[WIP] Add support for custom DeepSeek modelling in ACL Graph mode #677
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Yizhou Liu <[email protected]>
yiz-liu
force-pushed
the
feat-deepseek-graph
branch
from
3b0c106 to
593b87e
Compare
|
|
||
| return final_hidden_states.view(num_tokens, hidden_dim) | ||
|
|
||
| def _forward(self, hidden_states: torch.Tensor) -> torch.Tensor: |
There was a problem hiding this comment.
Have we tested torchair on this, access attn_metadata in global context may cause some graph related issue, better confirmed it with torchair
|
|
||
| return output_padded | ||
| if trace_flag: | ||
| torch.ops.vllm.unified_ascend_mla_attention_with_output( |
There was a problem hiding this comment.
Can we use the attention interface unified_ascend_attention_with_output which already registered in attention_v1.py
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What this PR does / why we need it?
Enable ACL Graph mode for custom DeepSeek modelling.
Does this PR introduce any user-facing change?
None.
How was this patch tested?
Test it with any DeepSeek model.