docs(model-providers): replace hand-maintained Cohere model lists and fix related drift - #516
Open
g-despot wants to merge 1 commit into
Open
docs(model-providers): replace hand-maintained Cohere model lists and fix related drift#516g-despot wants to merge 1 commit into
g-despot wants to merge 1 commit into
Conversation
… fix related drift Core stopped validating model names for the Cohere modules in v1.33.0 (backported to v1.31.17 and v1.32.10), so any model string is forwarded verbatim and the static lists on those pages are guaranteed to rot. Replace each list with the no-validation statement, the server default, and a link to Cohere's own model documentation. Also corrected while in these pages: - openai/embeddings.md said you can use any OpenAI embedding model, but text2vec-openai still validates against an allowlist. The page also never stated its server default (text-embedding-3-small), and its "(default)" markers refer to dimensions, not model choice. - openai/generative.md's list is Weaviate's token-limit table, not an availability list. Kept and demoted, with a lead-in saying what it does: it caps maxTokens for those models. - return_likelihoods dropped from four snippets. The Python client accepts and silently discards it, and core has no reader on any path. - Six Python/TypeScript snippet pairs showed different values for the same example (dimensions, the multimodal model, and one model left active in TypeScript while commented in Python). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017AryMRLot3XAaz2s44eNG6
![orca-security-eu[bot]](https://avatars.githubusercontent.com/in/276244?s=60&v=4){kind=small}
There was a problem hiding this comment.
Orca Security Scan Summary
| Status | Check | Issues by priority | |
|---|---|---|---|
 Passed |
Infrastructure as Code |  0  0  0  0 |
View in Orca |
| Passed |
SAST | 0 0 0 0 |
View in Orca |
| Passed |
Secrets | 0 0 0 0 |
View in Orca |
| Passed |
Vulnerabilities | 0 0 0 0 |
View in Orca |
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.
What's being changed:
Core stopped validating model names for the Cohere modules in
v1.33.0(backported tov1.31.17andv1.32.10), so the hand-maintained lists on those pages are guaranteed to rot. Each is replaced with the no-validation statement, the server default, and a link to Cohere's own model documentation.Four corrections found alongside:
openai/embeddings.mdsaid you can use any OpenAI embedding model. That is false —text2vec-openaistill validates against an allowlist (ent/class_settings.go:185-192). The page also never stated its server default (text-embedding-3-small), and its(default)markers refer to dimensions, not model choice.openai/generative.md's list is the token-limit table, not an availability list. Kept and demoted into a<details>, with a lead-in saying what it does: it capsmaxTokensfor those models.return_likelihoodsremoved from four snippets. The Python client accepts and silently discards it, and core has no reader on any path — so the snippets showed a live-looking option that does nothing.dimensions, the multimodal model, and one model left active in TypeScript while commented in Python).Type of change:
How has this been tested?
Not built locally. Marker names in the four
_includesfiles are byte-identical tomain(so no page lost a snippet),<details>/<Tabs>/<TabItem>pairing was re-checked on all six pages, both_includesPython files compile, and every#available-models/#vectorizer-parameters/#generative-parametersanchor still resolves. Relying on the deploy preview for the site build.