Open
Conversation
pzelasko
requested changes
Mar 5, 2026
| setup_speech_encoder(self, pretrained_weights=self.cfg.pretrained_weights) | ||
|
|
||
|
|
||
| # Load pretrained weights if provided |
Collaborator
There was a problem hiding this comment.
Why is this needed in the PR? Looks to me like a separate feature that we may want to implement in a generic way so that every speechlm2 model can re-use.
| if isinstance(self.modality_adapter, (QformerConnector, MultiLayerProjectionConnector)): | ||
| # Store as a plain Python attribute (not an nn.Module submodule) to avoid duplicate | ||
| # state_dict entries: encoder_multilayer.encoder IS self.encoder, so safetensors | ||
| # would de-duplicate them and produce spurious "missing key" warnings on reload. |
Collaborator
There was a problem hiding this comment.
It might be an overkill. Can we instead do:
# note: not self.encoder
encoder = self.from_config_dict(cfg.encoder)
if isinstance(self.modality_adapter, (QformerConnector, MultiLayerProjectionConnector)):
self.encoder_multilayer = ConformerMultiLayerFeatureExtractor(encoder, ...)
else:
self.encoder = encoder
pzelasko
reviewed
Mar 6, 2026
| missing_keys, unexpected_keys = self.load_state_dict(tensors, strict=False) | ||
| logging.warning(f"Missing keys: {missing_keys}") | ||
| logging.warning(f"Unexpected keys: {unexpected_keys}") | ||
| maybe_install_lora(self) |
Collaborator
There was a problem hiding this comment.
Can we move maybe_install_lora back to line 67?
Signed-off-by: Nune <ntadevosyan@nvidia.com>
Signed-off-by: ntadevosyan <ntadevosyan@nvidia.com>
nune-tadevosyan
force-pushed
the
projection_connector
branch
from
d48c89a to
acce061
Compare
Signed-off-by: nune-tadevosyan <nune-tadevosyan@users.noreply.github.com>
pzelasko
approved these changes
Mar 9, 2026
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
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.
Important
The
Update branchbutton must only be pressed in very rare occassions.An outdated branch is never blocking the merge of a PR.
Please reach out to the automation team before pressing that button.
What does this PR do ?
Adds
MultiLayerProjectionConnectorsupport for SALM models.Changelog
MultiLayerProjectionConnector.Usage
# Add a code snippet demonstrating how to use thisGitHub Actions CI
The Jenkins CI system has been replaced by GitHub Actions self-hosted runners.
The GitHub Actions CI will run automatically when the "Run CICD" label is added to the PR.
To re-run CI remove and add the label again.
To run CI on an untrusted fork, a NeMo user with write access must first click "Approve and run".
Before your PR is "Ready for review"
Pre checks:
PR Type:
If you haven't finished some of the above items you can still open "Draft" PR.
Who can review?
Anyone in the community is free to review the PR once the checks have passed.
Contributor guidelines contains specific people who can review PRs to various areas.
Additional Information