8626549506195685,
+      "grad_norm": 0.8713971326878898,
       "learning_rate": 2.4995932514915404e-06,
-      "loss": 1.0684,
+      "loss": 1.0683,
       "step": 325
     },
     {
       "epoch": 0.3468208092485549,
-      "grad_norm": 0.8534791671434315,
+      "grad_norm": 0.8822615502295114,
       "learning_rate": 2.478878642286253e-06,
       "loss": 1.0644,
       "step": 330
     },
     {
       "epoch": 0.35207566999474516,
-      "grad_norm": 0.8353900580922416,
+      "grad_norm": 0.8368326296787831,
       "learning_rate": 2.4578336451592705e-06,
-      "loss": 1.0813,
+      "loss": 1.0811,
       "step": 335
     },
     {
       "epoch": 0.35733053074093535,
-      "grad_norm": 0.8518466429904108,
+      "grad_norm": 0.8544515279738967,
       "learning_rate": 2.4364653631492774e-06,
-      "loss": 1.0626,
+      "loss": 1.0623,
       "step": 340
     },
     {
       "epoch": 0.3625853914871256,
-      "grad_norm": 0.8341915866164423,
+      "grad_norm": 0.891511057734901,
       "learning_rate": 2.414781008409014e-06,
       "loss": 1.0737,
       "step": 345
     },
     {
       "epoch": 0.3678402522333158,
-      "grad_norm": 0.8723412570499911,
+      "grad_norm": 0.9052547289541356,
       "learning_rate": 2.3927878997710575e-06,
       "loss": 1.0981,
       "step": 350
     },
     {
       "epoch": 0.37309511297950604,
-      "grad_norm": 0.841906976458512,
+      "grad_norm": 0.8951487470408807,
       "learning_rate": 2.3704934602775926e-06,
       "loss": 1.0827,
       "step": 355
     },
     {
       "epoch": 0.3783499737256963,
-      "grad_norm": 0.879119551017937,
+      "grad_norm": 0.9097434855008588,
       "learning_rate": 2.347905214675008e-06,
       "loss": 1.0713,
       "step": 360
     },
     {
       "epoch": 0.3836048344718865,
-      "grad_norm": 0.8718882110600734,
+      "grad_norm": 0.8583189613291908,
       "learning_rate": 2.3250307868741717e-06,
-      "loss": 1.0707,
+      "loss": 1.0706,
       "step": 365
     },
     {
       "epoch": 0.38885969521807673,
-      "grad_norm": 0.8002367155481277,
+      "grad_norm": 0.8216502318444575,
       "learning_rate": 2.3018778973772334e-06,
       "loss": 1.0573,
       "step": 370
     },
     {
       "epoch": 0.3941145559642669,
-      "grad_norm": 0.8221977480285654,
+      "grad_norm": 1.0166338065086127,
       "learning_rate": 2.278454360671823e-06,
-      "loss": 1.0867,
+      "loss": 1.0865,
       "step": 375
     },
     {
       "epoch": 0.39936941671045717,
-      "grad_norm": 0.8870576322337221,
+      "grad_norm": 0.8685946998953423,
       "learning_rate": 2.2547680825935325e-06,
-      "loss": 1.0851,
+      "loss": 1.0852,
       "step": 380
     },
     {
       "epoch": 0.4046242774566474,
-      "grad_norm": 0.8153258721073422,
+      "grad_norm": 0.8206385124004496,
       "learning_rate": 2.2308270576575657e-06,
       "loss": 1.0683,
       "step": 385
     },
     {
       "epoch": 0.4098791382028376,
-      "grad_norm": 0.8299248390745442,
+      "grad_norm": 0.8627531067721118,
       "learning_rate": 2.206639366360451e-06,
-      "loss": 1.0883,
+      "loss": 1.0885,
       "step": 390
     },
     {
       "epoch": 0.41513399894902786,
-      "grad_norm": 0.8891580311632679,
+      "grad_norm": 0.8850142996528242,
       "learning_rate": 2.1822131724527425e-06,
-      "loss": 1.0587,
+      "loss": 1.0588,
       "step": 395
     },
     {
       "epoch": 0.42038885969521805,
-      "grad_norm": 0.830348041290943,
+      "grad_norm": 0.8294172803163861,
       "learning_rate": 2.157556720183616e-06,
       "loss": 1.0542,
       "step": 400
     },
     {
       "epoch": 0.4256437204414083,
-      "grad_norm": 0.8254218651707308,
+      "grad_norm": 0.8307870053309333,
