End of training

README.md

@@ -2,9 +2,15 @@
 license: other
 base_model: lewtun/gemma-7b-sft-full-deita-10k-v0
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- argilla/dpo-mix-7k
 model-index:
 - name: gemma-7b-dpo-full-mix1-beta-0.4-epoch-3
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gemma-7b-dpo-full-mix1-beta-0.4-epoch-3
 
-This model is a fine-tuned version of [lewtun/gemma-7b-sft-full-deita-10k-v0](https://huggingface.co/lewtun/gemma-7b-sft-full-deita-10k-v0) on the None dataset.
+This model is a fine-tuned version of [lewtun/gemma-7b-sft-full-deita-10k-v0](https://huggingface.co/lewtun/gemma-7b-sft-full-deita-10k-v0) on the argilla/dpo-mix-7k dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0775
-- Rewards/chosen: -6.7226
-- Rewards/rejected: -12.3440
-- Rewards/accuracies: 0.7396
-- Rewards/margins: 5.6214
-- Logps/rejected: -482.4095
-- Logps/chosen: -470.2108
-- Logits/rejected: 99.2330
-- Logits/chosen: 105.2899
+- Loss: 1.3044
+- Rewards/chosen: -6.0191
+- Rewards/rejected: -11.1290
+- Rewards/accuracies: 0.75
+- Rewards/margins: 5.1099
+- Logps/rejected: -479.3722
+- Logps/chosen: -468.4523
+- Logits/rejected: 98.8291
+- Logits/chosen: 104.8679
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.96,
+    "eval_logits/chosen": 104.86785125732422,
+    "eval_logits/rejected": 98.8290786743164,
+    "eval_logps/chosen": -468.4523010253906,
+    "eval_logps/rejected": -479.3721618652344,
+    "eval_loss": 1.3044307231903076,
+    "eval_rewards/accuracies": 0.75,
+    "eval_rewards/chosen": -6.019142150878906,
+    "eval_rewards/margins": 5.109894752502441,
+    "eval_rewards/rejected": -11.129035949707031,
+    "eval_runtime": 51.0519,
+    "eval_samples": 750,
+    "eval_samples_per_second": 14.691,
+    "eval_steps_per_second": 0.47,
     "train_loss": 0.5786063394103295,
     "train_runtime": 1791.9811,
     "train_samples": 6750,

config.json

@@ -23,6 +23,6 @@
   "rope_theta": 10000.0,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 256000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.96,
+    "eval_logits/chosen": 104.86785125732422,
+    "eval_logits/rejected": 98.8290786743164,
+    "eval_logps/chosen": -468.4523010253906,
+    "eval_logps/rejected": -479.3721618652344,
+    "eval_loss": 1.3044307231903076,
+    "eval_rewards/accuracies": 0.75,
+    "eval_rewards/chosen": -6.019142150878906,
+    "eval_rewards/margins": 5.109894752502441,
+    "eval_rewards/rejected": -11.129035949707031,
+    "eval_runtime": 51.0519,
+    "eval_samples": 750,
+    "eval_samples_per_second": 14.691,
+    "eval_steps_per_second": 0.47
+}