You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When using the semantic or the hybrid search, hybrid.embedder is now a mandatory parameter in GET and POST /indexes/{:indexUid}/search
As a consequence, it is now mandatory to pass hybrid even for full-vector search (with only vector and not q)
embedder is now a mandatory parameter in GET and POST /indexes/{:indexUid}/similar
Ignore non-zero semanticRatio when vector is passed but not q: a semantic search will be performed.
The default model for OpenAI is now text-embedding-3-small instead of text-embedding-ada-002.
Changes:
A new sub setting in embedders setting to enable binary quantization and speed up indexing speed.
Limit the maximum length of a rendered document template: when the source of an embedder is set to huggingFace, openAi, rest or ollama, then documentTemplateMaxBytes is now available as an optional parameter. This parameter describes the number of bytes in which the rendered document template text should fit when trying to embed a document. Longer texts are truncated to fit.
TODO
Breaking changes section (see above)
Ensure the breaking changes are applied in the code base
Fix tests failing due to the of breaking changes
Ensure we can enable binary quantization: add the binaryQuantized to in the embedders settings (refer to usage page)
Ensure the documentTemplateMaxBytes parameter can be used with huggingFace, openAi, rest or ollama models
Add tests for the new added features
⚠️ Make PRs pointing to bump-meilisearch-v1.11.0 and NOT main. Please do 1 PR for all of these changes, and not several.
The text was updated successfully, but these errors were encountered:
Related to meilisearch/integration-guides#303
Explanation of the feature
Usage:
Breaking:
hybrid.embedder
is now a mandatory parameter inGET and POST /indexes/{:indexUid}/search
hybrid
even for full-vector search (with onlyvector
and notq
)embedder
is now a mandatory parameter inGET and POST /indexes/{:indexUid}/similar
semanticRatio
whenvector
is passed but notq
: a semantic search will be performed.text-embedding-3-small
instead oftext-embedding-ada-002
.Changes:
embedders
setting to enable binary quantization and speed up indexing speed.huggingFace
,openAi
,rest
orollama
, thendocumentTemplateMaxBytes
is now available as an optional parameter. This parameter describes the number of bytes in which the rendered document template text should fit when trying to embed a document. Longer texts are truncated to fit.TODO
binaryQuantized
to in theembedders
settings (refer to usage page)documentTemplateMaxBytes
parameter can be used withhuggingFace
,openAi
,rest
orollama
modelsbump-meilisearch-v1.11.0
and NOTmain
. Please do 1 PR for all of these changes, and not several.The text was updated successfully, but these errors were encountered: