End of training

README.md

@@ -3,9 +3,16 @@ library_name: transformers
 license: apache-2.0
 base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/hh-rlhf-h4
+- data/chatbot_arena_conversations
 model-index:
 - name: tinyllama-1.1b-chat-v1.0-arena-hh-dpo
   results: []
@@ -16,7 +23,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tinyllama-1.1b-chat-v1.0-arena-hh-dpo
 
-This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) on the None dataset.
+This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) on the HuggingFaceH4/hh-rlhf-h4 and the data/chatbot_arena_conversations datasets.
+It achieves the following results on the evaluation set:
+- Loss: 0.6741
+- Rewards/chosen: -0.5672
+- Rewards/rejected: -0.6232
+- Rewards/accuracies: 0.5643
+- Rewards/margins: 0.0561
+- Logps/rejected: -221.2706
+- Logps/chosen: -218.0917
+- Logits/rejected: -2.3759
+- Logits/chosen: -2.3758
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9983492901947838,
+    "eval_logits/chosen": -2.375783920288086,
+    "eval_logits/rejected": -2.3758819103240967,
+    "eval_logps/chosen": -218.0917205810547,
+    "eval_logps/rejected": -221.27059936523438,
+    "eval_loss": 0.6741424202919006,
+    "eval_rewards/accuracies": 0.564338207244873,
+    "eval_rewards/chosen": -0.5671599507331848,
+    "eval_rewards/margins": 0.056060947477817535,
+    "eval_rewards/rejected": -0.6232209205627441,
+    "eval_runtime": 162.069,
+    "eval_samples": 8652,
+    "eval_samples_per_second": 53.385,
+    "eval_steps_per_second": 0.839,
     "total_flos": 0.0,
     "train_loss": 0.6782766637347993,
     "train_runtime": 9304.817,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9983492901947838,
+    "eval_logits/chosen": -2.375783920288086,
+    "eval_logits/rejected": -2.3758819103240967,
+    "eval_logps/chosen": -218.0917205810547,
+    "eval_logps/rejected": -221.27059936523438,
+    "eval_loss": 0.6741424202919006,
+    "eval_rewards/accuracies": 0.564338207244873,
+    "eval_rewards/chosen": -0.5671599507331848,
+    "eval_rewards/margins": 0.056060947477817535,
+    "eval_rewards/rejected": -0.6232209205627441,
+    "eval_runtime": 162.069,
+    "eval_samples": 8652,
+    "eval_samples_per_second": 53.385,
+    "eval_steps_per_second": 0.839
+}