Update app.py

app.py

@@ -9,6 +9,9 @@ import chromadb
 from datasets import load_dataset
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import gradio as gr
+# import packages load LLM model
+from gpt4all import GPT4All
+from pathlib import Path
 
 # Function to clear the cache
 def clear_cache(model_name):
@@ -36,14 +39,11 @@ class VectorStore:
     # Method to populate the vector store with embeddings from a dataset
     def populate_vectors(self, dataset, batch_size=100):
         # Use dataset streaming
-        dataset = load_dataset('Thefoodprocessor/recipe_new_with_features_full', split='train', streaming=True)
+        dataset = load_dataset('Thefoodprocessor/recipe_new_with_features_full', split='train[:1500]')
 
         # Process in batches
         texts = []
-        max_examples = 15
         for i, example in enumerate(dataset):
-            if i >= max_examples:
-                break
             title = example['title_cleaned']
             recipe = example['recipe_new']
             meal_type = example['meal_type']
@@ -79,9 +79,14 @@ vector_store.populate_vectors(dataset=None)
 
 # Load the model and tokenizer
 # text generation model
-model_name = "meta-llama/Meta-Llama-3-8B"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(model_name)
+# model_name = "meta-llama/Meta-Llama-3-8B"
+# tokenizer = AutoTokenizer.from_pretrained(model_name)
+# model = AutoModelForCausalLM.from_pretrained(model_name)
+
+# load model orca-mini general purpose model
+model_name = 'mistral-7b-openorca.gguf2.Q4_0.gguf'
+model_path = Path.home() / '.cache' / 'gpt4all'
+model = GPT4All(model_name=model_name, model_path=model_path)
 
 # Define the chatbot response function
 def chatbot_response(user_input):