Repository navigation
fix(core): hide Bedrock bare IDs that require inference profiles - #48078
Merged
Merged
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
V2 sends Bedrock model IDs verbatim, but 15 bare catalog IDs are profile-only on AWS: selecting them always fails with ValidationException (on-demand throughput isn't supported, retry with an inference profile). V1 rewrites these to us./eu./etc. profiles at request time, so they must stay in models.dev — V2 instead disables them so they vanish from the picker/available/default resolution. Users pick the working prefixed twins.
The 15 (all verified: absent from on-demand foundation-model listings in us-east-1/us-east-2/us-west-2/eu-west-1/eu-central-1/ap-southeast-2, plus live Converse probes returning the profile-only error):
Single-region survivors (bare Llama3 in us-west-2/us-east-2, bare Qwen/DeepSeek-v3, bare Nova 1.x) were deliberately NOT included — a region-aware filter is the right fix for those and would wrongly kill them if blunt-dropped. This list is interim until per-entry source-region metadata lands.
Mechanics: enabled=false via the existing Bedrock catalog transform (pi-style exclusion, generalized from their 1 ID to the full verified set). Guarded by model.get so entries absent from the catalog are not recreated as skeletons. Explicit config/session refs still resolve with full metadata (and fail at AWS as before); user disabled:false can re-enable.
Tests: 2 new cases (disable-15 + keep controls incl. us. twin, bare Nova Micro, Mantle Astra; no skeleton creation). 16/16 pass, typecheck clean.