Adding ccl_enabled flag during model loading and passing CCL lists during compilation process #623
+1,738
−136
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
vjanfaza
requested review from
ochougul,
quic-amitraj,
quic-hemagnih and
quic-rishinr
as code owners
Contributor
|
@vjanfaza , Can you please resolve the conflicts on the PR and run lint/format checks? |
Contributor
Author
I resolved the conflicts and pushed the changes. |
…ring compilation process Signed-off-by: Vahid Janfaza <[email protected]>
…ring compilation process Signed-off-by: Vahid Janfaza <[email protected]>
…ring compilation process Signed-off-by: Vahid Janfaza <[email protected]>
…ring compilation process Signed-off-by: Vahid Janfaza <[email protected]>
…ring compilation process Signed-off-by: Vahid Janfaza <[email protected]>
…27b.yaml Signed-off-by: vjanfaza <[email protected]>
…4b.yaml Signed-off-by: vjanfaza <[email protected]>
quic-mamta
reviewed
Nov 27, 2025
| comp_ctx_lengths_prefill = [256, 512, ctx_len] | ||
| comp_ctx_lengths_decode = [256, 512, ctx_len] | ||
| # In moe models when compiling with prefill_seq_len=1 and non-continuous-batching mode, prefill and decode will share the same ccl specializations. | ||
| comp_ctx_lengths_prefill = [256, 512, ctx_len] # None # |
Contributor
There was a problem hiding this comment.
nit; please remove the #None # at the end of this line from other places/files as well.
|
|
||
| model_name = "Qwen/Qwen3-30B-A3B-Instruct-2507" | ||
| """ | ||
| # For CB inference, set continuous_batching to True and add full_batch_size,mxfp6,mint8 argument in compile function |
Contributor
There was a problem hiding this comment.
nit, should be mxint8 not mint8
| comp_ctx_lengths_prefill=comp_ctx_lengths_prefill, | ||
| comp_ctx_lengths_decode=comp_ctx_lengths_decode, | ||
| ) | ||
| # mos=1, |
Contributor
There was a problem hiding this comment.
please remove this line.
| processor=processor, | ||
| images=image_urls, | ||
| generation_len=100, | ||
| device_ids=[28, 29, 30, 31], |
Contributor
There was a problem hiding this comment.
make these as [0,1,2,3]
| inputs["pixel_values"] = inputs["pixel_values"].to(torch.float32) | ||
| streamer = TextStreamer(tokenizer) | ||
| output = qeff_model.generate(inputs=inputs, device_ids=[0, 1, 2, 3], generation_len=100) | ||
| output = qeff_model.generate(inputs=inputs, device_ids=[8, 9, 10, 11], generation_len=100) |
Contributor
There was a problem hiding this comment.
this should be kept as original.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
In these changes, instead of passing CCL lists during model loading, I passed a flag called ccl_enabled to specify whether CCL feature is enabled or not and moved passing CCL lists to compilation process.