Refactor so that environment and engine #65
Merged
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
so that they dont depend on llama specific stuff such as ModelArgs
FanhaiLu1
reviewed
May 2, 2024
Collaborator
FanhaiLu1
left a comment
FanhaiLu1
left a comment
There was a problem hiding this comment.
For 'tokenize_and_pad' error, jetstream removed tokenize_and_pad in recent change, I added it back in PR AI-Hypercomputer/JetStream#70, it should pass after re-run the test.
| @property | ||
| def max_prefill_length(self) -> int: | ||
| return self.param.max_seq_len | ||
| return self.env.max_input_sequence_length |
Collaborator
There was a problem hiding this comment.
Great, it's much clear now. Engineer might need to figure out and do mapping max_seq_len is max_prefill_length
Collaborator
Author
I added a command in |
FanhaiLu1
approved these changes
May 2, 2024
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Outdated suggestions cannot be applied.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.
so that they dont depend on llama
specific stuff such as ModelArgs