Model save

README.md

@@ -2,13 +2,11 @@
 license: apache-2.0
 base_model: mistralai/Mistral-7B-v0.1
 tags:
-- alignment-handbook
-- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- HuggingFaceH4/ultrachat_200k
+- generator
 model-index:
 - name: zephyr-7b-sft-full
   results: []
@@ -19,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-sft-full
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the HuggingFaceH4/ultrachat_200k dataset.
+This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9354
+- Loss: 0.9353
 
 ## Model description
 
@@ -57,7 +55,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 0.9077        | 1.0   | 1090 | 0.9354          |
+| 0.9075        | 1.0   | 1090 | 0.9353          |
 
 
 ### Framework versions

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 1.0,
-    "eval_loss": 0.9353834986686707,
-    "eval_runtime": 143.2861,
+    "eval_loss": 0.9353461265563965,
+    "eval_runtime": 141.7596,
     "eval_samples": 23110,
-    "eval_samples_per_second": 107.694,
-    "eval_steps_per_second": 1.689,
-    "train_loss": 0.9776670060026537,
-    "train_runtime": 5614.3798,
+    "eval_samples_per_second": 108.853,
+    "eval_steps_per_second": 1.707,
+    "train_loss": 0.9776473476252424,
+    "train_runtime": 5608.2856,
     "train_samples": 207865,
-    "train_samples_per_second": 24.834,
+    "train_samples_per_second": 24.861,
     "train_steps_per_second": 0.194
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "eval_loss": 0.9353834986686707,
-    "eval_runtime": 143.2861,
+    "eval_loss": 0.9353461265563965,
+    "eval_runtime": 141.7596,
     "eval_samples": 23110,
-    "eval_samples_per_second": 107.694,
-    "eval_steps_per_second": 1.689
+    "eval_samples_per_second": 108.853,
+    "eval_steps_per_second": 1.707
 }

model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8f0995a437e42c9046c3964bb7d858c34866d6b4f1de4ef12f98129888ee3fc
+oid sha256:ab3121e8d66828f8888b456267d5ce7e87063138a04ddde287016df065852277
 size 4943162336

model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b789d79949a4f5de8001f9e01a619675b6d2675041e49fb912ebe6c8e7d5b52
+oid sha256:9fa2d53c86d1b508a7440b1a6aa57652ac00fe7307f67b3b184cfbe2e84d8d55
 size 4999819336

model-00003-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d4b611a06358f2462ff5cc487cd28bd2de498fde53ba13e219b972869ebe1d5
+oid sha256:f5a120540ca1ffa8b3b7e66aa3a45760092305697b9e0b23a2399ab56b3f0771
 size 4540516344

runs/Jan10_01-40-06_ip-26-0-168-34/events.out.tfevents.1704850836.ip-26-0-168-34.1133327.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd358ecb98e337e7bdb0fadcbeedbec16960ffebf7e300cd81091cb11328b54b
-size 35928
+oid sha256:cd7f25fbf29b87da1eb52c6961c9817d4df7207797ebcaa5e3a8d23380bce5d5
+size 39379

runs/Jan10_01-40-06_ip-26-0-168-34/events.out.tfevents.1704856586.ip-26-0-168-34.1133327.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98419bcde11bdd300eba86cae7c9590c8fa1aa6a585ea4be1e4772dafbb18fc6
+size 359

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "train_loss": 0.9776670060026537,
-    "train_runtime": 5614.3798,
+    "train_loss": 0.9776473476252424,
+    "train_runtime": 5608.2856,
     "train_samples": 207865,
-    "train_samples_per_second": 24.834,
+    "train_samples_per_second": 24.861,
     "train_steps_per_second": 0.194
 }

trainer_state.json

@@ -23,7 +23,7 @@
     {
       "epoch": 0.01,
       "learning_rate": 1.8348623853211011e-06,
-      "loss": 1.3918,
+      "loss": 1.3919,
       "step": 10
     },
     {
@@ -41,13 +41,13 @@
     {
       "epoch": 0.02,
