End of training

README.md

@@ -2,9 +2,15 @@
 license: llama3
 base_model: Magpie-Align/Llama-3-8B-Magpie-Mix-RC
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- princeton-nlp/llama3-ultrafeedback-armorm
 model-index:
 - name: Llama-3-8B-Magpie-Mix-RC-UltraDPO-07
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/uw-nsl/huggingface/runs/67yyhon0)
 # Llama-3-8B-Magpie-Mix-RC-UltraDPO-07
 
-This model is a fine-tuned version of [Magpie-Align/Llama-3-8B-Magpie-Mix-RC](https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Mix-RC) on an unknown dataset.
+This model is a fine-tuned version of [Magpie-Align/Llama-3-8B-Magpie-Mix-RC](https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Mix-RC) on the princeton-nlp/llama3-ultrafeedback-armorm dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3872
-- Rewards/chosen: -4.3693
-- Rewards/rejected: -5.9670
-- Rewards/accuracies: 0.8710
-- Rewards/margins: 1.5976
-- Logps/rejected: -863.7308
-- Logps/chosen: -710.0647
-- Logits/rejected: -1.1495
-- Logits/chosen: -1.1283
+- Loss: 0.3853
+- Rewards/chosen: -4.4256
+- Rewards/rejected: -6.0430
+- Rewards/accuracies: 0.875
+- Rewards/margins: 1.6173
+- Logps/rejected: -871.3292
+- Logps/chosen: -715.6917
+- Logits/rejected: -1.1545
+- Logits/chosen: -1.1333
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9981298423724285,
+    "eval_logits/chosen": -1.1332677602767944,
+    "eval_logits/rejected": -1.1544803380966187,
+    "eval_logps/chosen": -715.691650390625,
+    "eval_logps/rejected": -871.3291625976562,
+    "eval_loss": 0.38531965017318726,
+    "eval_rewards/accuracies": 0.875,
+    "eval_rewards/chosen": -4.425616264343262,
+    "eval_rewards/margins": 1.6173373460769653,
+    "eval_rewards/rejected": -6.0429534912109375,
+    "eval_runtime": 126.4968,
+    "eval_samples": 1961,
+    "eval_samples_per_second": 15.502,
+    "eval_steps_per_second": 0.49,
     "total_flos": 0.0,
     "train_loss": 0.4754098514673286,
     "train_runtime": 9903.4254,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.42.3",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9981298423724285,
+    "eval_logits/chosen": -1.1332677602767944,
+    "eval_logits/rejected": -1.1544803380966187,
+    "eval_logps/chosen": -715.691650390625,
+    "eval_logps/rejected": -871.3291625976562,
+    "eval_loss": 0.38531965017318726,
+    "eval_rewards/accuracies": 0.875,
+    "eval_rewards/chosen": -4.425616264343262,
+    "eval_rewards/margins": 1.6173373460769653,
+    "eval_rewards/rejected": -6.0429534912109375,
+    "eval_runtime": 126.4968,
+    "eval_samples": 1961,
+    "eval_samples_per_second": 15.502,
+    "eval_steps_per_second": 0.49
+}

runs/Jul15_06-39-44_compute-permanent-node-365/events.out.tfevents.1721036237.compute-permanent-node-365.232942.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:264c859f9657023f0cbf57bd32dfab8fca59cacefe19506c0b3923d1ec1d120d
+size 828