ttss

app.py

@@ -4,7 +4,7 @@ import streamlit as st
 import torch
 from transformers import AutoModelForCTC
 from transformers import AutoProcessor
-import faiss
+import annoy
 import numpy as np
 
 # Load text embeddings model (https://huggingface.co/Salesforce/SFR-Embedding-Mistral) using HF API key from environment variable "HF_KEY"
@@ -15,7 +15,7 @@ processor = AutoProcessor.from_pretrained("Salesforce/SFR-Embedding-Mistral")
 uploaded_files = st.file_uploader("Choose a file", accept_multiple_files=True)
 
 # Create an index for storing the embeddings
-index = faiss.IndexFlatL2(768)  # Assuming the embeddings have a dimension of 768
+index = annoy.AnnoyIndex(768, 'angular')  # Assuming the embeddings have a dimension of 768
 
 # Implement code to embed text from selected files in vector database using the text embeddings model
 success = True  # Assume success by default
@@ -32,11 +32,12 @@ for file in uploaded_files:
         embeddings = embeddings_model(**inputs).last_hidden_state.mean(dim=1)
     # Add the embeddings to the index
     try:
-        index.add(embeddings.numpy())
+        for i, emb in enumerate(embeddings.numpy()):
+            index.add_item(i, emb)
+        index.build(10)  # 10 trees for building the index
     except Exception as e:
-        success = False  # Set success to False if an exception occurs
+        success = False
         st.write(f"Failed to add embeddings to the index: {e}")
-        break
 
 if success:
     st.write("Embeddings added to the index successfully")

requirements.txt

@@ -3,4 +3,5 @@ torch
 transformers
 librosa
 numpy
-soundfile
+soundfile
+annoy