Add support for KoboldCpp embeddings in Vector Storage (#3795)

* Add support for KoboldCpp embeddings in vector processing

* Add validation for KoboldCpp embeddings to handle empty data

* Improve toast handling
This commit is contained in:
Cohee
2025-04-01 21:21:29 +03:00
committed by GitHub
parent 9c4404cae9
commit 80e821d12d
4 changed files with 113 additions and 1 deletions

View File

@@ -13,6 +13,7 @@
<option value="cohere">Cohere</option>
<option value="extras">Extras (deprecated)</option>
<option value="palm">Google AI Studio</option>
<option value="koboldcpp">KoboldCpp</option>
<option value="llamacpp">llama.cpp</option>
<option value="transformers" data-i18n="Local (Transformers)">Local (Transformers)</option>
<option value="mistral">MistralAI</option>
@@ -55,6 +56,14 @@
Hint: Set the URL in the API connection settings.
</i>
</div>
<div class="flex-container flexFlowColumn" id="koboldcpp_vectorsModel">
<span>
Set the KoboldCpp URL in the Text Completion API connection settings.
</span>
<span>
Must use version 1.87 or higher and have an embedding model loaded.
</span>
</div>
<div class="flex-container flexFlowColumn" id="llamacpp_vectorsModel">
<span data-i18n="The server MUST be started with the --embedding flag to use this feature!">
The server MUST be started with the <code>--embedding</code> flag to use this feature!