Subscribe
Sign in
Mistral Large 2 — Top-tier reasoning for high-complexity tasks
Compared to its predecessor, Mistral Large 2 is significantly more capable in code generation, mathematics, and reasoning. It also provides a much stronger multilingual support, and advanced function calling capabilities.
19
Replies
Chris Messina
Hunter
📌
Immediately following Meta's release of Meta Llama 3.1, Mistral is out with Mistral Large 2, with "a 128k context window [that] supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash."
Share
@chrismessina Congratulations on your product launch! I have used Mistral, it's really cool. I am surprise there is only so few upvote. Thank you for your hard work and dedication.
When trying out the snippet shown in the hugging face page it returns an error:
from huggingface_hub import InferenceClient
import os
client = InferenceClient(
"mistralai/Mistral-Large-Instruct-2407",
token=os.getenv("hf_token"),
)
for message in client.chat_completion(
messages=[{"role": "user", "content": "What is the capital of France?"}],
max_tokens=500,
stream=True,
):
print(message.choices[0].delta.content, end="")
403 Forbidden: None.
Cannot access content at: https://api-inference.huggingfac....
If you are trying to create or update content, make sure you have a token with the `write` role.
The model mistralai/Mistral-Large-Instruct-2407 is too large to be loaded automatically (245GB > 10GB). Please use Spaces (https://huggingface.co/spaces) or Inference Endpoints (https://huggingface.co/inference...).
Another gold 🥇 for France 🇫🇷
Coding with Mistral is a real pleasure... Would had been even better if open-sourced thought 🤔
How does it improve upon its predecessor in terms of multilingual support and function calling?
Congrats on this excellent release! It’s really made a difference for me, @arthurmensch
