End of training

README.md

@@ -3,11 +3,15 @@ library_name: transformers
 license: apache-2.0
 base_model: mistralai/Mistral-7B-v0.3
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- generator
+- llama-duo/synth_closed_qa_dataset_dedup
 model-index:
 - name: mistral-7b-0.3-gpt4o_100k_closedqa-fft
   results: []
@@ -18,7 +22,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # mistral-7b-0.3-gpt4o_100k_closedqa-fft
 
-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_closed_qa_dataset_dedup dataset.
 It achieves the following results on the evaluation set:
 - Loss: 6.9495
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 0.9991416309012876,
+    "eval_loss": 6.949496746063232,
+    "eval_runtime": 1.8925,
+    "eval_samples": 15,
+    "eval_samples_per_second": 3.17,
+    "eval_steps_per_second": 0.528,
     "total_flos": 30438567444480.0,
     "train_loss": 6.936546556728402,
     "train_runtime": 8521.4757,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.45.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32770
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9991416309012876,
+    "eval_loss": 6.949496746063232,
+    "eval_runtime": 1.8925,
+    "eval_samples": 15,
+    "eval_samples_per_second": 3.17,
+    "eval_steps_per_second": 0.528
+}

runs/Sep28_15-37-37_637888cb4368/events.out.tfevents.1727547589.637888cb4368.549868.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c984346fd358121469540cae0af7d03dc79ce3694766d5b0acfbda25e950539c
+size 359