fix(provider): default top_p to 0.95 for deepseek-v4-flash - #40247
Open
hoooou wants to merge 2 commits into
Open
fix(provider): default top_p to 0.95 for deepseek-v4-flash#40247hoooou wants to merge 2 commits into
hoooou wants to merge 2 commits into
Conversation
github-actions
Bot
added
needs:compliance
Contributor
|
Thanks for your contribution! This PR doesn't have a linked issue. All PRs must reference an existing issue. Please:
See CONTRIBUTING.md for details. |
github-actions
Bot
removed
the
needs:compliance
Contributor
|
Thanks for updating your PR! It now meets our contributing guidelines. 👍 |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Issue for this PR
Closes #40249
Type of change
What does this PR do?
opencode currently sends deepseek-v4-flash requests without a top_p, so the API default of 1.0 is used. DeepSeek recommends top_p = 0.95 (with temperature = 1.0) for agentic scenarios with this model, and OpenCode is a coding agent (see https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731).
This adds a top_p = 0.95 default in the sampling defaults, matching the existing treatment of minimax-m2 and the kimi-k2.5 family. The match is scoped to deepseek-v4-flash only so other deepseek models (deepseek-chat / reasoner / v3 / v4-pro) are unaffected.
How did you verify your code works?
Added a unit test in packages/opencode/test/provider/transform.test.ts asserting topP() returns 0.95 for deepseek-v4-flash (temperature/topK stay unset).
bun test test/provider/transform.test.tspasses: 398 tests, 0 failures. Prettier check passes on both changed files.Screenshots / recordings
N/A, not a UI change.
Checklist