⚗️ Questions Eval: externalize dataset name

arkhn · Oct 18, 2024 · 8db08ef · 8db08ef
1 parent 27ce804
commit 8db08ef
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 1 deletion.
diff --git a/lib/questions_eval/configs/run.yaml b/lib/questions_eval/configs/run.yaml
@@ -6,6 +6,7 @@ defaults:
 
 samples: 2
 num_questions: 5
+dataset: "DataFog/medical-transcription-instruct"
 
 prompts:
   transcription: >-

diff --git a/lib/questions_eval/run.py b/lib/questions_eval/run.py
@@ -227,7 +227,7 @@ def main(cfg: DictConfig):
     evaluation_chain = create_chain(cfg.prompts.evaluation, question_llm, False)
 
     # Load and process dataset
-    loaded_dataset = load_dataset("DataFog/medical-transcription-instruct", split="train")
+    loaded_dataset = load_dataset(cfg.dataset, split="train")
     df = loaded_dataset.to_pandas().iloc[: cfg.samples]
 
     tqdm.pandas(desc="Generating data...")