All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Add the possibility to customize the prompt template used to instruct the LLM to create code documentation.
Just pass
--template
to the CLI, ortemplate
to the Docai module.
- Support to generate documentation locally with Ollama.
- Support for using Groq for document generation. This allows users to choose groq for content generation by specifying
groq
as themodelProvider
.
- Support for using Mistral as an alternative to ChatGPT for document generation. This allows users to choose Mistral for content generation by specifying
mistral
as themodelProvider
.
- Breaking Change:
modelName
is now required. The previous default value ofgpt-4
has been removed to encourage users to explicitly specify the model they wish to use. - Breaking Change:
modelProvider
is required. Users must choose betweenmistral
andopenAI
as the model provider. This change aims to clarify the source of the language model being used and to offer more flexibility. - The
openAi
configuration has been deleted and replaced with a newllm
structure, which includesapiKey
for specifying your API key, andmodelProvider
andmodelName
for defining the model provider and model name, respectively. This approach streamlines the configuration for better consistency and ease of use.
- The
openAi
object is officially deprecated and replaced byllm
for better representation of the language model configuration.
- The
OPEN_API_KEY
environment variable has been renamed toAPI_KEY
for the CLI version, to standardize the naming of environment variables and clarify their use.