       "learning_rate": 4.587155963302753e-06,
-      "loss": 1.1173,
+      "loss": 1.1172,
       "step": 25
     },
     {
       "epoch": 0.03,
       "learning_rate": 5.504587155963303e-06,
-      "loss": 1.0705,
+      "loss": 1.0704,
       "step": 30
     },
     {
@@ -59,7 +59,7 @@
     {
       "epoch": 0.04,
       "learning_rate": 7.3394495412844045e-06,
-      "loss": 1.0521,
+      "loss": 1.052,
       "step": 40
     },
     {
@@ -71,49 +71,49 @@
     {
       "epoch": 0.05,
       "learning_rate": 9.174311926605506e-06,
-      "loss": 1.0309,
+      "loss": 1.0308,
       "step": 50
     },
     {
       "epoch": 0.05,
       "learning_rate": 1.0091743119266055e-05,
-      "loss": 1.0092,
+      "loss": 1.0091,
       "step": 55
     },
     {
       "epoch": 0.06,
       "learning_rate": 1.1009174311926607e-05,
-      "loss": 1.0359,
+      "loss": 1.0356,
       "step": 60
     },
     {
       "epoch": 0.06,
       "learning_rate": 1.1926605504587156e-05,
-      "loss": 1.0081,
+      "loss": 1.0079,
       "step": 65
     },
     {
       "epoch": 0.06,
       "learning_rate": 1.2844036697247708e-05,
-      "loss": 1.0002,
+      "loss": 1.0001,
       "step": 70
     },
     {
       "epoch": 0.07,
       "learning_rate": 1.3761467889908258e-05,
-      "loss": 1.0003,
+      "loss": 1.0001,
       "step": 75
     },
     {
       "epoch": 0.07,
       "learning_rate": 1.4678899082568809e-05,
-      "loss": 0.992,
+      "loss": 0.9918,
       "step": 80
     },
     {
       "epoch": 0.08,
       "learning_rate": 1.559633027522936e-05,
-      "loss": 1.0151,
+      "loss": 1.0148,
       "step": 85
     },
     {
@@ -131,7 +131,7 @@
     {
       "epoch": 0.09,
       "learning_rate": 1.834862385321101e-05,
-      "loss": 1.0162,
+      "loss": 1.0163,
       "step": 100
     },
     {
@@ -143,13 +143,13 @@
     {
       "epoch": 0.1,
       "learning_rate": 1.999994872196626e-05,
-      "loss": 1.0054,
+      "loss": 1.0056,
       "step": 110
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.9998154046002822e-05,
-      "loss": 1.014,
+      "loss": 1.0141,
       "step": 115
     },
     {
@@ -161,37 +161,37 @@
     {
       "epoch": 0.11,
       "learning_rate": 1.9986875683942535e-05,
-      "loss": 1.0201,
+      "loss": 1.02,
       "step": 125
     },
     {
       "epoch": 0.12,
       "learning_rate": 1.9977394889447526e-05,
-      "loss": 1.0236,
+      "loss": 1.0237,
       "step": 130
     },
     {
       "epoch": 0.12,
       "learning_rate": 1.9965356041462954e-05,
-      "loss": 1.0303,
+      "loss": 1.0304,
       "step": 135
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.9950762226567783e-05,
-      "loss": 1.0282,
+      "loss": 1.0284,
       "step": 140
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.9933617186395917e-05,
-      "loss": 1.0044,
+      "loss": 1.0046,
       "step": 145
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.9913925316676946e-05,
-      "loss": 1.0025,
+      "loss": 1.0026,
       "step": 150
     },
     {
@@ -209,97 +209,97 @@
     {
       "epoch": 0.15,
       "learning_rate": 1.9839622474129595e-05,
-      "loss": 1.0161,
+      "loss": 1.016,
       "step": 165
     },
     {
       "epoch": 0.16,
       "learning_rate": 1.9809800282473014e-05,
-      "loss": 0.9981,
+      "loss": 0.9979,
       "step": 170
     },
     {
       "epoch": 0.16,
       "learning_rate": 1.977746300605507e-05,
-      "loss": 1.0223,
+      "loss": 1.0219,
       "step": 175
     },
     {
       "epoch": 0.17,
       "learning_rate": 1.9742618935665478e-05,