       "learning_rate": 2.1326783315182984e-06,
       "loss": 1.0666,
       "step": 405
     },
     {
       "epoch": 0.43089858118759855,
-      "grad_norm": 0.8211111371978261,
+      "grad_norm": 0.8178946165962943,
       "learning_rate": 2.1075864033292623e-06,
-      "loss": 1.0723,
+      "loss": 1.0725,
       "step": 410
     },
     {
       "epoch": 0.43615344193378874,
-      "grad_norm": 0.8308235610863787,
+      "grad_norm": 0.8236704022077929,
       "learning_rate": 2.082289404562144e-06,
-      "loss": 1.0767,
+      "loss": 1.0769,
       "step": 415
     },
     {
       "epoch": 0.441408302679979,
-      "grad_norm": 0.8127153392787219,
+      "grad_norm": 0.8321193181865734,
       "learning_rate": 2.0567958733773313e-06,
-      "loss": 1.0614,
+      "loss": 1.0613,
       "step": 420
     },
     {
       "epoch": 0.4466631634261692,
-      "grad_norm": 0.8123707093337108,
+      "grad_norm": 0.8122682546804041,
       "learning_rate": 2.0311144142681904e-06,
-      "loss": 1.064,
+      "loss": 1.0642,
       "step": 425
     },
     {
       "epoch": 0.45191802417235943,
-      "grad_norm": 0.829943736431802,
+      "grad_norm": 0.8236878035013919,
       "learning_rate": 2.005253695156909e-06,
       "loss": 1.0472,
       "step": 430
     },
     {
       "epoch": 0.4571728849185497,
-      "grad_norm": 0.831220074995737,
+      "grad_norm": 0.8493608509140299,
       "learning_rate": 1.9792224444689222e-06,
-      "loss": 1.0615,
+      "loss": 1.0616,
       "step": 435
     },
     {
       "epoch": 0.4624277456647399,
-      "grad_norm": 0.8745913548912436,
+      "grad_norm": 0.842399913675415,
       "learning_rate": 1.9530294481869286e-06,
-      "loss": 1.0802,
+      "loss": 1.0803,
       "step": 440
     },
     {
       "epoch": 0.4676826064109301,
-      "grad_norm": 0.8449818880595231,
+      "grad_norm": 0.8657084634018396,
       "learning_rate": 1.926683546885469e-06,
       "loss": 1.0588,
       "step": 445
     },
     {
       "epoch": 0.4729374671571203,
-      "grad_norm": 0.8068996721164886,
+      "grad_norm": 0.8242451456404312,
       "learning_rate": 1.9001936327470894e-06,
-      "loss": 1.0708,
+      "loss": 1.0709,
       "step": 450
     },
     {
       "epoch": 0.47819232790331057,
-      "grad_norm": 0.824592533003207,
+      "grad_norm": 0.8243645563523258,
       "learning_rate": 1.873568646561075e-06,
-      "loss": 1.0672,
+      "loss": 1.0674,
       "step": 455
     },
     {
       "epoch": 0.4834471886495008,
-      "grad_norm": 0.83022160250907,
+      "grad_norm": 0.8343407072556835,
       "learning_rate": 1.8468175747057898e-06,
       "loss": 1.0748,
       "step": 460
     },
     {
       "epoch": 0.488702049395691,
-      "grad_norm": 0.8333740262644744,
+      "grad_norm": 0.8704271811597876,
       "learning_rate": 1.8199494461156204e-06,
-      "loss": 1.0532,
+      "loss": 1.0533,
       "step": 465
     },
     {
       "epoch": 0.49395691014188126,
-      "grad_norm": 0.8922005131672666,
+      "grad_norm": 0.869601199009106,
       "learning_rate": 1.7929733292335591e-06,
       "loss": 1.0733,
       "step": 470
     },
     {
       "epoch": 0.49921177088807145,
-      "grad_norm": 0.9005753696617086,
+      "grad_norm": 0.8251688330803062,
       "learning_rate": 1.765898328950455e-06,
-      "loss": 1.0647,
+      "loss": 1.0651,
       "step": 475
     },
     {
       "epoch": 0.5044666316342616,
-      "grad_norm": 0.8247436750347265,
+      "grad_norm": 0.8182570293593837,
       "learning_rate": 1.738733583531959e-06,
       "loss": 1.0802,
       "step": 480
     },
     {
       "epoch": 0.509721492380452,
-      "grad_norm": 0.841670784750436,
+      "grad_norm": 0.861721166977276,
       "learning_rate": 1.7114882615342073e-06,
