bug fix : ASR_model = openai/whisper-tiny, csv_transcript, csv_result values were hardcoded, making them query string dependent inside generateTranscript

ASR_Server.py

@@ -11,7 +11,7 @@ from utils.generate_box_plot import box_plot_data
 
 # Set the cache directory for Hugging Face datasets
 os.environ["HF_HOME"] = "/tmp/huggingface"
-ASR_model = "openai/whisper-tiny"  # Replace with your ASR model
+# ASR_model = "openai/whisper-tiny"  # Replace with your ASR model
 #Check cpu score
 import timeit
 cpu_score = timeit.timeit("sum(range(1000000))", number=5)
@@ -27,8 +27,8 @@ job_status = {
 }
 
 csv_path = "test.csv"
-csv_transcript = f'test_with_{ASR_model.replace("/", "_")}.csv'
-csv_result = f'test_with_{ASR_model.replace("/","_")}_WER.csv'
+# csv_transcript = f'test_with_{ASR_model.replace("/", "_")}.csv'
+# csv_result = f'test_with_{ASR_model.replace("/","_")}_WER.csv'
 df = pd.read_csv(csv_path)
 print(f"CSV Loaded with {len(df)} rows")
 
@@ -54,12 +54,16 @@ def generateTranscript(ASR_model):
         "total": None
     })
 
+    csv_transcript = f'test_with_{ASR_model.replace("/", "_")}.csv'
+    csv_result = f'test_with_{ASR_model.replace("/","_")}_WER.csv'
     # Check if transcript already exists
     df_transcript = download_csv(csv_transcript)
     if(df_transcript is None):
         print(f"CSV not found in the dataset repo. Proceeding to generate transcript.")
     else:
         print(f"Transcript already exists for model {ASR_model}. Skipping transcription.")
+        job_status["running"] = False
+        job_status["message"] = "Transcription Already existss"
         return
 
     # # Load test.csv
@@ -170,7 +174,7 @@ def asr_models():
     ]
 
     def background_job():
-        generateTranscript(ASR_model)
+        generateTranscript("openai/whisper-tiny")
 
     # Start the background job in a separate thread
     threading.Thread(target=background_job).start()