-      "loss": 1.0088,
+      "loss": 1.0086,
       "step": 180
     },
     {
       "epoch": 0.17,
       "learning_rate": 1.9705277004798072e-05,
-      "loss": 0.9986,
+      "loss": 0.9983,
       "step": 185
     },
     {
       "epoch": 0.17,
       "learning_rate": 1.9665446787360444e-05,
-      "loss": 1.0107,
+      "loss": 1.0106,
       "step": 190
     },
     {
       "epoch": 0.18,
       "learning_rate": 1.9623138495219292e-05,
-      "loss": 1.0088,
+      "loss": 1.0084,
       "step": 195
     },
     {
       "epoch": 0.18,
       "learning_rate": 1.957836297558229e-05,
-      "loss": 1.0274,
+      "loss": 1.0272,
       "step": 200
     },
     {
       "epoch": 0.19,
       "learning_rate": 1.9531131708217005e-05,
-      "loss": 0.9892,
+      "loss": 0.989,
       "step": 205
     },
     {
       "epoch": 0.19,
       "learning_rate": 1.948145680250766e-05,
-      "loss": 1.0117,
+      "loss": 1.0114,
       "step": 210
     },
     {
       "epoch": 0.2,
       "learning_rate": 1.9429350994350483e-05,
-      "loss": 0.9933,
+      "loss": 0.993,
       "step": 215
     },
     {
       "epoch": 0.2,
       "learning_rate": 1.93748276428884e-05,
-      "loss": 1.0232,
+      "loss": 1.0229,
       "step": 220
     },
     {
       "epoch": 0.21,
       "learning_rate": 1.931790072708596e-05,
-      "loss": 1.0189,
+      "loss": 1.0191,
       "step": 225
     },
     {
       "epoch": 0.21,
       "learning_rate": 1.9258584842145342e-05,
-      "loss": 1.0018,
+      "loss": 1.0016,
       "step": 230
     },
     {
       "epoch": 0.22,
       "learning_rate": 1.9196895195764363e-05,
-      "loss": 0.9957,
+      "loss": 0.9955,
       "step": 235
     },
     {
       "epoch": 0.22,
       "learning_rate": 1.913284760423745e-05,
-      "loss": 1.0213,
+      "loss": 1.0212,
       "step": 240
     },
     {
@@ -311,13 +311,13 @@
     {
       "epoch": 0.23,
       "learning_rate": 1.8997744869421248e-05,
-      "loss": 1.0037,
+      "loss": 1.0036,
       "step": 250
     },
     {
       "epoch": 0.23,
       "learning_rate": 1.8926724364434447e-05,
-      "loss": 0.9953,
+      "loss": 0.9951,
       "step": 255
     },
     {
@@ -329,7 +329,7 @@
     {
       "epoch": 0.24,
       "learning_rate": 1.8777836117563894e-05,
-      "loss": 1.0198,
+      "loss": 1.0197,
       "step": 265
     },
     {
@@ -341,13 +341,13 @@
     {
       "epoch": 0.25,
       "learning_rate": 1.861994642880105e-05,
-      "loss": 1.0035,
+      "loss": 1.0036,
       "step": 275
     },
     {
       "epoch": 0.26,
       "learning_rate": 1.8537676285033886e-05,
-      "loss": 1.0062,
+      "loss": 1.0064,
       "step": 280
     },
     {
@@ -359,13 +359,13 @@
     {
       "epoch": 0.27,
       "learning_rate": 1.8366590857437182e-05,
-      "loss": 1.015,
+      "loss": 1.0154,
       "step": 290
     },
     {
       "epoch": 0.27,
       "learning_rate": 1.8277819437329577e-05,
-      "loss": 0.9928,
+      "loss": 0.993,
       "step": 295
     },
     {
@@ -377,61 +377,61 @@
     {
       "epoch": 0.28,
       "learning_rate": 1.809393297692334e-05,
-      "loss": 0.9891,
+      "loss": 0.9892,
       "step": 305
     },
     {
       "epoch": 0.28,
       "learning_rate": 1.799886508233829e-05,
-      "loss": 0.9866,
+      "loss": 0.9869,
       "step": 310
     },
     {
       "epoch": 0.29,
       "learning_rate": 1.790174639944997e-05,
-      "loss": 0.9891,
+      "loss": 0.9894,
       "step": 315
     },
     {
       "epoch": 0.29,
