End of training

README.md

@@ -2,9 +2,15 @@
 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-dpo-full
   results: []
@@ -15,16 +21,16 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-dpo-full
 
-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: 0.9106
-- Rewards/chosen: -6.3982
-- Rewards/rejected: -10.5970
-- Rewards/accuracies: 0.7639
-- Rewards/margins: 4.1988
-- Logps/rejected: -366.1889
-- Logps/chosen: -345.9361
-- Logits/rejected: -1.3900
+- Loss: 0.9109
+- Rewards/chosen: -6.4067
+- Rewards/rejected: -10.6017
+- Rewards/accuracies: 0.7659
+- Rewards/margins: 4.1951
+- Logps/rejected: -366.2361
+- Logps/chosen: -346.0206
+- Logits/rejected: -1.3898
 - Logits/chosen: -1.6525
 
 ## Model description

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 2.998430141287284,
+    "eval_logits/chosen": -1.6524701118469238,
+    "eval_logits/rejected": -1.3897571563720703,
+    "eval_logps/chosen": -346.02056884765625,
+    "eval_logps/rejected": -366.2360534667969,
+    "eval_loss": 0.910855770111084,
+    "eval_rewards/accuracies": 0.7658730149269104,
+    "eval_rewards/chosen": -6.40665864944458,
+    "eval_rewards/margins": 4.195058822631836,
+    "eval_rewards/rejected": -10.601717948913574,
+    "eval_runtime": 196.377,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 10.184,
+    "eval_steps_per_second": 0.321,
     "total_flos": 0.0,
     "train_loss": 0.21882489114921755,
     "train_runtime": 49934.9402,

config.json

@@ -21,6 +21,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.40.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 2.998430141287284,
+    "eval_logits/chosen": -1.6524701118469238,
+    "eval_logits/rejected": -1.3897571563720703,
+    "eval_logps/chosen": -346.02056884765625,
+    "eval_logps/rejected": -366.2360534667969,
+    "eval_loss": 0.910855770111084,
+    "eval_rewards/accuracies": 0.7658730149269104,
+    "eval_rewards/chosen": -6.40665864944458,
+    "eval_rewards/margins": 4.195058822631836,
+    "eval_rewards/rejected": -10.601717948913574,
+    "eval_runtime": 196.377,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 10.184,
+    "eval_steps_per_second": 0.321
+}

runs/May14_17-59-18_ubuntu/events.out.tfevents.1715760404.ubuntu.774053.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7461453ff6e80718704c100dbcb0ede4a71933f6ea4491e10e5259261a2555a
+size 828