End of training

README.md

@@ -2,10 +2,13 @@
 license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: mistralai/Mistral-7B-v0.1
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: zephyr-7b
   results: []
@@ -16,19 +19,19 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-qlora](https://huggingface.co/alignment-handbook/zephyr-7b-sft-qlora) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4405
-- Rewards/chosen: -3.0809
-- Rewards/rejected: -4.8915
-- Rewards/accuracies: 0.3438
-- Rewards/margins: 1.8106
-- Logps/rejected: -566.3419
-- Logps/chosen: -371.9976
-- Logits/rejected: 4.5207
-- Logits/chosen: 4.3874
-- Use Label: 5649.7188
-- Pred Label: 1650.2812
+- Loss: 0.4399
+- Rewards/chosen: -3.1655
+- Rewards/rejected: -5.0200
+- Rewards/accuracies: 0.3398
+- Rewards/margins: 1.8544
+- Logps/rejected: -579.1896
+- Logps/chosen: -380.4651
+- Logits/rejected: 4.5948
+- Logits/chosen: 4.4604
+- Use Label: 6575.7188
+- Pred Label: 2212.2812
 
 ## Model description
 

all_results.json

@@ -1,5 +1,20 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": 4.460368633270264,
+    "eval_logits/rejected": 4.594798564910889,
+    "eval_logps/chosen": -380.465087890625,
+    "eval_logps/rejected": -579.1896362304688,
+    "eval_loss": 0.4399436116218567,
+    "eval_pred_label": 2212.28125,
+    "eval_rewards/accuracies": 0.33984375,
+    "eval_rewards/chosen": -3.1655280590057373,
+    "eval_rewards/margins": 1.8544387817382812,
+    "eval_rewards/rejected": -5.019967079162598,
+    "eval_runtime": 125.2743,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 15.965,
+    "eval_steps_per_second": 0.255,
+    "eval_use_label": 6575.71875,
     "train_loss": 0.5378267840019562,
     "train_runtime": 9600.9753,
     "train_samples": 61135,

eval_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": 1.828277349472046,
-    "eval_logits/rejected": 1.9234589338302612,
-    "eval_logps/chosen": -131.74139404296875,
-    "eval_logps/rejected": -192.50405883789062,
-    "eval_loss": 0.5775244235992432,
-    "eval_pred_label": 1039.875,
-    "eval_rewards/accuracies": 0.3671875,
-    "eval_rewards/chosen": -0.678291380405426,
-    "eval_rewards/margins": 0.47481971979141235,
-    "eval_rewards/rejected": -1.153111219406128,
-    "eval_runtime": 125.2019,
+    "eval_logits/chosen": 4.460368633270264,
+    "eval_logits/rejected": 4.594798564910889,
+    "eval_logps/chosen": -380.465087890625,
+    "eval_logps/rejected": -579.1896362304688,
+    "eval_loss": 0.4399436116218567,
+    "eval_pred_label": 2212.28125,
+    "eval_rewards/accuracies": 0.33984375,
+    "eval_rewards/chosen": -3.1655280590057373,
+    "eval_rewards/margins": 1.8544387817382812,
+    "eval_rewards/rejected": -5.019967079162598,
+    "eval_runtime": 125.2743,
     "eval_samples": 2000,
-    "eval_samples_per_second": 15.974,
-    "eval_steps_per_second": 0.256,
-    "eval_use_label": 7748.125
+    "eval_samples_per_second": 15.965,
+    "eval_steps_per_second": 0.255,
+    "eval_use_label": 6575.71875
 }