Llama 4 Support #1508
Draft
Llama 4 Support #1508
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: Qubitium <Qubitium@modelcloud.ai>
Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>
4 tasks
|
We are missing layers/modules. The quantized sized vs non-quantized size is only 6.2% which is not normal. At 4bit, it whould be closer to 70% difference. |
added 2 commits
April 6, 2025 20:35
|
Need to fix this. MoE modules are parameters and not linear so they are skipped for quantization. https://github.com/user-attachments/assets/ac3b3fd4-50eb-420e-929f-c6ae1f2c8b11 |
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Outdated suggestions cannot be applied.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.
Llama 4 attention mask/casual mask caculations have issues/bugs at the moment. We are bypassing them by forcing
batch=1, padding, and removingattention_masksalltogether, for now until the bugs are fixed upstream.