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-0.05-epoch-3
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gemma-7b-dpo-full-mix1-beta-0.05-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: 0.4861
-- Rewards/chosen: -2.9739
-- Rewards/rejected: -4.4977
-- Rewards/accuracies: 0.6771
-- Rewards/margins: 1.5238
-- Logps/rejected: -541.5043
-- Logps/chosen: -512.8825
-- Logits/rejected: 87.7248
-- Logits/chosen: 93.8502
+- Loss: 0.4924
+- Rewards/chosen: -3.4915
+- Rewards/rejected: -5.2635
+- Rewards/accuracies: 0.7396
+- Rewards/margins: 1.7720
+- Logps/rejected: -556.8203
+- Logps/chosen: -523.2347
+- Logits/rejected: 83.0993
+- Logits/chosen: 88.9292
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.96,
+    "eval_logits/chosen": 88.92923736572266,
+    "eval_logits/rejected": 83.09931182861328,
+    "eval_logps/chosen": -523.2347412109375,
+    "eval_logps/rejected": -556.8203125,
+    "eval_loss": 0.4923539161682129,
+    "eval_rewards/accuracies": 0.7395833134651184,
+    "eval_rewards/chosen": -3.491511106491089,
+    "eval_rewards/margins": 1.7720268964767456,
+    "eval_rewards/rejected": -5.263538837432861,
+    "eval_runtime": 51.9214,
+    "eval_samples": 750,
+    "eval_samples_per_second": 14.445,
+    "eval_steps_per_second": 0.462,
     "train_loss": 0.28389122929328525,
     "train_runtime": 1811.0132,
     "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": 88.92923736572266,
+    "eval_logits/rejected": 83.09931182861328,
+    "eval_logps/chosen": -523.2347412109375,
+    "eval_logps/rejected": -556.8203125,
+    "eval_loss": 0.4923539161682129,
+    "eval_rewards/accuracies": 0.7395833134651184,
+    "eval_rewards/chosen": -3.491511106491089,
+    "eval_rewards/margins": 1.7720268964767456,
+    "eval_rewards/rejected": -5.263538837432861,
+    "eval_runtime": 51.9214,
+    "eval_samples": 750,
+    "eval_samples_per_second": 14.445,
+    "eval_steps_per_second": 0.462
+}