Model save

README.md

@@ -2,16 +2,10 @@
 license: mit
 base_model: HuggingFaceH4/mistral-7b-sft-beta
 tags:
-- alignment-handbook
-- trl
-- dpo
-- generated_from_trainer
 - trl
 - dpo
 - alignment-handbook
 - generated_from_trainer
-datasets:
-- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: DPO-Zephyr-7B
   results: []
@@ -22,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # DPO-Zephyr-7B
 
-This model is a fine-tuned version of [HuggingFaceH4/mistral-7b-sft-beta](https://huggingface.co/HuggingFaceH4/mistral-7b-sft-beta) on the HuggingFaceH4/ultrafeedback_binarized dataset.
+This model is a fine-tuned version of [HuggingFaceH4/mistral-7b-sft-beta](https://huggingface.co/HuggingFaceH4/mistral-7b-sft-beta) on an unknown dataset.
 
 ## Model description
 

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
     "total_flos": 0.0,
-    "train_loss": 0.030318551759704878,
-    "train_runtime": 314.312,
+    "train_loss": 0.030315694479590496,
+    "train_runtime": 321.4302,
     "train_samples": 15283,
-    "train_samples_per_second": 48.624,
-    "train_steps_per_second": 2.027
+    "train_samples_per_second": 47.547,
+    "train_steps_per_second": 1.982
 }

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.41.2",
-  "use_cache": true,
+  "use_cache": false,
   "vocab_size": 32000
 }

model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ae08a6faed6a61e5944bbacbf06f10a16c15d89ba131c1d0ace090fbd02a4bf
+oid sha256:7d53e3515878b0afe210acd491a71b83312ff85be681ba863db6064bcaac5dac
 size 4943162336

model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d1fb2c679d96030af45e036313e6ae0f7eb45339027b543f200f18eeab5c8363
+oid sha256:c1d72589ff788b09403e442b919aec91fd9a5896d66f004c08e7123819cf638d
 size 4999819336

model-00003-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26980bec61c9151f0b869f54f061bef53bc8b4c49466bced6ef7bb935a454d46
+oid sha256:2fc869f5a80ad958a13a649a7288bfe736b51cee3092fff57193169d90960ba6
 size 4540516344

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
     "total_flos": 0.0,
-    "train_loss": 0.030318551759704878,
-    "train_runtime": 314.312,
+    "train_loss": 0.030315694479590496,
+    "train_runtime": 321.4302,
     "train_samples": 15283,
-    "train_samples_per_second": 48.624,
-    "train_steps_per_second": 2.027
+    "train_samples_per_second": 47.547,
+    "train_steps_per_second": 1.982
 }

trainer_state.json

@@ -925,57 +925,57 @@
     },
     {
       "epoch": 0.957613814756672,
-      "grad_norm": 102.67877330344177,
+      "grad_norm": 101.21792438469406,
       "learning_rate": 2.734228528934679e-09,
-      "logits/chosen": -2.1245834827423096,
-      "logits/rejected": -2.1046948432922363,
-      "logps/chosen": -403.13818359375,
-      "logps/rejected": -488.44805908203125,
-      "loss": 0.5818,
-      "rewards/accuracies": 0.675000011920929,
-      "rewards/chosen": -1.370595932006836,
-      "rewards/margins": 0.6687647104263306,
-      "rewards/rejected": -2.039360523223877,
+      "logits/chosen": -2.12449312210083,
+      "logits/rejected": -2.1045031547546387,
+      "logps/chosen": -403.0607604980469,
+      "logps/rejected": -488.42059326171875,
+      "loss": 0.5817,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": -1.3698210716247559,
+      "rewards/margins": 0.6692644953727722,
+      "rewards/rejected": -2.0390853881835938,
       "step": 610
     },
     {
       "epoch": 0.9733124018838305,
-      "grad_norm": 62.505643361986195,
+      "grad_norm": 62.863474044060304,
       "learning_rate": 1.0851353912008642e-09,
-      "logits/chosen": -2.2792656421661377,
-      "logits/rejected": -2.1855509281158447,
-      "logps/chosen": -420.2193298339844,
-      "logps/rejected": -431.3525390625,
-      "loss": 0.5459,
+      "logits/chosen": -2.2791266441345215,
+      "logits/rejected": -2.18548583984375,
+      "logps/chosen": -420.2642517089844,
+      "logps/rejected": -431.43804931640625,
+      "loss": 0.5458,
       "rewards/accuracies": 0.737500011920929,
-      "rewards/chosen": -1.3569763898849487,
-      "rewards/margins": 0.5367645025253296,
-      "rewards/rejected": -1.8937408924102783,
+      "rewards/chosen": -1.357425332069397,
+      "rewards/margins": 0.5371707677841187,
+      "rewards/rejected": -1.8945963382720947,
       "step": 620
     },
     {
       "epoch": 0.989010989010989,
-      "grad_norm": 82.06301353927576,
+      "grad_norm": 82.02664725646991,
       "learning_rate": 1.840955480532924e-10,
-      "logits/chosen": -2.2514705657958984,
-      "logits/rejected": -2.204968214035034,
-      "logps/chosen": -372.6676330566406,
-      "logps/rejected": -425.01708984375,
+      "logits/chosen": -2.251556873321533,
+      "logits/rejected": -2.2051682472229004,
+      "logps/chosen": -372.6395568847656,
+      "logps/rejected": -425.1973571777344,
       "loss": 0.5177,
       "rewards/accuracies": 0.762499988079071,
-      "rewards/chosen": -1.1919991970062256,
-      "rewards/margins": 0.8975691795349121,
-      "rewards/rejected": -2.0895683765411377,
+      "rewards/chosen": -1.1917184591293335,
+      "rewards/margins": 0.899652361869812,
+      "rewards/rejected": -2.0913710594177246,
       "step": 630
     },
     {
       "epoch": 1.0,
       "step": 637,
       "total_flos": 0.0,
-      "train_loss": 0.030318551759704878,
-      "train_runtime": 314.312,
-      "train_samples_per_second": 48.624,
-      "train_steps_per_second": 2.027
+      "train_loss": 0.030315694479590496,
+      "train_runtime": 321.4302,
+      "train_samples_per_second": 47.547,
+      "train_steps_per_second": 1.982
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86f92d49570c8e77578cbebfa1f08306574d4aea2a55e4615fb1e8413abe47d0
+oid sha256:cd82e2e96a0037dfde71ec37184c9c39d84dea8a6e8371e90390c83ad86691d0
 size 6456