End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: llama3.1
 base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: llama-3-8b-dpo-full
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama-3-8b-dpo-full
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5582
-- Rewards/chosen: -1.3603
-- Rewards/rejected: -2.0529
-- Rewards/accuracies: 0.7262
-- Rewards/margins: 0.6926
-- Logps/rejected: -600.9839
-- Logps/chosen: -540.5128
-- Logits/rejected: -2.7438
-- Logits/chosen: -2.5853
+- Loss: 0.5576
+- Rewards/chosen: -1.3507
+- Rewards/rejected: -2.0430
+- Rewards/accuracies: 0.7341
+- Rewards/margins: 0.6923
+- Logps/rejected: -599.9959
+- Logps/chosen: -539.5551
+- Logits/rejected: -2.7918
+- Logits/chosen: -2.6247
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.998691442030882,
+    "eval_logits/chosen": -2.6246745586395264,
+    "eval_logits/rejected": -2.7917981147766113,
+    "eval_logps/chosen": -539.5550537109375,
+    "eval_logps/rejected": -599.9959106445312,
+    "eval_loss": 0.5575631260871887,
+    "eval_rewards/accuracies": 0.7341269850730896,
+    "eval_rewards/chosen": -1.3506927490234375,
+    "eval_rewards/margins": 0.6923374533653259,
+    "eval_rewards/rejected": -2.043030261993408,
+    "eval_runtime": 39.0144,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 51.263,
+    "eval_steps_per_second": 1.615,
     "total_flos": 0.0,
     "train_loss": 0.5939158873488068,
     "train_runtime": 3918.7641,

config.json

@@ -35,6 +35,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.45.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.998691442030882,
+    "eval_logits/chosen": -2.6246745586395264,
+    "eval_logits/rejected": -2.7917981147766113,
+    "eval_logps/chosen": -539.5550537109375,
+    "eval_logps/rejected": -599.9959106445312,
+    "eval_loss": 0.5575631260871887,
+    "eval_rewards/accuracies": 0.7341269850730896,
+    "eval_rewards/chosen": -1.3506927490234375,
+    "eval_rewards/margins": 0.6923374533653259,
+    "eval_rewards/rejected": -2.043030261993408,
+    "eval_runtime": 39.0144,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 51.263,
+    "eval_steps_per_second": 1.615
+}

runs/Oct07_16-47-37_a2ap-dgx001/events.out.tfevents.1728295433.a2ap-dgx001.921654.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7729f48883d3a9fb642d2d9e684720e05b02f5d2111491eaa5073896b4f80164
+size 828