       "learning_rate": 1.780260182802314e-05,
-      "loss": 0.998,
+      "loss": 0.9984,
       "step": 320
     },
     {
       "epoch": 0.3,
       "learning_rate": 1.7701456787229805e-05,
-      "loss": 0.9875,
+      "loss": 0.9878,
       "step": 325
     },
     {
       "epoch": 0.3,
       "learning_rate": 1.7598337209132142e-05,
-      "loss": 1.01,
+      "loss": 1.0103,
       "step": 330
     },
     {
       "epoch": 0.31,
       "learning_rate": 1.7493269532033882e-05,
-      "loss": 0.9939,
+      "loss": 0.994,
       "step": 335
     },
     {
       "epoch": 0.31,
       "learning_rate": 1.738628069370195e-05,
-      "loss": 0.9979,
+      "loss": 0.9978,
       "step": 340
     },
     {
       "epoch": 0.32,
       "learning_rate": 1.7277398124460022e-05,
-      "loss": 1.0069,
+      "loss": 1.0071,
       "step": 345
     },
     {
       "epoch": 0.32,
       "learning_rate": 1.71666497401558e-05,
-      "loss": 0.9926,
+      "loss": 0.9925,
       "step": 350
     },
     {
@@ -443,7 +443,7 @@
     {
       "epoch": 0.33,
       "learning_rate": 1.6939669574305565e-05,
-      "loss": 1.0071,
+      "loss": 1.0069,
       "step": 360
     },
     {
@@ -455,25 +455,25 @@
     {
       "epoch": 0.34,
       "learning_rate": 1.6705572958388576e-05,
-      "loss": 0.999,
+      "loss": 0.9989,
       "step": 370
     },
     {
       "epoch": 0.34,
       "learning_rate": 1.6585930722009602e-05,
-      "loss": 0.9953,
+      "loss": 0.9951,
       "step": 375
     },
     {
       "epoch": 0.35,
       "learning_rate": 1.6464599952375998e-05,
-      "loss": 0.9973,
+      "loss": 0.9971,
       "step": 380
     },
     {
       "epoch": 0.35,
       "learning_rate": 1.63416117568662e-05,
-      "loss": 0.9813,
+      "loss": 0.9811,
       "step": 385
     },
     {
@@ -485,31 +485,31 @@
     {
       "epoch": 0.36,
       "learning_rate": 1.6090789634342278e-05,
-      "loss": 1.003,
+      "loss": 1.0029,
       "step": 395
     },
     {
       "epoch": 0.37,
       "learning_rate": 1.5963020014335437e-05,
-      "loss": 0.9932,
+      "loss": 0.9934,
       "step": 400
     },
     {
       "epoch": 0.37,
       "learning_rate": 1.583372156597961e-05,
-      "loss": 0.9754,
+      "loss": 0.9753,
       "step": 405
     },
     {
       "epoch": 0.38,
       "learning_rate": 1.570292743944583e-05,
-      "loss": 1.0097,
+      "loss": 1.009,
       "step": 410
     },
     {
       "epoch": 0.38,
       "learning_rate": 1.557067116837444e-05,
-      "loss": 0.9678,
+      "loss": 0.968,
       "step": 415
     },
     {
@@ -521,43 +521,43 @@
     {
       "epoch": 0.39,
       "learning_rate": 1.530190819284555e-05,
-      "loss": 1.0128,
+      "loss": 1.0131,
       "step": 425
     },
     {
       "epoch": 0.39,
       "learning_rate": 1.5165470395159314e-05,
-      "loss": 0.9882,
+      "loss": 0.9881,
       "step": 430
     },
     {
       "epoch": 0.4,
       "learning_rate": 1.5027708248811331e-05,
-      "loss": 0.9888,
+      "loss": 0.9887,
       "step": 435
     },
     {
       "epoch": 0.4,
       "learning_rate": 1.4888657073937077e-05,
-      "loss": 0.979,
+      "loss": 0.9793,
       "step": 440
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.4748352521159492e-05,
-      "loss": 0.9767,
+      "loss": 0.9766,
       "step": 445
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.4606830562448692e-05,
-      "loss": 0.9875,
+      "loss": 0.9877,
       "step": 450
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.4464127481899312e-05,
-      "loss": 0.9999,
+      "loss": 0.9998,
