How to use pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # if on a CUDA device, also pip install mlx[cuda] # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit") prompt = "Once upon a time in" text = generate(model, tokenizer, prompt=prompt, verbose=True)
How to use pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit with MLX LM:
# Install MLX LM uv tool install mlx-lm # Generate some text mlx_lm.generate --model "pipenetwork/DeepSeek-V4.1-Flash-MLX-mixed-4_8bit" --prompt "Once upon a time"
I think that by "sudo sysctl iogpu.wired_limit_mb=516096" the engram-6 build (which is 477 GB and better) could work on the 512GB machine. It would be great to try it.
· Sign up or log in to comment
