readme file loading#

evaluations/models.py

@@ -3,7 +3,7 @@ from huggingface_hub import InferenceClient
 import os
 
 system_messages = { "STRICT": "You are a chatbot evaluating github repositories, their python codes and corresponding readme files. Strictly answer the questions with Yes or No.",
-                    "HELP": "You are a chatbot evaluating github repositories, their python codes and corresponding readme files. Please help me answer the following question." }
+                    "HELP": "You are a chatbot evaluating github repositories, their python codes and corresponding readme files. Please help me answer the following question. Keep your answers short, and informative." }
 
 class LocalLLM():
   def __init__(self, model_name):
@@ -33,4 +33,4 @@ class RemoteLLM():
         model=self.model_name, max_tokens=500, stream=False,
         messages=[{"role": "system", "content": system_messages[response_type]}, 
                   {"role": "user", "content": prompt}])
-    return message['choices'][0]
+    return message['choices'][0]['message']['content']

evaluations/repo_evaluations.py

@@ -41,9 +41,9 @@ def evaluate(llm, verbose, repo_url, title=None, year=None):
 
       zip = zipfile.ZipFile(repository_zip_name)
       readme = fetch_readme(zip)
-      
-      if (llm):
-        summary = llm.predict("HELP", "{readme}\nBased on the readme file above can you give a quick summary of this repository?")
+
+      if ((readme != "") & llm):
+        summary = llm.predict("HELP", f"{readme}\nBased on the readme file above can you give a quick summary of this repository?")
         log(verbose, "LOG", f"Summary: {summary}")
 
       results["pred_stars"] = fetch_repo_stars(verbose, repo_url, token)