-
Notifications
You must be signed in to change notification settings - Fork 31
⬆️ bump vllm upper bound to support 0.10.2 #463
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Signed-off-by: Prashant Gupta <[email protected]>
|
👋 Hi! Thank you for contributing to vLLM support on Spyre. Or this can be done with Now you are good to go 🚀 |
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
|
|
||
| with pytest.raises(BadRequestError, | ||
| match="This model's maximum context length is"): | ||
| with pytest.raises(BadRequestError, match="maximum context length is"): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a bug in vllm upstream - opened a PR vllm-project/vllm#24995
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Co-authored-by: Joe Runde <[email protected]> Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
maxdebayser
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The pooling correction works. So now we only need backwards compatibility tests for the cursor stuff.
Oh awesome! Do we want to review and get #468 in first so that we're not stuck in a state where we have to be backward compatible with 0.10.2 (which is going to be much harder vs 0.10.1.1) Edit: merged! |
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
|
Nope, all pass! Ready for review! |
|
bot:test |
1 similar comment
|
bot:test |
|
Some CI tests failing, looking... |
|
bot:test |
|
CI tests all pass! Time for a review |
rafvasq
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
just a typo
Signed-off-by: Prashant Gupta <[email protected]>
|
|
||
| if task == "embed": | ||
| self.pooler = Pooler.for_embed(pooler_config=pooler_config) | ||
| with set_current_vllm_config(self.vllm_config): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we need to set_current_vllm_config because now Pooler class needs the vllm config for it to read vllm_config.model_config.head_dtype
Signed-off-by: Max de Bayser <[email protected]>
Signed-off-by: Max de Bayser <[email protected]>
|
bot:test |
|
Merging as all non-main tests are passing |
Description
Bump upper bound of vllm to
0.10.2Related Issues