       "step": 455
     },
     {
@@ -581,13 +581,13 @@
     {
       "epoch": 0.44,
       "learning_rate": 1.3882240024460928e-05,
-      "loss": 0.9861,
+      "loss": 0.986,
       "step": 475
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.3734185865027061e-05,
-      "loss": 0.9833,
+      "loss": 0.9834,
       "step": 480
     },
     {
@@ -599,31 +599,31 @@
     {
       "epoch": 0.45,
       "learning_rate": 1.3435243583752294e-05,
-      "loss": 0.9752,
+      "loss": 0.9749,
       "step": 490
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.3284432106201233e-05,
-      "loss": 0.9796,
+      "loss": 0.9799,
       "step": 495
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.313277854982062e-05,
-      "loss": 0.9584,
+      "loss": 0.9585,
       "step": 500
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.2980321796293838e-05,
-      "loss": 0.9884,
+      "loss": 0.9886,
       "step": 505
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.2827100933231904e-05,
-      "loss": 0.9695,
+      "loss": 0.9694,
       "step": 510
     },
     {
@@ -641,7 +641,7 @@
     {
       "epoch": 0.48,
       "learning_rate": 1.2363247441059775e-05,
-      "loss": 0.971,
+      "loss": 0.9711,
       "step": 525
     },
     {
@@ -659,7 +659,7 @@
     {
       "epoch": 0.5,
       "learning_rate": 1.189394177207125e-05,
-      "loss": 0.9825,
+      "loss": 0.9826,
       "step": 540
     },
     {
@@ -671,13 +671,13 @@
     {
       "epoch": 0.5,
       "learning_rate": 1.1578576573543541e-05,
-      "loss": 0.9612,
+      "loss": 0.9614,
       "step": 550
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.1420266647205232e-05,
-      "loss": 0.9692,
+      "loss": 0.9689,
       "step": 555
     },
     {
@@ -689,13 +689,13 @@
     {
       "epoch": 0.52,
       "learning_rate": 1.1102595071354471e-05,
-      "loss": 0.9446,
+      "loss": 0.9447,
       "step": 565
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.0943314868040365e-05,
-      "loss": 0.9582,
+      "loss": 0.9584,
       "step": 570
     },
     {
@@ -713,13 +713,13 @@
     {
       "epoch": 0.54,
       "learning_rate": 1.0464186795625481e-05,
-      "loss": 0.9636,
+      "loss": 0.9637,
       "step": 585
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.0304184775393642e-05,
-      "loss": 0.9571,
+      "loss": 0.9573,
       "step": 590
     },
     {
@@ -737,25 +737,25 @@
     {
       "epoch": 0.56,
       "learning_rate": 9.823874962247565e-06,
-      "loss": 0.9651,
+      "loss": 0.9652,
       "step": 605
     },
     {
       "epoch": 0.56,
       "learning_rate": 9.663807268427197e-06,
-      "loss": 0.9511,
+      "loss": 0.951,
       "step": 610
     },
     {
       "epoch": 0.56,
       "learning_rate": 9.503825769350016e-06,
-      "loss": 0.9731,
+      "loss": 0.9735,
       "step": 615
     },
     {
       "epoch": 0.57,
       "learning_rate": 9.343971481858246e-06,
-      "loss": 0.9487,
+      "loss": 0.9491,
       "step": 620
     },
     {
@@ -767,7 +767,7 @@
     {
       "epoch": 0.58,
       "learning_rate": 9.024808435416435e-06,
-      "loss": 0.9431,
+      "loss": 0.9434,
       "step": 630
     },
     {
@@ -779,19 +779,19 @@
     {
       "epoch": 0.59,
       "learning_rate": 8.706645422477739e-06,
-      "loss": 0.9618,
+      "loss": 0.9619,
       "step": 640
     },
     {
       "epoch": 0.59,
       "learning_rate": 8.548040936496989e-06,
-      "loss": 0.9452,
+      "loss": 0.9455,
       "step": 645
     },
     {
       "epoch": 0.6,
