Conversation
Contributor
|
🌿 Preview your docs: https://nvidia-preview-flafrance-terminal-bench-4.docs.buildwithfern.com/nemo/gym Here are the markdown pages you've updated: |
lafrancef
force-pushed
the
flafrance/terminal-bench-4
branch
from
September 15, 2026 09:30
53ee407 to
8b41d2f
Compare
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
…acts Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Translate Harbor's static no-network policy into deny-all OpenSandbox egress for single-container environments. Keep public agent configuration intact and reject unsupported providers, Compose isolation, allowlists, and runtime policy switching. Validation: 75 Harbor integration tests passed; both affected TB4 tasks completed capped model runs and official separate verification without Harbor exceptions. Live probes confirmed hostname/direct-IP blocking and artifact transfer. Document the deployment's IPv6 loopback limitation. Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Pass Compose shared-memory sizes to OpenSandbox create metadata, preserve explicit root UID zero, and retain declared ports across sandbox serialization and reconnection. Validation: 113 targeted Compose, sandbox reconnect, and Harbor adapter tests passed. Ruff and diff checks passed. Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Add pinned TB4 resources with explicit session ownership, budgets, official verification, persisted retries, and borrowed sandbox handoffs. Run OpenCode and generic mini-SWE in their Gym agent servers, preserving existing clients. Add lifecycle, cancellation, artifact, MCP, and configuration coverage plus real smoke validation utilities and migration documentation. Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
Signed-off-by: Frederic Lafrance <flafrance@nvidia.com>
lafrancef
force-pushed
the
flafrance/terminal-bench-4
branch
from
September 15, 2026 22:40
8b41d2f to
ac92e3b
Compare
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.
This is based on #3332
It adds TB4 through the Harbor integration and includes several sub-changes to make all tasks run and be observable.
WIP:
Known limitations: