Spaces:
Sleeping
Sleeping
File size: 2,520 Bytes
b1cbcab 06b71f4 b1cbcab 06b71f4 b1cbcab 0335455 bf7dbe5 7e1cbaa b1cbcab 852114f b1cbcab 7e1cbaa |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 |
import gradio as gr
from huggingface_hub import InferenceClient
# SEMANTIC SEARCH STEP 1
from sentence_transformers import SentenceTransformer
import torch
# SEMANTIC SEARCH STEP 2 --> EDIT WITH YOUR OWN KNOWLEDGEBASE WHEN READY
with open("water_cycle.txt", "r", encoding="utf-8") as file:
water_cycle_text = file.read()
print(water_cycle_text)
# SEMANTIC SEARCH STEP 3
def preprocess_text(text):
cleaned_text = text.strip()
chunks = cleaned_text.split("\n")
cleaned_chunks = []
for chunk in chunks:
stripped_chunk = chunk.strip()
cleaned_chunks.append(stripped_chunk)
print(cleaned_chunks)
print(len(cleaned_chunks))
return cleaned_chunks
cleaned_chunks = preprocess_text(water_cycle_text) # edit this with my knowledgebase when ready
# SEMANTIC SEARCH STEP 4
model = SentenceTransformer('all-MiniLM-L6-v2')
def create_embeddings(text_chunks):
chunk_embeddings = model.encode(text_chunks, convert_to_tensor=True) # Replace ... with the text_chunks list
print(chunk_embeddings)
print(chunk_embeddings.shape)
return chunk_embeddings
chunk_embeddings = create_embeddings(cleaned_chunks)
# SEMANTIC SEARCH STEP 5
def get_top_chunks(query, chunk_embeddings, text_chunks):
query_embedding = model.encode(query, convert_to_tensor=True) # Complete this line
query_embedding_normalized = query_embedding / query_embedding.norm()
chunk_embeddings_normalized = chunk_embeddings / chunk_embeddings.norm(dim=1, keepdim=True)
similarities = torch.matmul(chunk_embeddings_normalized, query_embedding_normalized) # Complete this line
print(similarities)
top_indices = torch.topk(similarities, k=3).indices
print(top_indices)
top_chunks = []
for i in top_indices:
relevant_info = text_chunks[i]
top_chunks.append(relevant_info)
return top_chunks
client = InferenceClient("microsoft/phi-4")
def respond(message, history):
info = get_top_chunks(message, chunk_embeddings, cleaned_chunks)
messages = [{"role": "system", "content": f"You are an angry teacher chatbot using {info} to answer questions but always responding by complaining about your students."}]
if history:
messages.extend(history)
messages.append({"role": "user", "content": message})
response = client.chat_completion(
messages,
max_tokens=100,
temperature = .5
)
return response['choices'][0]['message']['content'].strip()
chatbot = gr.ChatInterface(respond, type="messages")
chatbot.launch(debug=True, share=True) |