Upload 5 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* 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
+faiss_index/faiss_index.faiss filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,16 @@
+import streamlit as st
+from rag_utils_unsloth import load_faiss_index, get_embedding_model, query_index, generate_answer
+
+st.title("🎓 EduPilot (Unsloth + Mistral 4bit CPU)")
+
+index, documents = load_faiss_index()
+model_embed = get_embedding_model()
+
+user_input = st.text_input("Pose ta question ici :")
+
+if user_input:
+    top_docs = query_index(user_input, index, documents, model_embed)
+    context = "\n".join(top_docs)
+    response = generate_answer(user_input, context)
+    st.markdown("### ✨ Réponse du chatbot :")
+    st.write(response)

faiss_index/documents.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cc6100b51468166d2e0b5e0ca119f239e648cb5d539dd256dc886ef39f45f46
+size 36366182

faiss_index/faiss_index.faiss

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43efadcf8c063cedf2414c75df1dd801404c0ba263b0e35de8f21c25436d0694
+size 165167661

rag_utils_unsloth.py

@@ -0,0 +1,42 @@
+import faiss
+import pickle
+import numpy as np
+import torch
+import os
+
+from sentence_transformers import SentenceTransformer
+from unsloth import FastLanguageModel
+
+def load_faiss_index(index_path="faiss_index/faiss_index.faiss", doc_path="faiss_index/documents.pkl"):
+    index = faiss.read_index(index_path)
+    with open(doc_path, "rb") as f:
+        documents = pickle.load(f)
+    return index, documents
+
+def get_embedding_model():
+    return SentenceTransformer("sentence-transformers/multi-qa-MiniLM-L6-cos-v1")
+
+def query_index(question, index, documents, model, k=3):
+    question_embedding = model.encode([question])
+    _, indices = index.search(np.array(question_embedding).astype("float32"), k)
+    results = [documents[i] for i in indices[0]]
+    return results
+
+def generate_answer(question, context):
+    model_id = "unsloth/mistral-7b-instruct-v0.1-bnb-4bit"
+
+    model, tokenizer = FastLanguageModel.from_pretrained(
+        model_name=model_id,
+        max_seq_length=4096,
+        dtype="float32",            # pour CPU uniquement
+        load_in_4bit=True,
+        device_map="auto"
+    )
+
+    tokenizer.pad_token = tokenizer.eos_token
+
+    prompt = f"Voici un contexte :\n{context}\n\nQuestion : {question}\nRéponse :"
+    inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True).to(model.device)
+
+    outputs = model.generate(**inputs, max_new_tokens=256, pad_token_id=tokenizer.eos_token_id)
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)

requirements.txt

@@ -0,0 +1,5 @@
+streamlit
+faiss-cpu
+sentence-transformers
+unsloth
+torch