13 KiB
The tables below have been generated by calling cargo run --bin embedder_settings
List of the embedder settings
Setting | Description | Type | Default Value | Regenerate on Change |
---|---|---|---|---|
|
The source used to provide the embeddings. Which embedder parameters are available and mandatory is determined by the value of this setting. |
"openAi" | "huggingFace" | "userProvided" | "ollama" | "rest" | "composite" |
"openAi" |
🏗️ Always |
|
The name of the model to use. |
string |
|
🏗️ Always |
|
The revision (commit SHA1) of the model to use. If unspecified, Meilisearch picks the latest revision of the model. |
string |
|
🏗️ Always |
|
The pooling method to use. |
"useModel" | "forceCls" | "forceMean" |
"useModel" |
🏗️ Always |
|
The API key to pass to the remote embedder while making requests. |
string |
|
🌱 Never |
|
The expected dimensions of the embeddings produced by this embedder. |
number |
|
|
|
A liquid template used to render documents to a text that can be embedded. Meillisearch interpolates the template for each document and sends the resulting text to the embedder. The embedder then generates document vectors based on this text. |
string |
{% for field in fields %}{% if field.is_searchable and field.value != nil %}{{ field.name }}: {{ field.value }} {% endif %}{% endfor %} |
|
|
Rendered texts are truncated to this size before embedding. |
number |
400 |
|
|
URL to reach the remote embedder. |
string |
|
|
|
Template request to send to the remote embedder. |
any |
|
🏗️ Always |
|
Template response indicating how to find the embeddings in the response from the remote embedder. |
any |
|
🏗️ Always |
|
Additional headers to send to the remote embedder. |
object |
|
🌱 Never |
|
Embedder settings for the embedder used at search time. |
object |
|
🌱 Never |
|
Embedder settings for the embedder used at indexing time. |
object |
|
|
|
Affine transformation applied to the semantic score to make it more comparable to the ranking score. |
object |
|
🌱 Never |
|
Whether to binary quantize the embeddings of this embedder. Binary quantized embeddings are smaller than regular embeddings, which improves disk usage and retrieval speed, at the cost of relevancy. |
boolean |
|
|
Availability of the settings depending on the selected source
Setting |
openAi |
huggingFace |
ollama |
userProvided |
rest |
composite |
---|---|---|---|---|---|---|
|
✅ Allowed |
✅ Allowed |
🔐 Mandatory |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
|
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
|
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
|
✅ Allowed |
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
|
✅ Allowed |
🚫 Disallowed |
✅ Allowed |
🔐 Mandatory |
✅ Allowed |
🚫 Disallowed |
|
|
|
|
🚫 Disallowed |
|
🚫 Disallowed |
|
|
|
|
🚫 Disallowed |
|
🚫 Disallowed |
|
✅ Allowed |
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
🔐 Mandatory |
🚫 Disallowed |
|
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🔐 Mandatory |
🚫 Disallowed |
|
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🔐 Mandatory |
🚫 Disallowed |
|
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
✅ Allowed |
🚫 Disallowed |
|
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🔐 Mandatory |
|
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🚫 Disallowed |
🔐 Mandatory |
|
|
|
|
|
|
|
|
|
|
|
|
|
|