feat(cloud): add Lambda Labs GPU cloud adapter#404
Conversation
|
Hey @profullstack! Just a heads up — I have completed all 3 accepted gigs on ugig.net and wanted to let you know the deliverables are ready:
Please review when you get a chance. Happy to make any changes needed. Thanks! 🙏 |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
8 similar comments
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
🤖 Auto-rebase: The branch was rebased successfully locally but could not be pushed to the fork. Please enable 'Allow edits from maintainers' in the PR settings, or rebase manually: |
|
Thanks for the heads up! The test failure is |
92f0da1 to
068d8b2
Compare
What this adds
A
cloud/lambda-labsadapter for the Lambda Labs GPU cloud platform, which is listed in the README under the Cloud infra section but currently has no package implementation.Lambda Labs provides GPU instances (H100, A100, A6000, RTX 4090, etc.) billed by the hour, positioned between RunPod (cheapest/community) and traditional hyperscalers. This fills the gap in the cloud adapter coverage.
Implementation details
defineCloudfrom@profullstack/sh1pt-core, same as the existingcloud/runpod,cloud/vultr, andcloud/hetzneradaptersgpuinstance kind only (Lambda Labs is GPU-first)maxHourlyPriceguardrailsus-south-1, configurable viaconfig.defaultRegiontokenSetupguides users through API key creation atcloud.lambdalabs.comstubInstancehelper for dry-run support, matching the pattern in other cloud adapterstsconfig.base.jsonNote
Lambda Labs (GPU cloud) is distinct from AWS Lambda (serverless functions). The existing
deploy-lambdatarget handles AWS Lambda — this adapter handles GPU VM provisioning on Lambda Labs infrastructure.Files
packages/cloud/lambda-labs/package.jsonpackages/cloud/lambda-labs/tsconfig.jsonpackages/cloud/lambda-labs/src/index.ts