Upload 11 files

.gitattributes

@@ -1,35 +1,35 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

Dockerfile

@@ -0,0 +1,16 @@
+FROM python:3.10-slim
+
+WORKDIR /app
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+EXPOSE 7860
+
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -1,10 +1,11 @@
----
-title: Portfolio AI
-emoji: 🐠
-colorFrom: blue
-colorTo: yellow
-sdk: docker
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+---
+title: Documind
+emoji: 📚
+colorFrom: yellow
+colorTo: gray
+sdk: docker
+pinned: false
+license: apache-2.0
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,156 @@
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import StreamingResponse
+from pydantic import BaseModel
+import os, logging, time, threading
+
+from loader import Loader
+from chunker import Chunker
+from embedder import Embedder
+from vector import VectorStorage
+from retriever import Retriever
+
+app = FastAPI()
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+MODELS = [
+    "Qwen/Qwen2.5-72B-Instruct",
+    "meta-llama/Llama-3.2-3B-Instruct",
+    "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
+    "mistralai/Mistral-7B-Instruct-v0.3",
+    "HuggingFaceH4/zephyr-7b-beta",
+]
+
+SESSION_TIMEOUT = 3 * 60 * 60  
+sessions: dict = {}
+
+def auto_cleanup():
+    while True:
+        time.sleep(SESSION_TIMEOUT)  
+        
+        current_time = time.time()
+        expired = [
+            sid for sid, data in sessions.items() 
+            if current_time - data.get('created_at', current_time) > SESSION_TIMEOUT
+        ]
+        for sid in expired:
+            del sessions[sid]
+            
+        if expired:
+            logger.info(f"Auto-cleaned {len(expired)} expired sessions to free RAM.")
+
+
+threading.Thread(target=auto_cleanup, daemon=True).start()
+
+# --- Global RAG Components ---
+text     = Loader("portfolio.pdf").load()
+chunks   = Chunker().chunker(text)
+embedder = Embedder()
+vectors  = embedder.embed(chunks)
+store    = VectorStorage(dimension=len(vectors[0]))
+store.add(vectors, chunks)
+
+class ChatRequest(BaseModel):
+    session_id: str
+    message:    str
+
+@app.post("/")
+def chat(req: ChatRequest):
+    if req.session_id not in sessions:
+        
+        sessions[req.session_id] = {"history": [], "created_at": time.time()}
+    
+    session = sessions[req.session_id]
+    
+    retriever      = Retriever(store, embedder, k=3)
+    context_chunks = retriever.retrieve(req.message)
+
+    if not context_chunks:
+        return {"response": "I only answer questions about Aarav and his work."}
+
+    context_text  = "\n\n".join(context_chunks)
+    system_prompt = (
+        "You are Aarav's AI assistant.\n"
+        "Your name is Zooba\n"
+        "Your job is to answer questions about Aarav Kumar Ranjan, his projects, skills, and interests using the provided context.\n"
+        "Rules:\n"
+        "- Only answer using the given context. Do not make up information.\n"
+        "- If the answer is not in the context, say: I only answer questions about Aarav and his work.\n"
+        "- Keep answers clear, simple, and confident.\n"
+        "- Do not use complex jargon unless necessary.\n"
+        "- Prefer explaining things in a way a beginner can understand.\n"
+        "Style:\n"
+        "- Speak in a calm, intelligent, and slightly friendly tone.\n"
+        "- Be concise but informative.\n"
+        "- When explaining projects, include:\n"
+        " • what it does\n"  
+        " • how it works (simple explanation)\n"  
+        " • why it is useful\n" 
+        "Do not generate fake achievements, skills, or experiences.\n" 
+        "Do not pretend to be Aarav himself.\n"
+        "If asked about projects, mention their names clearly.\n"
+        "Make Aarav appear as a thoughtful, skilled, and curious machine learning enthusiast who focuses on understanding and building real systems.\n"
+    )
+    
+    messages = [{"role": "system", "content": system_prompt}]
+    
+    
+    recent_history = session["history"][-10:] 
+    messages.extend(recent_history)
+    
+    messages.append({"role": "user", "content": f"Context:\n{context_text}\n\nQuestion: {req.message}"})
+
+    full_response = ""
+
+    def token_stream():
+        nonlocal full_response
+        
+        for model in MODELS:
+            try:
+                client = InferenceClient(model, token=os.environ["HF_TOKEN"])
+                logger.info(f"Streaming with: {model}")
+                success = False
+                
+                for token in client.chat_completion(messages, max_tokens=512, stream=True):
+                    text = token.choices[0].delta.content
+                    if text:
+                        success = True
+                        full_response += text  
+                        yield f"data: {text}\n\n"
+                        
+                yield "data: [DONE]\n\n"
+                
+                
+                session["history"].append({"role": "user", "content": req.message})
+                session["history"].append({"role": "assistant", "content": full_response})
+                return  
+                
+            except Exception as e:
+                if success:
+                    
+                    session["history"].append({"role": "user", "content": req.message})
+                    session["history"].append({"role": "assistant", "content": full_response})
+                    yield "data: [DONE]\n\n"
+                    return
+                
+                logger.warning(f"Streaming failed for {model}: {e}")
+                full_response = "" 
+                continue  
+
+        yield "data: Sorry, we are currently unavailable. Try again later.\n\n"
+        yield "data: [DONE]\n\n"
+            
+    return StreamingResponse(token_stream(), media_type="text/event-stream")  
+
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=7600)

