EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[Model][2/N] Remove deepseek_mtp modeling. (#3561)

Browse Source 
This PR is step 2 of deepseek model refactoring and removes
deepseek_mtp.

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: whx-sjtu <2952154980@qq.com>
This commit is contained in:
whx
2025-10-21 20:17:09 +08:00
committed by GitHub
parent ffb42a8daa
commit 220df60c61
7 changed files with 38 additions and 422 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm_ascend/models/__init__.py
View File

@@ -33,10 +33,6 @@ def register_model():
"DeepseekV32ForCausalLM",
"vllm_ascend.models.deepseek_v3_2: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(