Add OrcaReplay to Benchmarks & Evaluation - #60
Open
xizhuomengcontin wants to merge 2 commits into
Open
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds OrcaReplay to Benchmarks & Evaluation.
The entries there measure AI coding capability across tasks. OrcaReplay measures it on your task: fork a recorded run from any checkpoint onto a different model, with every earlier turn still served from the recording, so both models face byte-identical context and file state. That is a head-to-head on the work you actually do, rather than on SWE-bench.
The recording side is what makes it possible. Capture sits at the process and socket boundary, so the model traffic, the shell commands with their exit codes, the per-turn file changes and the MCP calls all land on one timeline; replaying with the network off reproduces the run byte-for-byte, which turns an intermittent agent failure into something you can rerun.
Row matches the table's three columns (Tool / Type / Why It's Awesome). Apache-2.0,
npm i -g orcareplay, Node 20+. Works with Claude Code, Codex, opencode, Qwen Code, Cursor and others.