End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: apache-2.0
 base_model: alignment-handbook/zephyr-7b-sft-full
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: zephyr-7b-align-scan-9e-07-0.86-linear-2.0
   results: []
@@ -16,17 +22,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-align-scan-9e-07-0.86-linear-2.0
 
-This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on an unknown dataset.
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.2247
-- Rewards/chosen: 0.1937
-- Rewards/rejected: -1.8484
-- Rewards/accuracies: 0.3214
-- Rewards/margins: 2.0420
-- Logps/rejected: -83.2776
-- Logps/chosen: -74.2660
-- Logits/rejected: -2.5909
-- Logits/chosen: -2.6070
+- Loss: 1.2207
+- Rewards/chosen: 0.7526
+- Rewards/rejected: -1.3224
+- Rewards/accuracies: 0.3294
+- Rewards/margins: 2.0750
+- Logps/rejected: -82.6661
+- Logps/chosen: -73.6161
+- Logits/rejected: -2.6026
+- Logits/chosen: -2.6190
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.0,
+    "eval_logits/chosen": -2.6189608573913574,
+    "eval_logits/rejected": -2.6026179790496826,
+    "eval_logps/chosen": -73.61612701416016,
+    "eval_logps/rejected": -82.66606140136719,
+    "eval_loss": 1.2207262516021729,
+    "eval_rewards/accuracies": 0.329365074634552,
+    "eval_rewards/chosen": 0.7525964975357056,
+    "eval_rewards/margins": 2.075026512145996,
+    "eval_rewards/rejected": -1.32243013381958,
+    "eval_runtime": 113.5924,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.607,
+    "eval_steps_per_second": 0.555,
     "total_flos": 0.0,
     "train_loss": 0.6794358088995106,
     "train_runtime": 6413.3344,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.0,
+    "eval_logits/chosen": -2.6189608573913574,
+    "eval_logits/rejected": -2.6026179790496826,
+    "eval_logps/chosen": -73.61612701416016,
+    "eval_logps/rejected": -82.66606140136719,
+    "eval_loss": 1.2207262516021729,
+    "eval_rewards/accuracies": 0.329365074634552,
+    "eval_rewards/chosen": 0.7525964975357056,
+    "eval_rewards/margins": 2.075026512145996,
+    "eval_rewards/rejected": -1.32243013381958,
+    "eval_runtime": 113.5924,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 17.607,
+    "eval_steps_per_second": 0.555
+}

runs/Sep27_11-09-17_qa-a100-005.crc.nd.edu/events.out.tfevents.1727456629.qa-a100-005.crc.nd.edu.4081404.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95fd9abf85adac17d7329172b9d4686e1e3d76ced98033a459a29956ef7cef5b
+size 828