       "learning_rate": 8.389808710909881e-06,
-      "loss": 0.9739,
+      "loss": 0.9741,
       "step": 650
     },
     {
@@ -809,25 +809,25 @@
     {
       "epoch": 0.61,
       "learning_rate": 7.917750740470116e-06,
-      "loss": 0.9695,
+      "loss": 0.9696,
       "step": 665
     },
     {
       "epoch": 0.61,
       "learning_rate": 7.761412129735853e-06,
-      "loss": 0.9626,
+      "loss": 0.9628,
       "step": 670
     },
     {
       "epoch": 0.62,
       "learning_rate": 7.605647459150961e-06,
-      "loss": 0.9482,
+      "loss": 0.9485,
       "step": 675
     },
     {
       "epoch": 0.62,
       "learning_rate": 7.4504966644264775e-06,
-      "loss": 0.9482,
+      "loss": 0.9485,
       "step": 680
     },
     {
@@ -839,7 +839,7 @@
     {
       "epoch": 0.63,
       "learning_rate": 7.142195648261747e-06,
-      "loss": 0.9646,
+      "loss": 0.9648,
       "step": 690
     },
     {
@@ -851,31 +851,31 @@
     {
       "epoch": 0.64,
       "learning_rate": 6.83682523588902e-06,
-      "loss": 0.9436,
+      "loss": 0.944,
       "step": 700
     },
     {
       "epoch": 0.65,
       "learning_rate": 6.685336991505122e-06,
-      "loss": 0.9576,
+      "loss": 0.9575,
       "step": 705
     },
     {
       "epoch": 0.65,
       "learning_rate": 6.5346985766979384e-06,
-      "loss": 0.9497,
+      "loss": 0.9498,
       "step": 710
     },
     {
       "epoch": 0.66,
       "learning_rate": 6.384948612883872e-06,
-      "loss": 0.9474,
+      "loss": 0.9475,
       "step": 715
     },
     {
       "epoch": 0.66,
       "learning_rate": 6.2361254936939e-06,
-      "loss": 0.9382,
+      "loss": 0.9385,
       "step": 720
     },
     {
@@ -893,31 +893,31 @@
     {
       "epoch": 0.67,
       "learning_rate": 5.79559751711138e-06,
-      "loss": 0.9349,
+      "loss": 0.9352,
       "step": 735
     },
     {
       "epoch": 0.68,
       "learning_rate": 5.650860813791786e-06,
-      "loss": 0.9444,
+      "loss": 0.9446,
       "step": 740
     },
     {
       "epoch": 0.68,
       "learning_rate": 5.507239164130501e-06,
-      "loss": 0.9522,
+      "loss": 0.9523,
       "step": 745
     },
     {
       "epoch": 0.69,
       "learning_rate": 5.364769390551225e-06,
-      "loss": 0.9507,
+      "loss": 0.9508,
       "step": 750
     },
     {
       "epoch": 0.69,
       "learning_rate": 5.223488020154028e-06,
-      "loss": 0.9461,
+      "loss": 0.946,
       "step": 755
     },
     {
@@ -935,13 +935,13 @@
     {
       "epoch": 0.71,
       "learning_rate": 4.807134973085036e-06,
-      "loss": 0.9437,
+      "loss": 0.9438,
       "step": 770
     },
     {
       "epoch": 0.71,
       "learning_rate": 4.670966253826027e-06,
-      "loss": 0.9515,
+      "loss": 0.9502,
       "step": 775
     },
     {
@@ -953,25 +953,25 @@
     {
       "epoch": 0.72,
       "learning_rate": 4.402762228128531e-06,
-      "loss": 0.9412,
+      "loss": 0.9408,
       "step": 785
     },
     {
       "epoch": 0.72,
       "learning_rate": 4.270795685155001e-06,
-      "loss": 0.9489,
+      "loss": 0.9482,
       "step": 790
     },
     {
       "epoch": 0.73,
       "learning_rate": 4.140298023713416e-06,
-      "loss": 0.9292,
+      "loss": 0.9288,
       "step": 795
     },
     {
       "epoch": 0.73,
       "learning_rate": 4.0113027014349374e-06,
-      "loss": 0.944,
+      "loss": 0.9438,
       "step": 800
     },
     {
@@ -983,49 +983,49 @@
     {
       "epoch": 0.74,
       "learning_rate": 3.757950970527249e-06,
-      "loss": 0.9365,
+      "loss": 0.9366,
       "step": 810
