oproj TP support acl graph#4073
Conversation
Signed-off-by: zzhx1 <zzh_201018@outlook.com>
|
👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:
If CI fails, you can run linting and testing checks locally according Contributing and Testing. |
There was a problem hiding this comment.
Code Review
This pull request aims to enable oproj tensor parallelism for ACL graphs. The change in vllm_ascend/ascend_config.py correctly modifies the condition to allow this feature in any graph mode by checking if eager mode is enforced, instead of checking for a specific graph implementation (torchair). This change is correct, safe, and effectively addresses the goal of the pull request. I have no issues to report.
|
@jianzs Please push this PR for merging. |
|
@wangxiyuan Please merge this PR, this is a small feature support, no other conflicts. |
### What this PR does / why we need it? Reference vllm-project#2167 and orpoj TP supports ACL graph. - vLLM version: v0.11.0 - vLLM main: vllm-project/vllm@83f478b --------- Signed-off-by: zzhx1 <zzh_201018@outlook.com> Signed-off-by: luolun <luolun1995@cmbchina.com>
### What this PR does / why we need it? Reference vllm-project#2167 and orpoj TP supports ACL graph. - vLLM version: v0.11.0 - vLLM main: vllm-project/vllm@83f478b --------- Signed-off-by: zzhx1 <zzh_201018@outlook.com> Signed-off-by: hwhaokun <haokun0405@163.com>
### What this PR does / why we need it? Reference vllm-project#2167 and orpoj TP supports ACL graph. - vLLM version: v0.11.0 - vLLM main: vllm-project/vllm@83f478b --------- Signed-off-by: zzhx1 <zzh_201018@outlook.com> Signed-off-by: nsdie <yeyifan@huawei.com>
### What this PR does / why we need it? Reference vllm-project#2167 and orpoj TP supports ACL graph. - vLLM version: v0.11.0 - vLLM main: vllm-project/vllm@83f478b --------- Signed-off-by: zzhx1 <zzh_201018@outlook.com>
What this PR does / why we need it?
Reference #2167 and orpoj TP supports ACL graph.