End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: llama3
 base_model: meta-llama/Meta-Llama-3-8B-Instruct
 tags:
+- alignment-handbook
 - trl
 - simpo
 - generated_from_trainer
+- trl
+- simpo
+- generated_from_trainer
+datasets:
+- yakazimir/llama3-ultrafeedback-armorm
 model-index:
 - name: llama3_l5_best_entropy
   results: []
@@ -16,18 +22,18 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3_l5_best_entropy
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the yakazimir/llama3-ultrafeedback-armorm dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.4402
-- Rewards/chosen: -8.2804
-- Rewards/rejected: -21.2503
+- Loss: 1.4398
+- Rewards/chosen: -8.4238
+- Rewards/rejected: -21.7089
 - Rewards/accuracies: 0.8795
-- Rewards/margins: 12.9699
-- Logps/rejected: -2.1250
-- Logps/chosen: -0.8280
-- Logits/rejected: -1.4434
-- Logits/chosen: -1.5360
-- Semantic Entropy: 0.8377
+- Rewards/margins: 13.2850
+- Logps/rejected: -2.1709
+- Logps/chosen: -0.8424
+- Logits/rejected: -1.4192
+- Logits/chosen: -1.5108
+- Semantic Entropy: 0.8327
 
 ## Model description
 

all_results.json

@@ -1,5 +1,19 @@
 {
     "epoch": 1.9978142076502732,
+    "eval_logits/chosen": -1.510839581489563,
+    "eval_logits/rejected": -1.4192240238189697,
+    "eval_logps/chosen": -0.842383861541748,
+    "eval_logps/rejected": -2.170888900756836,
+    "eval_loss": 1.4397584199905396,
+    "eval_rewards/accuracies": 0.8795180916786194,
+    "eval_rewards/chosen": -8.423839569091797,
+    "eval_rewards/margins": 13.28504753112793,
+    "eval_rewards/rejected": -21.708887100219727,
+    "eval_runtime": 33.4155,
+    "eval_samples": 1318,
+    "eval_samples_per_second": 39.443,
+    "eval_semantic_entropy": 0.8327488899230957,
+    "eval_steps_per_second": 2.484,
     "total_flos": 0.0,
     "train_loss": 1.6402891297830795,
     "train_runtime": 11806.3913,

config.json

@@ -24,6 +24,6 @@
   "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": 1.9978142076502732,
+    "eval_logits/chosen": -1.510839581489563,
+    "eval_logits/rejected": -1.4192240238189697,
+    "eval_logps/chosen": -0.842383861541748,
+    "eval_logps/rejected": -2.170888900756836,
+    "eval_loss": 1.4397584199905396,
+    "eval_rewards/accuracies": 0.8795180916786194,
+    "eval_rewards/chosen": -8.423839569091797,
+    "eval_rewards/margins": 13.28504753112793,
+    "eval_rewards/rejected": -21.708887100219727,
+    "eval_runtime": 33.4155,
+    "eval_samples": 1318,
+    "eval_samples_per_second": 39.443,
+    "eval_semantic_entropy": 0.8327488899230957,
+    "eval_steps_per_second": 2.484
+}