End of training

README.md

@@ -3,9 +3,16 @@ 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:
+- data/uf
+- data/rlced_conifer
 model-index:
 - name: zephyr-7b-uf-rc-small-dpo
   results: []
@@ -16,7 +23,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-uf-rc-small-dpo
 
-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 data/uf and the data/rlced_conifer datasets.
+It achieves the following results on the evaluation set:
+- Loss: 0.3443
+- Rewards/chosen: -1.6459
+- Rewards/rejected: -3.1222
+- Rewards/accuracies: 0.8701
+- Rewards/margins: 1.4764
+- Logps/rejected: -758.1630
+- Logps/chosen: -587.9950
+- Logits/rejected: -2.7637
+- Logits/chosen: -2.7840
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9874476987447699,
+    "eval_logits/chosen": -2.7839648723602295,
+    "eval_logits/rejected": -2.763737201690674,
+    "eval_logps/chosen": -587.9949951171875,
+    "eval_logps/rejected": -758.1629638671875,
+    "eval_loss": 0.3443484902381897,
+    "eval_rewards/accuracies": 0.8700980544090271,
+    "eval_rewards/chosen": -1.6458674669265747,
+    "eval_rewards/margins": 1.476360559463501,
+    "eval_rewards/rejected": -3.122227668762207,
+    "eval_runtime": 296.2496,
+    "eval_samples": 6491,
+    "eval_samples_per_second": 21.911,
+    "eval_steps_per_second": 0.344,
     "total_flos": 0.0,
     "train_loss": 0.47203683449050127,
     "train_runtime": 1681.7427,

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": 0.9874476987447699,
+    "eval_logits/chosen": -2.7839648723602295,
+    "eval_logits/rejected": -2.763737201690674,
+    "eval_logps/chosen": -587.9949951171875,
+    "eval_logps/rejected": -758.1629638671875,
+    "eval_loss": 0.3443484902381897,
+    "eval_rewards/accuracies": 0.8700980544090271,
+    "eval_rewards/chosen": -1.6458674669265747,
+    "eval_rewards/margins": 1.476360559463501,
+    "eval_rewards/rejected": -3.122227668762207,
+    "eval_runtime": 296.2496,
+    "eval_samples": 6491,
+    "eval_samples_per_second": 21.911,
+    "eval_steps_per_second": 0.344
+}