     },
     {
       "epoch": 0.75,
       "learning_rate": 3.633659517458736e-06,
-      "loss": 0.9312,
+      "loss": 0.931,
       "step": 815
     },
     {
       "epoch": 0.75,
       "learning_rate": 3.511000298021098e-06,
-      "loss": 0.95,
+      "loss": 0.9497,
       "step": 820
     },
     {
       "epoch": 0.76,
       "learning_rate": 3.39000476018726e-06,
-      "loss": 0.944,
+      "loss": 0.9438,
       "step": 825
     },
     {
       "epoch": 0.76,
       "learning_rate": 3.2707039253872796e-06,
-      "loss": 0.9354,
+      "loss": 0.9353,
       "step": 830
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.153128380554941e-06,
-      "loss": 0.9418,
+      "loss": 0.9419,
       "step": 835
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.037308270285709e-06,
-      "loss": 0.9204,
+      "loss": 0.9201,
       "step": 840
     },
     {
       "epoch": 0.78,
       "learning_rate": 2.923273289108115e-06,
-      "loss": 0.9462,
+      "loss": 0.9461,
       "step": 845
     },
     {
@@ -1037,7 +1037,7 @@
     {
       "epoch": 0.78,
       "learning_rate": 2.700675196245288e-06,
-      "loss": 0.9466,
+      "loss": 0.9465,
       "step": 855
     },
     {
@@ -1049,55 +1049,55 @@
     {
       "epoch": 0.79,
       "learning_rate": 2.485562370502279e-06,
-      "loss": 0.9283,
+      "loss": 0.9282,
       "step": 865
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.3808821740669608e-06,
-      "loss": 0.9294,
+      "loss": 0.9293,
       "step": 870
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.2781554044688015e-06,
-      "loss": 0.9404,
+      "loss": 0.9402,
       "step": 875
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.1774083993013715e-06,
-      "loss": 0.9363,
+      "loss": 0.9362,
       "step": 880
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.0786669885765044e-06,
-      "loss": 0.9345,
+      "loss": 0.9342,
       "step": 885
     },
     {
       "epoch": 0.82,
       "learning_rate": 1.981956488101898e-06,
-      "loss": 0.9244,
+      "loss": 0.9245,
       "step": 890
     },
     {
       "epoch": 0.82,
       "learning_rate": 1.8873016929904942e-06,
-      "loss": 0.9336,
+      "loss": 0.9338,
       "step": 895
     },
     {
       "epoch": 0.83,
       "learning_rate": 1.7947268713034128e-06,
-      "loss": 0.9481,
+      "loss": 0.9483,
       "step": 900
     },
     {
       "epoch": 0.83,
       "learning_rate": 1.704255757827963e-06,
-      "loss": 0.935,
+      "loss": 0.9351,
       "step": 905
     },
     {
@@ -1109,13 +1109,13 @@
     {
       "epoch": 0.84,
       "learning_rate": 1.529716891919074e-06,
-      "loss": 0.9213,
+      "loss": 0.9212,
       "step": 915
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.4456938886170413e-06,
-      "loss": 0.9381,
+      "loss": 0.9379,
       "step": 920
     },
     {
@@ -1127,19 +1127,19 @@
     {
       "epoch": 0.85,
       "learning_rate": 1.2842484469453365e-06,
-      "loss": 0.9158,
+      "loss": 0.9156,
       "step": 930
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.2068674007506787e-06,
-      "loss": 0.9288,
+      "loss": 0.9285,
       "step": 935
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.1317407810650372e-06,
-      "loss": 0.9213,
+      "loss": 0.9214,
       "step": 940
     },
     {
@@ -1163,37 +1163,37 @@
     {
       "epoch": 0.88,
       "learning_rate": 8.541550213737171e-07,
-      "loss": 0.924,
+      "loss": 0.9241,
       "step": 960
     },
     {
       "epoch": 0.89,
       "learning_rate": 7.905777244947954e-07,
