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
 - cpo
 - generated_from_trainer
+- trl
+- cpo
+- generated_from_trainer
+datasets:
+- princeton-nlp/llama3-ultrafeedback
 model-index:
 - name: llama3.1-cpo-full
   results: []
@@ -16,18 +22,18 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3.1-cpo-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 princeton-nlp/llama3-ultrafeedback dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6704
-- Rewards/chosen: -15.1684
-- Rewards/rejected: -16.0433
-- Rewards/accuracies: 0.6293
-- Rewards/margins: 0.8749
-- Logps/rejected: -160.4333
-- Logps/chosen: -151.6842
-- Logits/rejected: -0.3369
-- Logits/chosen: -0.3497
-- Nll Loss: 0.4209
+- Loss: 1.6689
+- Rewards/chosen: -15.0012
+- Rewards/rejected: -15.8900
+- Rewards/accuracies: 0.6336
+- Rewards/margins: 0.8888
+- Logps/rejected: -158.8998
+- Logps/chosen: -150.0119
+- Logits/rejected: -0.3381
+- Logits/chosen: -0.3504
+- Nll Loss: 0.4161
 
 ## Model description
 

all_results.json

@@ -1,5 +1,19 @@
 {
     "epoch": 2.9930715935334873,
+    "eval_logits/chosen": -0.35042497515678406,
+    "eval_logits/rejected": -0.3381310701370239,
+    "eval_logps/chosen": -150.01190185546875,
+    "eval_logps/rejected": -158.89979553222656,
+    "eval_loss": 1.6688746213912964,
+    "eval_nll_loss": 0.416054904460907,
+    "eval_rewards/accuracies": 0.6336206793785095,
+    "eval_rewards/chosen": -15.001189231872559,
+    "eval_rewards/margins": 0.8887884616851807,
+    "eval_rewards/rejected": -15.88998031616211,
+    "eval_runtime": 41.8137,
+    "eval_samples": 1826,
+    "eval_samples_per_second": 43.67,
+    "eval_steps_per_second": 0.694,
     "total_flos": 0.0,
     "train_loss": 1.6959601876176433,
     "train_runtime": 15481.5304,

config.json

@@ -34,6 +34,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 2.9930715935334873,
+    "eval_logits/chosen": -0.35042497515678406,
+    "eval_logits/rejected": -0.3381310701370239,
+    "eval_logps/chosen": -150.01190185546875,
+    "eval_logps/rejected": -158.89979553222656,
+    "eval_loss": 1.6688746213912964,
+    "eval_nll_loss": 0.416054904460907,
+    "eval_rewards/accuracies": 0.6336206793785095,
+    "eval_rewards/chosen": -15.001189231872559,
+    "eval_rewards/margins": 0.8887884616851807,
+    "eval_rewards/rejected": -15.88998031616211,
+    "eval_runtime": 41.8137,
+    "eval_samples": 1826,
+    "eval_samples_per_second": 43.67,
+    "eval_steps_per_second": 0.694
+}

runs/Sep05_22-32-35_jjb_prism_dev2/events.out.tfevents.1725559716.jjb_prism_dev2.38742.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f11139865d90f784ba00d04630e4fbee5ed067e5e7bcdc0be9c24f6a72a90e33
+size 880