End of training

README.md

@@ -2,9 +2,10 @@
 license: gemma
 library_name: peft
 tags:
+- llama-factory
+- lora
 - trl
 - dpo
-- llama-factory
 - generated_from_trainer
 base_model: google/gemma-7b-it
 model-index:
@@ -17,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Gemma-7B-It-ORPO-SALT-HALF
 
-This model is a fine-tuned version of [google/gemma-7b-it](https://huggingface.co/google/gemma-7b-it) on the None dataset.
+This model is a fine-tuned version of [google/gemma-7b-it](https://huggingface.co/google/gemma-7b-it) on the dpo_mix_en and the bct_non_cot_dpo_500 datasets.
 It achieves the following results on the evaluation set:
 - Loss: 1.3159
 - Rewards/chosen: -0.1249

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 2.9974597798475866,
+    "eval_logits/chosen": 253.5438995361328,
+    "eval_logits/rejected": 253.86448669433594,
+    "eval_logps/chosen": -1.2487919330596924,
+    "eval_logps/rejected": -1.470909833908081,
+    "eval_loss": 1.3159173727035522,
+    "eval_odds_ratio_loss": 0.6712530851364136,
+    "eval_rewards/accuracies": 0.561904788017273,
+    "eval_rewards/chosen": -0.12487921118736267,
+    "eval_rewards/margins": 0.02221176214516163,
+    "eval_rewards/rejected": -0.14709095656871796,
+    "eval_runtime": 222.5011,
+    "eval_samples_per_second": 4.719,
+    "eval_sft_loss": 1.2487919330596924,
+    "eval_steps_per_second": 2.36,
+    "total_flos": 2.135641671600046e+18,
+    "train_loss": 1.3953795637788071,
+    "train_runtime": 23493.5825,
+    "train_samples_per_second": 1.206,
+    "train_steps_per_second": 0.075
+}

eval_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 2.9974597798475866,
+    "eval_logits/chosen": 253.5438995361328,
+    "eval_logits/rejected": 253.86448669433594,
+    "eval_logps/chosen": -1.2487919330596924,
+    "eval_logps/rejected": -1.470909833908081,
+    "eval_loss": 1.3159173727035522,
+    "eval_odds_ratio_loss": 0.6712530851364136,
+    "eval_rewards/accuracies": 0.561904788017273,
+    "eval_rewards/chosen": -0.12487921118736267,
+    "eval_rewards/margins": 0.02221176214516163,
+    "eval_rewards/rejected": -0.14709095656871796,
+    "eval_runtime": 222.5011,
+    "eval_samples_per_second": 4.719,
+    "eval_sft_loss": 1.2487919330596924,
+    "eval_steps_per_second": 2.36
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9974597798475866,
+    "total_flos": 2.135641671600046e+18,
+    "train_loss": 1.3953795637788071,
+    "train_runtime": 23493.5825,
+    "train_samples_per_second": 1.206,
+    "train_steps_per_second": 0.075
+}