Instructions to use mlx-community/Mixtral-8x7B-Instruct-v0.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/Mixtral-8x7B-Instruct-v0.1 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Mixtral-8x7B-Instruct-v0.1 mlx-community/Mixtral-8x7B-Instruct-v0.1
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Mixtral went stupid/lazy and disobeyed the instructions when it couldn't print all outputs at one time
#12
by yinzhu - opened
When I was asking Mixtral to compose some video scripts (around 100 scripts) according to a specific structure, with the settings max tokens 1024 and Mixtral can output 7 scripts each time, the first batch of 7 scripts went well with the structure I instructed, but since the second batch:
- It started disobeying the structure.
- It became stupid (maybe lazy) because it couldn't understand some ordinary knowledge.
How should I resolve this?

