gghfez/c4ai-command-a-03-2025-AWQ

⚠️ Compatibility Warning ⚠️

This quantized model has known issues with vLLM versions > 0.9.2 due to architecture compatibility problems.

Use this model instead: gghfez/command-a-03-2025-AWQ

Issues:

• Created before proper Cohere2 support in AWQ
• Uses legacy "Cohere" architecture workaround
• Breaks with newer vLLM versions

References:

• vLLM Issue #24038
• Discussion thread

Alternatives:

• Command-A Reasoning AWQ (The reasoning version of this model), see these working quants: 4-bit | 8-bit by cpatonn
• Command-A EXl3: ExLlamaV3 3.12bpw by Downtown-Case