ggml
sql
spider
text-to-sql
sql finetune
Eval Results
richardr1126's picture
Update README.md
2b64934
|
raw
history blame
4.65 kB
---
tags:
- sql
- spider
- text-to-sql
- sql finetune
model-index:
- name: spider-skeleton-wizard-coder-ggml
results:
- task:
type: text-generation
dataset:
type: richardr1126/spider-context-validation
name: Spider Dev w/ Database Context
metrics:
- name: Test Suite SQL Eval - Execution Accuracy
type: exec
value: 0.61
verified: false
- task:
type: text-generation
dataset:
type: richardr1126/spider-context-validation
name: Spider Dev w/ Database Context
metrics:
- name: Test Suite SQL Eval - Exact Matching Accuracy
type: exact
value: 0.568
verified: false
datasets:
- spider
- richardr1126/spider-skeleton-context-instruct
library_name: ggml
license: bigcode-openrail-m
---
### Spider Skeleton Wizard Coder GGML Summary
- This model was created by finetuning [WizardLM/WizardCoder-15B-V1.0](https://huggingface.co/WizardLM/WizardCoder-15B-V1.0) on an enhanced Spider context training dataset: [richardr1126/spider-skeleton-context-instruct](https://huggingface.co/datasets/richardr1126/spider-skeleton-context-instruct).
- Finetuning was performed using QLoRa on 3x RTX6000 48GB.
- This is the GGML version of the model.
- If you want just the QLoRa/LoRA adapter: [richardr1126/spider-skeleton-wizard-coder-qlora](https://huggingface.co/richardr1126/spider-skeleton-wizard-coder-qlora)
### Running the GGML model
- The best way to run this model is to use the [4-bit GGML version](https://huggingface.co/richardr1126/spider-skeleton-wizard-coder-ggml) on [koboldcpp](https://github.com/LostRuins/koboldcpp), with CuBlas support.
### Spider Dataset
[Spider](https://arxiv.org/abs/1809.08887) is a large-scale complex and cross-domain semantic parsing and text-to-SQL dataset annotated by 11 Yale students
The goal of the Spider challenge is to develop natural language interfaces to cross-domain databases.
This dataset was used to finetune this model.
### Spider Skeleton WizardCoder - [test-suite-sql-eval](https://github.com/taoyds/test-suite-sql-eval) Results
With temperature set to 0.0, top_p set to 0.9, and top_k set to 0, the model achieves **61% execution accuracy** on the Spider dev set.
<img src="https://raw.githubusercontent.com/cuplv/text-to-sql-wizardcoder/main/eval/plots/spiderwizard-plus-chatgpt.svg" height="300">
<img src="https://raw.githubusercontent.com/cuplv/text-to-sql-wizardcoder/main/eval/plots/spiderwizard-vs-chatgpt.svg" height="300">
Note:
- ChatGPT was evaluated with the default hyperparameters and with the system message `You are a sophisticated AI assistant capable of converting text into SQL queries. You can only output SQL, don't add any other text.`
- Both models were evaluated with `--plug_value` in `evaluation.py` using the Spider dev set with database context.
- `--plug_value`: If set, the gold value will be plugged into the predicted query. This is suitable if your model does not predict values. This is set to `False` by default.
## Citations
```
@misc{luo2023wizardcoder,
title={WizardCoder: Empowering Code Large Language Models with Evol-Instruct},
author={Ziyang Luo and Can Xu and Pu Zhao and Qingfeng Sun and Xiubo Geng and Wenxiang Hu and Chongyang Tao and Jing Ma and Qingwei Lin and Daxin Jiang},
year={2023},
}
```
```
@article{yu2018spider,
title={Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-sql task},
author={Yu, Tao and Zhang, Rui and Yang, Kai and Yasunaga, Michihiro and Wang, Dongxu and Li, Zifan and Ma, James and Li, Irene and Yao, Qingning and Roman, Shanelle and others},
journal={arXiv preprint arXiv:1809.08887},
year={2018}
}
```
```
@article{dettmers2023qlora,
title={QLoRA: Efficient Finetuning of Quantized LLMs},
author={Dettmers, Tim and Pagnoni, Artidoro and Holtzman, Ari and Zettlemoyer, Luke},
journal={arXiv preprint arXiv:2305.14314},
year={2023}
}
```
## Disclaimer
The resources, including code, data, and model weights, associated with this project are restricted for academic research purposes only and cannot be used for commercial purposes. The content produced by any version of WizardCoder is influenced by uncontrollable variables such as randomness, and therefore, the accuracy of the output cannot be guaranteed by this project. This project does not accept any legal liability for the content of the model output, nor does it assume responsibility for any losses incurred due to the use of associated resources and output results.