cleanup: remove unused non-streaming summarize_transcript function

summarize_transcript.py

@@ -32,43 +32,6 @@ def read_transcript(file_path):
         content = f.read()
     return content
 
-def summarize_transcript(llm, transcript, language='zh-TW'):
-    """Summarize the transcript using the loaded model."""
-    # Truncate the transcript to fit within the context window
-    # Account for the prompt tokens as well
-    max_transcript_length = 1000  # Leave room for prompt and response
-
-    if len(transcript) > max_transcript_length:
-        transcript = transcript[:max_transcript_length]
-        print(f"Transcript truncated to {max_transcript_length} characters to fit context window.")
-
-    # Use the model's chat format based on its template
-    if language == 'en':
-        messages = [
-            {"role": "system", "content": "You are a helpful assistant that summarizes transcripts."},
-            {"role": "user", "content": f"Please summarize the following transcript:\n\n{transcript}"}
-        ]
-    else:  # Default to zh-TW
-        messages = [
-            {"role": "system", "content": "你是一個有助的助手，負責總結轉錄內容。"},
-            {"role": "user", "content": f"請總結以下內容：\n\n{transcript}"}
-        ]
-
-    # Generate the summary using chat completion
-    output = llm.create_chat_completion(
-        messages=messages,
-        max_tokens=512,
-        temperature=0.3,
-        top_p=0.9,
-        repeat_penalty=1.1,
-        stop=["<|end_of_text|>", "<|eot_id|>", "<|eom_id|>"]
-    )
-
-    # Reset the model state to ensure clean state for next call
-    llm.reset()
-
-    return output['choices'][0]['message']['content'].strip()
-
 def stream_summarize_transcript(llm, transcript, language='zh-TW'):
     """
     Perform live streaming summary by getting real-time token output from the model.