Upload folder using huggingface_hub

app.py

@@ -4,7 +4,9 @@ import json
 import os
 import requests
 from pypdf import PdfReader
+from pathlib import Path
 import gradio as gr
+import time
 
 
 load_dotenv(override=True)
@@ -73,26 +75,93 @@ tools = [{"type": "function", "function": record_user_details_json},
         {"type": "function", "function": record_unknown_question_json}]
 
 
+def load_pdf_with_cache(pdf_path: str, cache_dir="me/cache") -> str:
+    pdf_path = Path(pdf_path)
+    cache_dir = Path(cache_dir)
+    cache_dir.mkdir(parents=True, exist_ok=True)
+
+    cache_file = cache_dir / f"{pdf_path.stem}.txt"
+
+    # Use cached text if available
+    if cache_file.exists():
+        with open(cache_file, "r", encoding="utf-8") as f:
+            return f.read()
+
+    # Otherwise parse PDF (slow path)
+    reader = PdfReader(str(pdf_path))
+    text = ""
+
+    for page in reader.pages:
+        page_text = page.extract_text()
+        if page_text:
+            text += page_text + "\n"
+
+    # Save cache for future cold starts
+    with open(cache_file, "w", encoding="utf-8") as f:
+        f.write(text)
+
+    return text
+
+
+
 class Me:
 
     def __init__(self):
         self.openai = OpenAI()
+        # self.gemini = OpenAI(api_key=os.getenv("GOOGLE_API_KEY"), base_url="https://generativelanguage.googleapis.com/v1beta/openai/")
+        # self.groq = OpenAI(api_key=os.getenv("GROQ_API_KEY"), base_url="https://api.groq.com/openai/v1")
+        # self.groq_model_name = "openai/gpt-oss-120b"
+        self.cur_model = 'gpt'
         self.name = "Mohit Kumar"
-        reader = PdfReader("me/linkedin.pdf")
-        self.linkedin = ""
-        for page in reader.pages:
-            text = page.extract_text()
-            if text:
-                self.linkedin += text
+        self.linkedin = load_pdf_with_cache("me/linkedin.pdf")
+        # reader = PdfReader("me/linkedin.pdf")
+        # self.linkedin = ""
+        # for page in reader.pages:
+            # text = page.extract_text()
+            # if text:
+                # self.linkedin += text
         with open("me/summary.txt", "r", encoding="utf-8") as f:
             self.summary = f.read()
-        reader = PdfReader("me/mkt_v1_2pg.pdf")
-        self.resume = ""
-        for page in reader.pages:
-            text = page.extract_text()
-            if text:
-                self.resume += text
-
+        # reader = PdfReader("me/mkt_v1_2pg.pdf")
+        # self.resume = ""
+        # for page in reader.pages:
+        #     text = page.extract_text()
+        #     if text:
+        #         self.resume += text
+        self.resume = load_pdf_with_cache("me/mkt_v1_2pg.pdf")
+        print("Linkedin and resume loaded successfully.", flush=True)
+        self._system_prompt = self.system_prompt()
+
+
+
+    def stream_llm(self, messages):
+        """
+        Tries Groq streaming first, falls back to OpenAI streaming
+        """
+        if self.cur_model == 'groq':
+            try:
+                return self.groq.chat.completions.create(
+                    model=self.groq_model_name,
+                    messages=messages,
+                    # tools=tools,
+                    stream=True,
+                )
+            except Exception as e:
+                print("Groq streaming failed:", e, flush=True)
+                return self.openai.chat.completions.create(
+                    model="gpt-5-nano",
+                    messages=messages,
+                    # tools=tools,
+                    stream=True,
+                )
+        else:
+            print("self.cur_model changed. Using gpt nano for streaming.", flush=True)
+            return self.openai.chat.completions.create(
+                    model="gpt-5-nano",
+                    messages=messages,
+                    # tools=tools,
+                    stream=True,
+                )
 
     def handle_tool_call(self, tool_calls):
         results = []
@@ -121,10 +190,27 @@ Be professional and engaging, as if talking to a potential client or future empl
         return system_prompt
     
     def chat(self, message, history):
-        messages = [{"role": "system", "content": self.system_prompt()}] + history + [{"role": "user", "content": message}]
+        messages = [{"role": "system", "content": self._system_prompt}] + history + [{"role": "user", "content": message}]
+        # if not history:
+        #     # messages = [{"role": "system", "content": self.system_prompt()}] + history + [{"role": "user", "content": message}]
+        #     messages = [{"role": "system", "content": self.system_prompt()}]
+        # else:
+        #     messages = []
+        # messages += history
+        # messages.append({"role": "user", "content": message})
         done = False
+        
         while not done:
-            response = self.openai.chat.completions.create(model="gpt-5-mini", messages=messages, tools=tools)
+            response = self.openai.chat.completions.create(model="gpt-5-nano", messages=messages, tools=tools)
+            # try:
+            #     # response = self.gemini.chat.completions.create(model="gemini-2.5-flash", messages=messages, tools=tools)
+            #     response = self.groq.chat.completions.create(model=self.groq_model_name, messages=messages, tools=tools)
+            #     print("Groq successful")
+            # except Exception as e:
+            #     print("Groq failed:", e, flush=True)
+            #     self.cur_model = 'openai'
+            #     response = self.openai.chat.completions.create(model="gpt-5-nano", messages=messages, tools=tools)
+            #     print("GPT 5 nano successful.", flush=True)
             if response.choices[0].finish_reason=="tool_calls":
                 message = response.choices[0].message
                 tool_calls = message.tool_calls
@@ -134,6 +220,16 @@ Be professional and engaging, as if talking to a potential client or future empl
             else:
                 done = True
         return response.choices[0].message.content
+        # stream = self.stream_llm(messages)
+
+        # partial = ""
+        # for chunk in stream:
+        #     delta = chunk.choices[0].delta
+
+        #     if delta and delta.content:
+        #         partial += delta.content
+        #         yield partial
+        #         time.sleep(0.01)
     
 
 if __name__ == "__main__":