Skip to content

Commit

Permalink
update readme
Browse files Browse the repository at this point in the history
  • Loading branch information
wangzaistone committed Sep 27, 2023

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
1 parent 5905887 commit 3e7ec80
Showing 2 changed files with 18 additions and 23 deletions.
22 changes: 10 additions & 12 deletions README.md
Original file line number Diff line number Diff line change
@@ -44,17 +44,15 @@ The following publicly available text-to-sql datasets are used for this project:

DB-GPT-Hub currently supports the following base models:

* LLaMa/LLaMa2
* CodeLlama
* alpaca
* vicuna
* guanaco

* Falcon
* BLOOM
* ChatGLM
* WizardLLM

- [x] CodeLlama
- [x] Baichuan2
- [x] LLaMa/LLaMa2
- [x] Falcon
- [x] Qwen
- [x] XVERSE
- [x] ChatGLM2
- [x] internlm

The approximate hardware resources required to quantize and fine-tune the model are as follows:

| Model Parameters | GPU RAM | CPU RAM | DISK |
@@ -194,7 +192,7 @@ You can find weights from huggingface. [hg-eosphoros-ai
To evaluate model performance on the dataset, default is spider dataset.
Run the following command:
```bash
python eval/evaluation.py --plug_value --input Your_model_pred_file
python dbgpt_hub/eval/evaluation.py --plug_value --input Your_model_pred_file
```
You can find the results of our latest review [here](docs/eval_llm_result.md)

19 changes: 8 additions & 11 deletions README.zh.md
Original file line number Diff line number Diff line change
@@ -42,16 +42,14 @@ DB-GPT-Hub是一个利用LLMs实现Text-to-SQL解析的实验项目,主要包

DB-GPT-HUB目前支持的base模型有:

* LLaMa/LLaMa2系列
* codeLlama
* alpaca
* vicuna
* guanaco

* Falcon系列
* BLOOM系列
* ChatGLM系列
* WizardLLM
- [x] CodeLlama
- [x] Baichuan2
- [x] LLaMa/LLaMa2
- [x] Falcon
- [x] Qwen
- [x] XVERSE
- [x] ChatGLM2
- [x] internlm

模型量化微调所需的硬件资源大概如下:

@@ -83,7 +81,6 @@ cd DB-GPT-Hub
conda create -n dbgpt_hub python=3.10
conda activate dbgpt_hub
pip install -r requirements.txt
mkdir model
```
你可以将下载的大模型文件放在新建model文件夹下面

0 comments on commit 3e7ec80

Please sign in to comment.