Instead of fragile inline Dockerfile patching, just copy a modified utils.py (with _post_quant_fix call) into the image, same pattern as deepseek_v4.py and deepseek_v4_attention.py patches.
6 lines
47 B
Plaintext
6 lines
47 B
Plaintext
docker/
|
|
*.egg-info/
|
|
.git/
|
|
.gitignore
|
|
README.md
|