       "loss": 1.056,
       "step": 485
     },
     {
       "epoch": 0.5149763531266421,
-      "grad_norm": 0.8182688706522747,
+      "grad_norm": 0.8025950431569188,
       "learning_rate": 1.6841715587092798e-06,
-      "loss": 1.0783,
+      "loss": 1.0785,
       "step": 490
     },
     {
       "epoch": 0.5202312138728323,
-      "grad_norm": 1.239386954886359,
+      "grad_norm": 0.8462319262578287,
       "learning_rate": 1.6567926949014804e-06,
       "loss": 1.0745,
       "step": 495
     },
     {
       "epoch": 0.5254860746190226,
-      "grad_norm": 0.8030922261487655,
+      "grad_norm": 0.8532567164974908,
       "learning_rate": 1.6293609109354836e-06,
-      "loss": 1.0612,
+      "loss": 1.0611,
       "step": 500
     },
     {
       "epoch": 0.5307409353652128,
-      "grad_norm": 0.8189314991097896,
+      "grad_norm": 0.8238719502721198,
       "learning_rate": 1.601885465497404e-06,
-      "loss": 1.065,
+      "loss": 1.0651,
       "step": 505
     },
     {
       "epoch": 0.535995796111403,
-      "grad_norm": 0.9050946114254789,
+      "grad_norm": 0.8290145975340624,
       "learning_rate": 1.5743756320098334e-06,
       "loss": 1.0643,
       "step": 510
     },
     {
       "epoch": 0.5412506568575933,
-      "grad_norm": 0.8807760866261877,
+      "grad_norm": 0.814187446370962,
       "learning_rate": 1.5468406955019059e-06,
-      "loss": 1.0569,
+      "loss": 1.0571,
       "step": 515
     },
     {
       "epoch": 0.5465055176037835,
-      "grad_norm": 0.8569862964159628,
+      "grad_norm": 0.8640835720468265,
       "learning_rate": 1.5192899494754443e-06,
       "loss": 1.0731,
       "step": 520
     },
     {
       "epoch": 0.5517603783499737,
-      "grad_norm": 0.8091790353711866,
+      "grad_norm": 0.8096973874507514,
       "learning_rate": 1.4917326927682494e-06,
-      "loss": 1.0703,
+      "loss": 1.0705,
       "step": 525
     },
     {
       "epoch": 0.5570152390961639,
-      "grad_norm": 0.822047948500721,
+      "grad_norm": 0.8060488155757439,
       "learning_rate": 1.4641782264155852e-06,
-      "loss": 1.0737,
+      "loss": 1.0738,
       "step": 530
     },
     {
       "epoch": 0.5622700998423542,
-      "grad_norm": 0.8252210935515614,
+      "grad_norm": 0.8272465004554679,
       "learning_rate": 1.4366358505109237e-06,
       "loss": 1.0857,
       "step": 535
     },
     {
       "epoch": 0.5675249605885444,
-      "grad_norm": 0.8152512103821862,
+      "grad_norm": 0.808405985194556,
       "learning_rate": 1.4091148610670098e-06,
-      "loss": 1.0697,
+      "loss": 1.0699,
       "step": 540
     },
     {
       "epoch": 0.5727798213347346,
-      "grad_norm": 0.8348760975293736,
+      "grad_norm": 0.8401268224107731,
       "learning_rate": 1.3816245468782988e-06,
-      "loss": 1.0598,
+      "loss": 1.0601,
       "step": 545
     },
     {
       "epoch": 0.5780346820809249,
-      "grad_norm": 0.8008142952039858,
+      "grad_norm": 0.8212169210412936,
       "learning_rate": 1.3541741863858352e-06,
-      "loss": 1.0532,
+      "loss": 1.0534,
       "step": 550
     },
     {
       "epoch": 0.5832895428271151,
-      "grad_norm": 0.8525665050865836,
+      "grad_norm": 0.909077361069008,
       "learning_rate": 1.326773044545621e-06,
       "loss": 1.0672,
       "step": 555
     },
     {
       "epoch": 0.5885444035733053,
-      "grad_norm": 0.828489906999064,
+      "grad_norm": 0.8144193123335469,
       "learning_rate": 1.299430369701541e-06,
-      "loss": 1.0578,
+      "loss": 1.058,
       "step": 560
     },
     {
       "epoch": 0.5937992643194955,
-      "grad_norm": 0.8071915455528251,
+      "grad_norm": 0.8024091331374241,
       "learning_rate": 1.272155390463889e-06,
       "loss": 1.084,
