Ask questions to your code base using the power of LLMs.
pip install git+https://github.com/abdeladim-s/askcodeaskcode --help
NAME
askcode - Chat with your code base with the power of LLMs.
SYNOPSIS
askcode <flags>
DESCRIPTION
Chat with your code base with the power of LLMs.
FLAGS
-c, --codebase_path=CODEBASE_PATH
Type: str
Default: '.'
path to your codebase
--language=LANGUAGE
Type: str
Default: 'python'
programming language ['python', 'javascript'] at the moment
-p, --parser_threshold=PARSER_THRESHOLD
Type: int
Default: 0
minimum lines needed to activate parsing (0 by default).
--text_splitter_chunk_size=TEXT_SPLITTER_CHUNK_SIZE
Type: int
Default: 256
Maximum size of chunks to return
--text_splitter_chunk_overlap=TEXT_SPLITTER_CHUNK_OVERLAP
Type: int
Default: 50
Overlap in characters between chunks
--use_HF=USE_HF
Type: bool
Default: True
use hugging face models, if False OpenAI models will be used
--llm_model=LLM_MODEL
Type: str
Default: 'TheBloke/CodeLlama-7B-...
Large language model name (HF model name or OpenAI model)
-e, --embeddings_model=EMBEDDINGS_MODEL
Type: str
Default: 'sentence-...
Embeddings model (HF model name or OpenAI model)
--retriever_search_type=RETRIEVER_SEARCH_TYPE
Type: str
Default: 'mmr'
Defines the type of search that the Retriever should perform. Can be "similarity" (default), "mmr", or "similarity_score_threshold".
--retriever_k=RETRIEVER_K
Type: int
Default: 4
Amount of documents to return (Default: 4)
-m, --max_new_tokens=MAX_NEW_TOKENS
Type: int
Default: 50
Maximum tokens to generate
--temperature=TEMPERATURE
Type: float
Default: 0.1
sampling temperature
--top_p=TOP_P
Type: float
Default: 0.9
sampling top_p
--repetition_penalty=REPETITION_PENALTY
Type: float
Default: 1.0
sampling repetition_penalty
--use_autogptq=USE_AUTOGPTQ
Type: bool
Default: True
Set it to True to use Quantized AutoGPTQ modelsMIT