End of training

README.md

@@ -1,10 +1,11 @@
 ---
 base_model: mistralai/Mistral-7B-v0.3
 datasets:
-- generator
+- llama-duo/synth_classification_dataset_dedup
 library_name: peft
 license: apache-2.0
 tags:
+- alignment-handbook
 - trl
 - sft
 - generated_from_trainer
@@ -18,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # mistral-7b-0.3-gpt4o_100k_classification-lora
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the generator dataset.
+This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the llama-duo/synth_classification_dataset_dedup dataset.
 It achieves the following results on the evaluation set:
 - Loss: 1.4824
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 1.0,
+    "eval_loss": 1.4824345111846924,
+    "eval_runtime": 0.9072,
+    "eval_samples": 16,
+    "eval_samples_per_second": 2.205,
+    "eval_steps_per_second": 1.102,
     "total_flos": 3.7501312597753856e+17,
     "train_loss": 0.975214945736216,
     "train_runtime": 1874.6509,

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "eval_loss": 1.4816477298736572,
-    "eval_runtime": 1.519,
+    "eval_loss": 1.4824345111846924,
+    "eval_runtime": 0.9072,
     "eval_samples": 16,
-    "eval_samples_per_second": 1.317,
-    "eval_steps_per_second": 0.658
+    "eval_samples_per_second": 2.205,
+    "eval_steps_per_second": 1.102
 }

runs/Sep29_05-51-01_637888cb4368/events.out.tfevents.1727591057.637888cb4368.957079.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f491d8f5be7fdeb7b8bf0d8d63751383173568e30bc587441b129e313fed2e5
+size 359