       "step": 565
     },
     {
       "epoch": 0.5990541250656858,
-      "grad_norm": 0.8213168186088048,
+      "grad_norm": 0.8132457228764496,
       "learning_rate": 1.2449573125945607e-06,
-      "loss": 1.0678,
+      "loss": 1.0677,
       "step": 570
     },
     {
       "epoch": 0.604308985811876,
-      "grad_norm": 0.808856584314896,
+      "grad_norm": 0.8142277094157844,
       "learning_rate": 1.2178453158999509e-06,
-      "loss": 1.0705,
+      "loss": 1.0706,
       "step": 575
     },
     {
       "epoch": 0.6095638465580662,
-      "grad_norm": 0.8097966068652359,
+      "grad_norm": 0.7999185759745931,
       "learning_rate": 1.1908285511326195e-06,
-      "loss": 1.0561,
+      "loss": 1.056,
       "step": 580
     },
     {
       "epoch": 0.6148187073042565,
-      "grad_norm": 0.8103079561814972,
+      "grad_norm": 0.8042827578821733,
       "learning_rate": 1.1639161369027564e-06,
-      "loss": 1.0543,
+      "loss": 1.0544,
       "step": 585
     },
     {
       "epoch": 0.6200735680504467,
-      "grad_norm": 0.7934957921328578,
+      "grad_norm": 0.7974533954875164,
       "learning_rate": 1.1371171566004986e-06,
-      "loss": 1.0507,
+      "loss": 1.0508,
       "step": 590
     },
     {
       "epoch": 0.6253284287966369,
-      "grad_norm": 0.7898748331831235,
+      "grad_norm": 0.7901099674485127,
       "learning_rate": 1.1104406553301357e-06,
-      "loss": 1.0637,
+      "loss": 1.0636,
       "step": 595
     },
     {
       "epoch": 0.6305832895428272,
-      "grad_norm": 0.8261021360086062,
+      "grad_norm": 0.8076939152475928,
       "learning_rate": 1.0838956368572335e-06,
       "loss": 1.0526,
       "step": 600
     },
     {
       "epoch": 0.6358381502890174,
-      "grad_norm": 0.8086241727940497,
+      "grad_norm": 0.799554008565015,
       "learning_rate": 1.0574910605697135e-06,
       "loss": 1.0715,
       "step": 605
     },
     {
       "epoch": 0.6410930110352075,
-      "grad_norm": 0.9088888967600997,
+      "grad_norm": 0.7979721444226595,
       "learning_rate": 1.03123583845391e-06,
       "loss": 1.0692,
       "step": 610
     },
     {
       "epoch": 0.6463478717813977,
-      "grad_norm": 0.8079809849272349,
+      "grad_norm": 0.8069555984252168,
       "learning_rate": 1.0051388320866258e-06,
       "loss": 1.0583,
       "step": 615
     },
     {
       "epoch": 0.651602732527588,
-      "grad_norm": 0.8422109756331945,
+      "grad_norm": 0.8547163693801487,
       "learning_rate": 9.792088496441992e-07,
-      "loss": 1.0819,
+      "loss": 1.082,
       "step": 620
     },
     {
       "epoch": 0.6568575932737782,
-      "grad_norm": 0.8172134359132237,
+      "grad_norm": 0.7889726381679567,
       "learning_rate": 9.53454642929601e-07,
       "loss": 1.0772,
       "step": 625
     },
     {
       "epoch": 0.6621124540199684,
-      "grad_norm": 0.9233197165497478,
+      "grad_norm": 0.8667983764401236,
       "learning_rate": 9.278849044185509e-07,
-      "loss": 1.0637,
+      "loss": 1.0636,
       "step": 630
     },
     {
       "epoch": 0.6673673147661587,
-      "grad_norm": 0.9171808214559961,
+      "grad_norm": 0.8030729729500611,
       "learning_rate": 9.025082643256647e-07,
-      "loss": 1.043,
+      "loss": 1.0431,
       "step": 635
     },
     {
       "epoch": 0.6726221755123489,
-      "grad_norm": 0.8043456622976703,
+      "grad_norm": 0.9156958348117665,
       "learning_rate": 8.77333287691609e-07,
-      "loss": 1.0584,
+      "loss": 1.0585,
       "step": 640
     },
     {
       "epoch": 0.6778770362585391,
-      "grad_norm": 0.8108655439985347,
+      "grad_norm": 0.8108123796700908,
       "learning_rate": 8.523684714922608e-07,
       "loss": 1.0742,
       "step": 645
     },
     {
