github/DB-GPT
1
0
Fork 0
mirror of https://github.com/csunny/DB-GPT.git synced 2025-08-16 07:24:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(model): support Baichuan2 model

Browse Source
This commit is contained in:
FangYin Cheng 2023-09-06 15:39:40 +08:00
parent f12258bbac
commit 38ebd34d58
3 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
README.md
View File

@ -86,6 +86,7 @@ Currently, we have released multiple key features, which are listed below to dem
- Unified vector storage/indexing of knowledge base - Unified vector storage/indexing of knowledge base
- Support for unstructured data such as PDF, TXT, Markdown, CSV, DOC, PPT, and WebURL - Support for unstructured data such as PDF, TXT, Markdown, CSV, DOC, PPT, and WebURL
- Multi LLMs Support, Supports multiple large language models, currently supporting - Multi LLMs Support, Supports multiple large language models, currently supporting
- 🔥 Baichuan2(7b,13b)
- 🔥 Vicuna-v1.5(7b,13b) - 🔥 Vicuna-v1.5(7b,13b)
- 🔥 llama-2(7b,13b,70b) - 🔥 llama-2(7b,13b,70b)
- WizardLM-v1.2(13b) - WizardLM-v1.2(13b)

1
README.zh.md
View File

@ -116,6 +116,7 @@ DB-GPT 是一个开源的以数据库为基础的GPT实验项目，使用本地
- 非结构化数据支持包括PDF、MarkDown、CSV、WebURL - 非结构化数据支持包括PDF、MarkDown、CSV、WebURL
- 多模型支持 - 多模型支持
- 支持多种大语言模型, 当前已支持如下模型: - 支持多种大语言模型, 当前已支持如下模型:
- 🔥 Baichuan2(7b,13b)
- 🔥 Vicuna-v1.5(7b,13b) - 🔥 Vicuna-v1.5(7b,13b)
- 🔥 llama-2(7b,13b,70b) - 🔥 llama-2(7b,13b,70b)
- WizardLM-v1.2(13b) - WizardLM-v1.2(13b)

2
pilot/configs/model_config.py
View File

@ -77,6 +77,8 @@ LLM_MODEL_CONFIG = {
"baichuan-13b": os.path.join(MODEL_PATH, "Baichuan-13B-Chat"), "baichuan-13b": os.path.join(MODEL_PATH, "Baichuan-13B-Chat"),
# please rename "fireballoon/baichuan-vicuna-chinese-7b" to "baichuan-7b" # please rename "fireballoon/baichuan-vicuna-chinese-7b" to "baichuan-7b"
"baichuan-7b": os.path.join(MODEL_PATH, "baichuan-7b"), "baichuan-7b": os.path.join(MODEL_PATH, "baichuan-7b"),
"baichuan2-7b": os.path.join(MODEL_PATH, "Baichuan2-7B-Chat"),
"baichuan2-13b": os.path.join(MODEL_PATH, "Baichuan2-13B-Chat"),
# (Llama2 based) We only support WizardLM-13B-V1.2 for now, which is trained from Llama-2 13b, see https://huggingface.co/WizardLM/WizardLM-13B-V1.2 # (Llama2 based) We only support WizardLM-13B-V1.2 for now, which is trained from Llama-2 13b, see https://huggingface.co/WizardLM/WizardLM-13B-V1.2
"wizardlm-13b": os.path.join(MODEL_PATH, "WizardLM-13B-V1.2"), "wizardlm-13b": os.path.join(MODEL_PATH, "WizardLM-13B-V1.2"),
"llama-cpp": os.path.join(MODEL_PATH, "ggml-model-q4_0.bin"), "llama-cpp": os.path.join(MODEL_PATH, "ggml-model-q4_0.bin"),