Repository navigation
Explain Kaggle runs that get no GPU and list live Kaggle kernels - #16
Merged
Merged
Conversation
The kernel driver now checks for the GPU it asked for and for internet access before running. Kaggle silently drops both on accounts without a verified phone number; the run now stops at once and its status says so instead of a bare exit code. Running instances also lists kernels that are queued or running on any saved Kaggle account, including ones orx no longer follows. Also formats src/local/skills.rs, which failed cargo fmt on main. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LnYzVVarTNKxtq2vcYnzmE
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Before: a Kaggle run on an account that is not phone-verified ran on CPU with no internet, failed while downloading packages, and showed only "Job failed: exited with code 2". Running instances listed only orx runs of the selected project, so Kaggle kernels orx no longer followed (or started elsewhere) were invisible.
After: the kernel stops right away when Kaggle withholds the GPU, and the run's status reads "Kaggle gave the kernel no GPU. Kaggle grants GPUs and internet access only to phone-verified accounts: verify this account at https://www.kaggle.com/settings and run again." A run that fails offline says it had no internet access. Running instances also shows kernels queued or running on any saved Kaggle account, with a link to each.
How: the driver checks
nvidia-smiand a TCP connection to PyPI/GitHub before extracting the snapshot and exits 98 without a GPU. The controller writes a one-linefailure_reasonnext toexit_code, whichlocalbox::inspect_jobnow uses as the error message when present. A new/api/settings/kaggle/kernelsendpoint checks the five most recently run kernels per account withkernels status(cached 30s), and the UI adds them to the Running instances table, skipping ones that already appear as orx runs.Also runs
cargo fmtonsrc/local/skills.rs, which madefmt, clippy, testfail on main.Checks:
cargo fmt --check,cargo clippy --all-targets -D warnings, Kaggle and localbox tests, i18n lint,tsc,pnpm build((void 0)()count 0). The driver was run locally without a GPU and exits 98 with the message.🤖 Generated with Claude Code
https://claude.ai/code/session_01LnYzVVarTNKxtq2vcYnzmE
Generated by Claude Code