       "epoch": 0.6831318970047294,
-      "grad_norm": 0.7968848199287993,
+      "grad_norm": 0.8023252184856209,
       "learning_rate": 8.276222417708309e-07,
       "loss": 1.0557,
       "step": 650
     },
     {
       "epoch": 0.6883867577509196,
-      "grad_norm": 0.8129711567557044,
+      "grad_norm": 0.8088733267219154,
       "learning_rate": 8.031029507939401e-07,
-      "loss": 1.0548,
+      "loss": 1.0549,
       "step": 655
     },
     {
       "epoch": 0.6936416184971098,
-      "grad_norm": 1.004235271804759,
+      "grad_norm": 0.7972317042691273,
       "learning_rate": 7.788188742325803e-07,
-      "loss": 1.0612,
+      "loss": 1.0615,
       "step": 660
     },
     {
       "epoch": 0.6988964792433,
-      "grad_norm": 0.8208103251521766,
+      "grad_norm": 0.8305529275791657,
       "learning_rate": 7.547782083689479e-07,
       "loss": 1.0643,
       "step": 665
     },
     {
       "epoch": 0.7041513399894903,
-      "grad_norm": 0.8005037088376794,
+      "grad_norm": 0.8071361500941268,
       "learning_rate": 7.309890673300506e-07,
-      "loss": 1.045,
+      "loss": 1.0452,
       "step": 670
     },
     {
       "epoch": 0.7094062007356805,
-      "grad_norm": 1.3197940343995904,
+      "grad_norm": 0.8100694184976672,
       "learning_rate": 7.074594803490618e-07,
-      "loss": 1.0518,
+      "loss": 1.0517,
       "step": 675
     },
     {
       "epoch": 0.7146610614818707,
-      "grad_norm": 0.8555830044827688,
+      "grad_norm": 0.8505474154250443,
       "learning_rate": 6.841973890553168e-07,
-      "loss": 1.0611,
+      "loss": 1.0612,
       "step": 680
     },
     {
       "epoch": 0.719915922228061,
-      "grad_norm": 0.8099691174064823,
+      "grad_norm": 0.8165179858505918,
       "learning_rate": 6.6121064479388e-07,
       "loss": 1.0365,
       "step": 685
     },
     {
       "epoch": 0.7251707829742512,
-      "grad_norm": 0.8053501843842861,
+      "grad_norm": 0.8100990471560885,
       "learning_rate": 6.385070059755846e-07,
-      "loss": 1.074,
+      "loss": 1.0739,
       "step": 690
     },
     {
       "epoch": 0.7304256437204414,
-      "grad_norm": 0.8346149088986957,
+      "grad_norm": 0.8184474658493306,
       "learning_rate": 6.160941354584404e-07,
-      "loss": 1.0609,
+      "loss": 1.0611,
       "step": 695
     },
     {
       "epoch": 0.7356805044666316,
-      "grad_norm": 0.8155402643621729,
+      "grad_norm": 0.8252657802847051,
       "learning_rate": 5.93979597961289e-07,
       "loss": 1.0505,
       "step": 700
     },
     {
       "epoch": 0.7409353652128219,
-      "grad_norm": 0.7796064768407052,
+      "grad_norm": 0.7762171975020372,
       "learning_rate": 5.721708575105861e-07,
       "loss": 1.0589,
       "step": 705
     },
     {
       "epoch": 0.7461902259590121,
-      "grad_norm": 0.8275975218466692,
+      "grad_norm": 0.8192139022265624,
       "learning_rate": 5.506752749211673e-07,
       "loss": 1.0546,
       "step": 710
     },
     {
       "epoch": 0.7514450867052023,
-      "grad_norm": 0.8163230049703886,
+      "grad_norm": 0.8083286951981082,
       "learning_rate": 5.295001053118499e-07,
-      "loss": 1.0562,
+      "loss": 1.0564,
       "step": 715
     },
     {
       "epoch": 0.7566999474513926,
-      "grad_norm": 0.985195116035074,
+      "grad_norm": 0.792026256039253,
       "learning_rate": 5.086524956567084e-07,
-      "loss": 1.0713,
+      "loss": 1.0714,
       "step": 720
     },
     {
       "epoch": 0.7619548081975828,
-      "grad_norm": 0.96356977644872,
+      "grad_norm": 0.7949670797179523,
       "learning_rate": 4.88139482372852e-07,
       "loss": 1.0375,
       "step": 725
     },
     {
       "epoch": 0.767209668943773,
-      "grad_norm": 0.8163211083671094,
+      "grad_norm": 0.9772289752161802,
       "learning_rate": 4.679679889455153e-07,
-      "loss": 1.0525,
+      "loss": 1.0526,
       "step": 730
     },
     {
       "epoch": 0.7724645296899633,
-      "grad_norm": 0.7987340798976842,
+      "grad_norm": 0.8027717879760866,
       "learning_rate": 4.4814482359126713e-07,
-      "loss": 1.082,
+      "loss": 1.0819,
       "step": 735
     },
     {
       "epoch": 0.7777193904361535,
-      "grad_norm": 0.8117501301968848,
+      "grad_norm": 0.8018900351097608,
       "learning_rate": 4.2867667696012255e-07,
-      "loss": 1.0626,
+      "loss": 1.0627,
       "step": 740
     },
     {
       "epoch": 0.7829742511823437,
-      "grad_norm": 0.8929835779245712,
+      "grad_norm": 1.2040107178094928,
       "learning_rate": 4.0957011987733655e-07,
-      "loss": 1.076,
+      "loss": 1.0759,
       "step": 745
     },
     {
       "epoch": 0.7882291119285338,
-      "grad_norm": 0.8088648176966625,
+      "grad_norm": 0.8007696827925082,
       "learning_rate": 3.908316011256419e-07,
       "loss": 1.0744,
       "step": 750
     },
     {
       "epoch": 0.7934839726747241,
-      "grad_norm": 0.7980716806439362,
+      "grad_norm": 0.7873570653528454,
       "learning_rate": 3.7246744526867525e-07,
-      "loss": 1.0566,
+      "loss": 1.0568,
       "step": 755
     },
     {
       "epoch": 0.7987388334209143,
-      "grad_norm": 0.8089436395697871,
+      "grad_norm": 0.7923946383936664,
       "learning_rate": 3.5448385051633225e-07,
-      "loss": 1.0465,
+      "loss": 1.0464,
       "step": 760
     },
     {
       "epoch": 0.8039936941671045,
-      "grad_norm": 0.8015612514297874,
+      "grad_norm": 0.7863315246660161,
       "learning_rate": 3.368868866327678e-07,
-      "loss": 1.0611,
+      "loss": 1.0612,
       "step": 765
     },
     {
       "epoch": 0.8092485549132948,
-      "grad_norm": 0.7877310543664608,
+      "grad_norm": 1.479921699105945,
       "learning_rate": 3.1968249288774887e-07,
-      "loss": 1.0604,
+      "loss": 1.0605,
       "step": 770
     },
     {
       "epoch": 0.814503415659485,
-      "grad_norm": 0.8023241081523638,
+      "grad_norm": 0.8008655410629794,
       "learning_rate": 3.0287647605205155e-07,
-      "loss": 1.0723,
+      "loss": 1.0724,
       "step": 775
     },
     {
       "epoch": 0.8197582764056752,
-      "grad_norm": 0.8051580485406397,
+      "grad_norm": 0.8062826939041355,
       "learning_rate": 2.86474508437579e-07,
       "loss": 1.0579,
       "step": 780
     },
     {
       "epoch": 0.8250131371518655,
-      "grad_norm": 0.8081087810817446,
+      "grad_norm": 0.8029021628515398,
       "learning_rate": 2.704821259828608e-07,
       "loss": 1.0684,
       "step": 785
     },
     {
       "epoch": 0.8302679978980557,
-      "grad_norm": 0.8036521909829868,
+      "grad_norm": 0.8105469057959347,
       "learning_rate": 2.5490472638458195e-07,
-      "loss": 1.0484,
+      "loss": 1.0485,
       "step": 790
     },
     {
       "epoch": 0.8355228586442459,
-      "grad_norm": 0.8069169918745295,
+      "grad_norm": 0.7927228754247995,
       "learning_rate": 2.3974756727576886e-07,
       "loss": 1.0698,
       "step": 795
     },
     {
       "epoch": 0.8407777193904361,
-      "grad_norm": 0.8010632009065842,
+      "grad_norm": 0.7849174985906052,
       "learning_rate": 2.2501576445125077e-07,
-      "loss": 1.0592,
+      "loss": 1.0595,
       "step": 800
     },
     {
       "epoch": 0.8460325801366264,
-      "grad_norm": 0.8112748895796706,
+      "grad_norm": 0.8186938863488736,
       "learning_rate": 2.1071429014099365e-07,
-      "loss": 1.063,
+      "loss": 1.0631,
       "step": 805
     },
     {
       "epoch": 0.8512874408828166,
-      "grad_norm": 0.7987887307844111,
