I am trying to run GPT4All's embedding model on my M1 Macbook with the following code:
import json
import numpy as np
from gpt4all import GPT4All, Embed4All
# Load the cleaned JSON data
with open('coursesclean.json') as file:
data = json.load(file)
# Create an index and embeddings array
index = {}
embeddings = []
# Iterate over each course
for course_code, course_info in data.items():
course_name = course_info['course_name']
course_desc = course_info['course_desc']
text = f"{course_name} {course_desc}"
embedder = Embed4All()
embeddings_response = embedder.embed(text)
course_embeddings = np.array(embeddings_response)
# Store the embeddings in the array
embeddings.append(course_embeddings)
# Index the embeddings
index[course_code] = len(embeddings) - 1
# Convert the embeddings array to a NumPy array
embeddings_array = np.stack(embeddings)
save the index and embeddings array to NumPy files
np.save('embeddings.npy', embeddings_array)
with open('index.json', 'w') as file:
json.dump(index, file, indent=2)
However, I seem to be running into a Metal library conflict that puts me in an infinite loop of the Found model file statement.
Found model file at /Users/MY_USERNAME/.cache/gpt4all/ggml-all-MiniLM-L6-v2-f16.bin
objc[42338]: Class GGMLMetalClass is implemented in both /opt/homebrew/lib/python3.11/site-packages/gpt4all/llmodel_DO_NOT_MODIFY/build/libreplit-mainline-metal.dylib (0x11eb0c208) and /opt/homebrew/lib/python3.11/site-packages/gpt4all/llmodel_DO_NOT_MODIFY/build/libllamamodel-mainline-metal.dylib (0x11ef38208). One of the two will be used. Which one is undefined.
Found model file at /Users/MY_USERNAME//.cache/gpt4all/ggml-all-MiniLM-L6-v2-f16.bin
Found model file at /Users/MY_USERNAME//.cache/gpt4all/ggml-all-MiniLM-L6-v2-f16.bin
...
I've tried reinstalling GPT4All, cleared the model from cache and tried through Langchain but I'm having the same issue.
