End of training

README.md

@@ -2,9 +2,10 @@
 license: apache-2.0
 library_name: peft
 tags:
+- llama-factory
+- lora
 - trl
 - dpo
-- llama-factory
 - generated_from_trainer
 base_model: mistralai/Mistral-7B-Instruct-v0.2
 model-index:
@@ -17,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Mistral-7B-Instruct-v0.2-ORPO-SALT
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the None dataset.
+This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the dpo_mix_en and the bct_non_cot_dpo_1000 datasets.
 It achieves the following results on the evaluation set:
 - Loss: 0.8630
 - Rewards/chosen: -0.0798

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 2.9969690846635686,
+    "eval_logits/chosen": -2.8472585678100586,
+    "eval_logits/rejected": -2.8558220863342285,
+    "eval_logps/chosen": -0.7975095510482788,
+    "eval_logps/rejected": -1.0328320264816284,
+    "eval_loss": 0.8629826903343201,
+    "eval_odds_ratio_loss": 0.6547309160232544,
+    "eval_rewards/accuracies": 0.5618181824684143,
+    "eval_rewards/chosen": -0.07975095510482788,
+    "eval_rewards/margins": 0.02353225089609623,
+    "eval_rewards/rejected": -0.10328320413827896,
+    "eval_runtime": 194.6972,
+    "eval_samples_per_second": 5.65,
+    "eval_sft_loss": 0.7975095510482788,
+    "eval_steps_per_second": 2.825,
+    "total_flos": 2.1013894560546816e+18,
+    "train_loss": 0.9013287582572352,
+    "train_runtime": 18144.1457,
+    "train_samples_per_second": 1.637,
+    "train_steps_per_second": 0.102
+}

eval_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 2.9969690846635686,
+    "eval_logits/chosen": -2.8472585678100586,
+    "eval_logits/rejected": -2.8558220863342285,
+    "eval_logps/chosen": -0.7975095510482788,
+    "eval_logps/rejected": -1.0328320264816284,
+    "eval_loss": 0.8629826903343201,
+    "eval_odds_ratio_loss": 0.6547309160232544,
+    "eval_rewards/accuracies": 0.5618181824684143,
+    "eval_rewards/chosen": -0.07975095510482788,
+    "eval_rewards/margins": 0.02353225089609623,
+    "eval_rewards/rejected": -0.10328320413827896,
+    "eval_runtime": 194.6972,
+    "eval_samples_per_second": 5.65,
+    "eval_sft_loss": 0.7975095510482788,
+    "eval_steps_per_second": 2.825
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9969690846635686,
+    "total_flos": 2.1013894560546816e+18,
+    "train_loss": 0.9013287582572352,
+    "train_runtime": 18144.1457,
+    "train_samples_per_second": 1.637,
+    "train_steps_per_second": 0.102
+}