+      "grad_norm": 0.7864326637111713,
       "learning_rate": 1.9684797133188865e-07,
-      "loss": 1.0396,
+      "loss": 1.0397,
       "step": 810
     },
     {
       "epoch": 0.8565423016290068,
-      "grad_norm": 0.7974743093384434,
+      "grad_norm": 0.7954693830348463,
       "learning_rate": 1.8342148813856414e-07,
-      "loss": 1.0497,
+      "loss": 1.0498,
       "step": 815
     },
     {
       "epoch": 0.8617971623751971,
-      "grad_norm": 0.9606632973010945,
+      "grad_norm": 0.8900900173262289,
       "learning_rate": 1.7043937222376766e-07,
       "loss": 1.0484,
       "step": 820
     },
     {
       "epoch": 0.8670520231213873,
-      "grad_norm": 0.8027436257960178,
+      "grad_norm": 0.8002588089694325,
       "learning_rate": 1.579060052688548e-07,
-      "loss": 1.0674,
+      "loss": 1.0676,
       "step": 825
     },
     {
       "epoch": 0.8723068838675775,
-      "grad_norm": 0.8077058467523247,
+      "grad_norm": 0.7947929611802326,
       "learning_rate": 1.4582561749489847e-07,
       "loss": 1.0658,
       "step": 830
     },
     {
       "epoch": 0.8775617446137677,
-      "grad_norm": 0.796090164625106,
+      "grad_norm": 0.8288459681225425,
       "learning_rate": 1.3420228623491742e-07,
-      "loss": 1.0339,
+      "loss": 1.034,
       "step": 835
     },
     {
       "epoch": 0.882816605359958,
-      "grad_norm": 0.7956190292731135,
+      "grad_norm": 0.8346010909012226,
       "learning_rate": 1.2303993455770946e-07,
-      "loss": 1.0678,
+      "loss": 1.0679,
       "step": 840
     },
     {
       "epoch": 0.8880714661061482,
-      "grad_norm": 0.8060093191984734,
+      "grad_norm": 0.8436829522982748,
       "learning_rate": 1.1234232994374916e-07,
-      "loss": 1.0562,
+      "loss": 1.0561,
       "step": 845
     },
     {
       "epoch": 0.8933263268523384,
-      "grad_norm": 0.8052954491456666,
+      "grad_norm": 0.8025541491307114,
       "learning_rate": 1.0211308301360039e-07,
       "loss": 1.0635,
       "step": 850
     },
     {
       "epoch": 0.8985811875985287,
-      "grad_norm": 0.789706910907941,
+      "grad_norm": 0.8305527668792659,
       "learning_rate": 9.235564630927196e-08,
-      "loss": 1.0497,
+      "loss": 1.0499,
       "step": 855
     },
     {
       "epoch": 0.9038360483447189,
-      "grad_norm": 0.9407705619653095,
+      "grad_norm": 0.798582022149453,
       "learning_rate": 8.307331312892601e-08,
       "loss": 1.0573,
       "step": 860
     },
     {
       "epoch": 0.9090909090909091,
-      "grad_norm": 0.7849160686891212,
+      "grad_norm": 0.7805401930107103,
       "learning_rate": 7.426921641533562e-08,
-      "loss": 1.0535,
+      "loss": 1.0536,
       "step": 865
     },
     {
       "epoch": 0.9143457698370994,
-      "grad_norm": 0.7819582818532375,
+      "grad_norm": 0.7779984836642342,
       "learning_rate": 6.594632769846354e-08,
       "loss": 1.0605,
       "step": 870
     },
     {
       "epoch": 0.9196006305832896,
-      "grad_norm": 0.8235274235744744,
+      "grad_norm": 0.7849919719549066,
       "learning_rate": 5.810745609252166e-08,
-      "loss": 1.0481,
+      "loss": 1.048,
       "step": 875
     },
     {
       "epoch": 0.9248554913294798,
-      "grad_norm": 0.8326201665614436,
+      "grad_norm": 0.7974346377505517,
       "learning_rate": 5.0755247347847814e-08,
-      "loss": 1.05,
+      "loss": 1.0503,
       "step": 880
     },
     {
       "epoch": 0.9301103520756699,
-      "grad_norm": 0.8084764417080952,
+      "grad_norm": 0.8239903131727321,
       "learning_rate": 4.389218295792002e-08,
-      "loss": 1.0649,
+      "loss": 1.065,
       "step": 885
     },
     {
       "epoch": 0.9353652128218602,
-      "grad_norm": 0.8589755183334531,
