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Outdated suggestions cannot be applied.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.
First change: Decode only the new tokens as they are generated during streaming instead of the entire sequence. This makes a difference when the context is large (like 40,000 tokens).
Here is a test with ExLlama_HF and a 13b model. The prompt is cached to make the difference more pronounced.
Before:
After:
Second change: Restrict the UI streaming updates to 5 per second when
--listen
or--share
are provided. I find that this makes the UI a lot more responsive over a network.