-      "loss": 0.9192,
+      "loss": 0.9194,
       "step": 965
     },
     {
       "epoch": 0.89,
       "learning_rate": 7.293615845160196e-07,
-      "loss": 0.9125,
+      "loss": 0.9124,
       "step": 970
     },
     {
       "epoch": 0.89,
       "learning_rate": 6.705222963319191e-07,
-      "loss": 0.9282,
+      "loss": 0.9281,
       "step": 975
     },
     {
       "epoch": 0.9,
       "learning_rate": 6.140749454480932e-07,
-      "loss": 0.9384,
+      "loss": 0.9382,
       "step": 980
     },
     {
       "epoch": 0.9,
       "learning_rate": 5.600340041135133e-07,
-      "loss": 0.9259,
+      "loss": 0.9257,
       "step": 985
     },
     {
@@ -1211,25 +1211,25 @@
     {
       "epoch": 0.92,
       "learning_rate": 4.124850842338779e-07,
-      "loss": 0.9252,
+      "loss": 0.9249,
       "step": 1000
     },
     {
       "epoch": 0.92,
       "learning_rate": 3.6820211191520127e-07,
-      "loss": 0.9231,
+      "loss": 0.9233,
       "step": 1005
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.263885872300343e-07,
-      "loss": 0.9087,
+      "loss": 0.9086,
       "step": 1010
     },
     {
       "epoch": 0.93,
       "learning_rate": 2.870552305351382e-07,
-      "loss": 0.9291,
+      "loss": 0.929,
       "step": 1015
     },
     {
@@ -1247,7 +1247,7 @@
     {
       "epoch": 0.94,
       "learning_rate": 1.840338184455881e-07,
-      "loss": 0.9148,
+      "loss": 0.9144,
       "step": 1030
     },
     {
@@ -1265,43 +1265,43 @@
     {
       "epoch": 0.96,
       "learning_rate": 1.0365852555447642e-07,
-      "loss": 0.9407,
+      "loss": 0.9406,
       "step": 1045
     },
     {
       "epoch": 0.96,
       "learning_rate": 8.19327959602012e-08,
-      "loss": 0.9346,
+      "loss": 0.9343,
       "step": 1050
     },
     {
       "epoch": 0.97,
       "learning_rate": 6.274990913221035e-08,
-      "loss": 0.9198,
+      "loss": 0.92,
       "step": 1055
     },
     {
       "epoch": 0.97,
       "learning_rate": 4.6114783273213395e-08,
-      "loss": 0.9214,
+      "loss": 0.9212,
       "step": 1060
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.203168337845508e-08,
-      "loss": 0.9125,
+      "loss": 0.9127,
       "step": 1065
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.05042201422323e-08,
-      "loss": 0.9268,
+      "loss": 0.9267,
       "step": 1070
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.1535349032167908e-08,
-      "loss": 0.9249,
+      "loss": 0.9248,
       "step": 1075
     },
     {
@@ -1313,30 +1313,30 @@
     {
       "epoch": 1.0,
       "learning_rate": 1.2819245493955746e-09,
-      "loss": 0.9161,
+      "loss": 0.9159,
       "step": 1085
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.0,
-      "loss": 0.9077,
+      "loss": 0.9075,
       "step": 1090
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.9353834986686707,
-      "eval_runtime": 144.309,
-      "eval_samples_per_second": 106.93,
-      "eval_steps_per_second": 1.677,
+      "eval_loss": 0.9353461265563965,
+      "eval_runtime": 142.6393,
+      "eval_samples_per_second": 108.182,
+      "eval_steps_per_second": 1.697,
       "step": 1090
     },
     {
       "epoch": 1.0,
       "step": 1090,
       "total_flos": 456447649382400.0,
-      "train_loss": 0.9776670060026537,
-      "train_runtime": 5614.3798,
-      "train_samples_per_second": 24.834,
+      "train_loss": 0.9776473476252424,
+      "train_runtime": 5608.2856,
+      "train_samples_per_second": 24.861,
       "train_steps_per_second": 0.194
     }
   ],