+      "grad_norm": 0.782976096757697,
       "learning_rate": 3.7520579321812186e-08,
-      "loss": 1.073,
+      "loss": 1.0731,
       "step": 890
     },
     {
       "epoch": 0.9406200735680504,
-      "grad_norm": 0.8084706845759793,
+      "grad_norm": 0.7800809248031729,
       "learning_rate": 3.1642586962369765e-08,
       "loss": 1.0363,
       "step": 895
     },
     {
       "epoch": 0.9458749343142406,
-      "grad_norm": 0.7879594862220682,
+      "grad_norm": 0.7848631462241675,
       "learning_rate": 2.6260189800372757e-08,
       "loss": 1.0517,
       "step": 900
     },
     {
       "epoch": 0.9511297950604309,
-      "grad_norm": 0.7751896271813346,
+      "grad_norm": 0.7779665883276563,
       "learning_rate": 2.13752044849288e-08,
-      "loss": 1.0625,
+      "loss": 1.0626,
       "step": 905
     },
     {
       "epoch": 0.9563846558066211,
-      "grad_norm": 0.8037723865057425,
+      "grad_norm": 0.7949647408081179,
       "learning_rate": 1.698927978032383e-08,
-      "loss": 1.0486,
+      "loss": 1.0485,
       "step": 910
     },
     {
       "epoch": 0.9616395165528113,
-      "grad_norm": 0.7949419082977591,
+      "grad_norm": 0.7920892473556865,
       "learning_rate": 1.3103896009537208e-08,
       "loss": 1.0476,
       "step": 915
     },
     {
       "epoch": 0.9668943772990016,
-      "grad_norm": 0.7711064252783653,
+      "grad_norm": 0.7894649006593766,
       "learning_rate": 9.720364554606898e-09,
-      "loss": 1.0553,
+      "loss": 1.0554,
       "step": 920
     },
     {
       "epoch": 0.9721492380451918,
-      "grad_norm": 0.9190212136355437,
+      "grad_norm": 0.7926655176731906,
       "learning_rate": 6.839827414016675e-09,
       "loss": 1.0636,
       "step": 925
     },
     {
       "epoch": 0.977404098791382,
-      "grad_norm": 0.7895845548954091,
+      "grad_norm": 0.7738940217826289,
       "learning_rate": 4.463256817252792e-09,
       "loss": 1.0626,
       "step": 930
     },
     {
       "epoch": 0.9826589595375722,
-      "grad_norm": 0.7799515201158175,
+      "grad_norm": 0.7796303093908586,
       "learning_rate": 2.5914548966596285e-09,
-      "loss": 1.0584,
+      "loss": 1.0583,
       "step": 935
     },
     {
       "epoch": 0.9879138202837625,
-      "grad_norm": 0.7896292102395543,
+      "grad_norm": 0.7880712824013714,
       "learning_rate": 1.2250534167067561e-09,
-      "loss": 1.0562,
+      "loss": 1.0563,
       "step": 940
     },
     {
       "epoch": 0.9931686810299527,
-      "grad_norm": 0.8122355011762691,
+      "grad_norm": 0.7995046619036974,
       "learning_rate": 3.6451356075817287e-10,
-      "loss": 1.051,
+      "loss": 1.0511,
       "step": 945
     },
     {
       "epoch": 0.9984235417761429,
-      "grad_norm": 0.7870290469964676,
+      "grad_norm": 0.7975204845224678,
       "learning_rate": 1.0125775414981941e-11,
-      "loss": 1.096,
+      "loss": 1.0962,
       "step": 950
     },
     {
       "epoch": 0.999474513925381,
-      "eval_loss": 1.0638214349746704,
-      "eval_runtime": 593.8184,
-      "eval_samples_per_second": 22.692,
-      "eval_steps_per_second": 0.711,
+      "eval_loss": 1.0638784170150757,
+      "eval_runtime": 594.284,
+      "eval_samples_per_second": 22.674,
+      "eval_steps_per_second": 0.71,
       "step": 951
     },
     {
       "epoch": 0.999474513925381,
       "step": 951,
       "total_flos": 905758069751808.0,
-      "train_loss": 1.1023603343061092,
-      "train_runtime": 21224.38,
-      "train_samples_per_second": 5.738,
-      "train_steps_per_second": 0.045
+      "train_loss": 1.102385032064155,
+      "train_runtime": 21864.5552,
+      "train_samples_per_second": 5.57,
+      "train_steps_per_second": 0.043
     }
   ],
   "logging_steps": 5,