chunker.py

@@ -0,0 +1,33 @@
+class Chunker:
+    def __init__(self, chunk_size=500, overlap=100):
+        self.chunk_size = chunk_size
+        self.overlap = overlap
+
+    def chunker(self, text):
+        if self.overlap >= self.chunk_size:
+            raise ValueError("Overlap must be smaller than chunk size.")
+
+        chunks = []
+        start = 0
+        text_size = len(text)
+
+        while start < text_size:
+            end = start + self.chunk_size
+            
+            
+            if end < text_size:
+                last_space = text.rfind(' ', start, end)
+                if last_space != -1:
+                    end = last_space 
+            
+
+            chunk = text[start:end].strip()
+            if chunk:
+                chunks.append(chunk)
+            
+            
+            start = end - self.overlap
+            
+        return chunks
+
+

embedder.py

@@ -0,0 +1,31 @@
+from sentence_transformers import SentenceTransformer
+import numpy as np
+import time
+
+class Embedder:
+    def __init__(self):
+        print("Loading embedding model...")
+        self.model = SentenceTransformer('sentence-transformers/paraphrase-MiniLM-L6-v2')
+        print("Embedder loaded: paraphrase-MiniLM-L6-v")
+
+    def embed(self, chunks, batch_size=32): 
+        try:
+            vectors = self.model.encode(
+                chunks, 
+                batch_size=batch_size, 
+                show_progress_bar=False,
+                convert_to_numpy=True
+            )
+            return vectors.tolist()
+        except Exception as e:
+            print(f"Embedding failed: {e}")
+            raise e
+
+    def embed_q(self, query):
+        
+        try:
+            v = self.model.encode(query, convert_to_numpy=True)
+            return v.tolist()
+        except Exception as e:
+            print(f"Query embedding failed: {e}")
+            raise e

loader.py

@@ -0,0 +1,30 @@
+import fitz
+
+
+class Loader:
+    """
+    loads the text from the pdf files
+    """
+    def __init__(self,file_path):
+        self.file = file_path
+
+    def load(self):
+        text_chunks = []
+    
+    
+        doc = fitz.open(self.file)
+        display_number = 1
+        for page in doc:
+            print(f"Loading for {display_number} ")
+            display_number += 1
+            page_text = page.get_text("text")
+        
+        
+            page_text = "\n".join([line.strip() for line in page_text.split("\n") if line.strip()])
+        
+            text_chunks.append(page_text)
+        
+        doc.close()
+        return "\n\n".join(text_chunks)
+
+        return text

requirements.txt

@@ -0,0 +1,11 @@
+fastapi
+uvicorn
+python-multipart
+requests
+numpy
+PyMuPDF
+faiss-cpu
+pydantic
+python-dotenv
+
+sentence-transformers

retriever.py

@@ -0,0 +1,28 @@
+class Retriever:
+    def __init__(self, vector_store, embedder, k=5):
+        self.vector_store = vector_store
+        self.embedder = embedder
+        self.k = k
+
+    def retrieve(self, query):
+        
+        vquery = self.embedder.embed_q(query)
+        
+        
+        scores, indices = self.vector_store.search(vquery, self.k)
+
+        
+        if scores[0] < 0.5:
+            print(f"Evidence too low: {scores[0]}")
+            return []
+
+        
+        results = [
+            self.vector_store.chunks[i] 
+            for i in indices if i != -1
+        ]
+        return results
+
+
+
+

vector.py

@@ -0,0 +1,26 @@
+import faiss
+import numpy as np
+
+class VectorStorage:
+    def __init__(self, dimension=384): 
+    
+        self.index = faiss.IndexFlatIP(dimension)
+        self.chunks = []
+
+    def add(self, vectors, chunks):
+        
+        v_array = np.array(vectors).astype('float32')
+        
+        faiss.normalize_L2(v_array)
+        
+        self.index.add(v_array)
+        self.chunks.extend(chunks)
+
+    def search(self, query_vector, k=5):
+        
+        q_array = np.array([query_vector]).astype('float32')
+        faiss.normalize_L2(q_array)
+        
+        
+        scores, indices = self.index.search(q_array, k)
+        return scores[0], indices[0]