search

carlrobertoh / CodeGPT

JetBrains extension providing access to state-of-the-art LLMs, such as GPT-4, Claude 3, Code Llama, and others, all for free
https://codegpt.ee
Apache License 2.0
912 stars 186 forks source link

fix: Storing HuggingFaceModel by modelName instead of quantization only #529

Closed PhilKes closed 2 months ago

PhilKes commented 2 months ago

This PR fixes how the HuggingFaceModel for Llama is stored. Scenario: Select CodeGemma 4-bit precision -> Save

CodeGPT_LlamaSettings.xml Before:

<!-- ambiguous -->
<option name="huggingFaceModel" value="4-bit precision" />

After:

<option name="huggingFaceModel" value="codegemma-7b-it-Q4_K_M.gguf" />

Also removed unnecessary CodeGemma and Phi-3 models