Instructions to use sweepai/sweep-next-edit-v2-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Inference
mlx conversion possible?
#1
by adampotocki - opened
Could you provide MLX 8bit quant and MLX BF16 for this model?
Just download all files, instal mlx-lm ( pip install mlx-lm) and then:
- BF16
python -m mlx_lm.convert \
--hf-path /path/to/safetensors_folder \
--mlx-path /path/to/output_mlx_bf16
- 8-bit
python -m mlx_lm.convert \
--hf-path /path/to/safetensors_folder \
-q 8 \
--mlx-path /path/to/output_mlx_8bit