optimum
8e7588b0 - default to exllama when exllamav2 is disabled (#1494)

Commit
1 year ago
default to exllama when exllamav2 is disabled (#1494) * fix logic * simplify tests
Author
Parents
  • docs/source/llm_quantization/usage_guides
    • File
      quantization.mdx
  • optimum/gptq
    • File
      quantizer.py
  • tests/gptq
    • File
      test_quantization.py