Files
xc-llm-ascend/vllm_ascend/models/__init__.py
Mengqing Cao 344228a5da [deepseek][bugfix] support deepseek quant (#469)
- support deepseek quant
  - add w8a8_dynamic quant
see #391

Signed-off-by: MengqingCao <cmq0113@163.com>
Co-authored-by: zzzzwwjj <1183291235@qq.com>
2025-04-07 10:56:12 +08:00

18 lines
564 B
Python

from vllm import ModelRegistry
def register_model():
from .qwen2_vl import CustomQwen2VLForConditionalGeneration # noqa: F401
ModelRegistry.register_model(
"Qwen2VLForConditionalGeneration",
"vllm_ascend.models.qwen2_vl:CustomQwen2VLForConditionalGeneration")
ModelRegistry.register_model(
"DeepseekV2ForCausalLM",
"vllm_ascend.models.deepseek_v2:CustomDeepseekV2ForCausalLM")
ModelRegistry.register_model(
"DeepseekV3ForCausalLM",
"vllm_ascend.models.deepseek_v2:CustomDeepseekV3ForCausalLM")