94dd832815
### What this PR does / why we need it? 1. Move additional functionalities from fused_moe.py to common_fused_moe.py and remove fused_moe.py 2. Remove unnecessary custom classes from qwen3_moe.py, and it will be completely removed after we release vllm-ascend v0.11.0 ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? Qwen3-30B-A3B/Qwen3-30B-A3B-W8A8/DeepSeek-V3-W4A8-Pruing/deepseek-mtp/pangu-pro-moe-pruing: 1. Enable/Disable EP 3. Aclgraph & eager 4. SP - vLLM version: v0.11.0 --------- Signed-off-by: Pr0Wh1teGivee <calvin_zhu0210@outlook.com> Co-authored-by: weijinqian0 <12153182+weijinqian0@users.noreply.github.com>
57 lines
2.1 KiB
Python
57 lines
2.1 KiB
Python
from vllm import ModelRegistry
|
|
|
|
import vllm_ascend.envs as envs_ascend
|
|
|
|
|
|
def register_model():
|
|
ModelRegistry.register_model(
|
|
"Qwen2VLForConditionalGeneration",
|
|
"vllm_ascend.models.qwen2_vl:AscendQwen2VLForConditionalGeneration")
|
|
|
|
ModelRegistry.register_model(
|
|
"Qwen3VLMoeForConditionalGeneration",
|
|
"vllm_ascend.models.qwen2_5_vl_without_padding:AscendQwen3VLMoeForConditionalGeneration"
|
|
)
|
|
|
|
ModelRegistry.register_model(
|
|
"Qwen3VLForConditionalGeneration",
|
|
"vllm_ascend.models.qwen2_5_vl_without_padding:AscendQwen3VLForConditionalGeneration"
|
|
)
|
|
|
|
if envs_ascend.USE_OPTIMIZED_MODEL:
|
|
ModelRegistry.register_model(
|
|
"Qwen2_5_VLForConditionalGeneration",
|
|
"vllm_ascend.models.qwen2_5_vl:AscendQwen2_5_VLForConditionalGeneration"
|
|
)
|
|
else:
|
|
ModelRegistry.register_model(
|
|
"Qwen2_5_VLForConditionalGeneration",
|
|
"vllm_ascend.models.qwen2_5_vl_without_padding:AscendQwen2_5_VLForConditionalGeneration_Without_Padding"
|
|
)
|
|
|
|
ModelRegistry.register_model(
|
|
"DeepseekV2ForCausalLM",
|
|
"vllm_ascend.models.deepseek_v2:CustomDeepseekV2ForCausalLM")
|
|
|
|
ModelRegistry.register_model(
|
|
"DeepseekV3ForCausalLM",
|
|
"vllm_ascend.models.deepseek_v2:CustomDeepseekV3ForCausalLM")
|
|
|
|
ModelRegistry.register_model(
|
|
"DeepseekV32ForCausalLM",
|
|
"vllm_ascend.models.deepseek_v2:CustomDeepseekV3ForCausalLM")
|
|
|
|
ModelRegistry.register_model(
|
|
"DeepSeekMTPModel",
|
|
"vllm_ascend.models.deepseek_mtp:CustomDeepSeekMTP")
|
|
|
|
# There is no PanguProMoEForCausalLM in vLLM, so we should register it before vLLM config initialization
|
|
# to make sure the model can be loaded correctly. This register step can be removed once vLLM support PanguProMoEForCausalLM.
|
|
ModelRegistry.register_model(
|
|
"PanguProMoEForCausalLM",
|
|
"vllm_ascend.torchair.models.torchair_pangu_moe:PanguProMoEForCausalLM"
|
|
)
|
|
ModelRegistry.register_model(
|
|
"Qwen3NextForCausalLM",
|
|
"vllm_ascend.models.qwen3_next:CustomQwen3NextForCausalLM")
|