Commit
·
c7e7a9d
1
Parent(s):
f7e9f3e
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,36 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
+
<h1>
|
| 5 |
+
<a alt="Ask2Democracy project" href="https://github.com/jorge-henao/ask2democracy">Ask2Democracy project</a>
|
| 6 |
+
</h1>
|
| 7 |
+
<hr>
|
| 8 |
+
|
| 9 |
+
## What's baizemocracy-lora-7B-cfqa model?
|
| 10 |
+
|
| 11 |
+
This model is an open-source chat model fine-tuned with [LoRA](https://github.com/microsoft/LoRA) inspired by [Baize project](https://github.com/project-baize/baize-chatbot/tree/main/). It was trained with the Baize datasets and the ask2democracy-cfqa-salud-pension dataset, wich contains almost 4k instructions to answers questions based on a context relevant to citizen concerns and public debate in spanish.
|
| 12 |
+
|
| 13 |
+
- **Developed by:**
|
| 14 |
+
- 🇨🇴 [Jorge Henao](https://huggingface.co/jorge-henao)
|
| 15 |
+
- 🇨🇴 [David Torres ](https://github.com/datorresb)
|
| 16 |
+
|
| 17 |
+
## Training Parameters
|
| 18 |
+
|
| 19 |
+
- Base Model: [LLaMA-7B](https://arxiv.org/pdf/2302.13971.pdf)
|
| 20 |
+
- Training Epoch: 1
|
| 21 |
+
- Batch Size: 16
|
| 22 |
+
- Maximum Input Length: 512
|
| 23 |
+
- Learning Rate: 2e-4
|
| 24 |
+
- LoRA Rank: 8
|
| 25 |
+
- Updated Modules: All Linears
|
| 26 |
+
|
| 27 |
+
## Training Dataset
|
| 28 |
+
|
| 29 |
+
- [Ask2Democracy-cfqa-salud-pension](https://huggingface.co/datasets/hackathon-somos-nlp-2023/ask2democracy-cfqa-salud-pension) (3,806)
|
| 30 |
+
- [Standford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) (51,942)
|
| 31 |
+
- [Quora Dialogs](https://github.com/project-baize/baize) (54,456):
|
| 32 |
+
- [StackOverflow Dialogs](https://github.com/project-baize/baize) (57,046)
|
| 33 |
+
- [Alpacaca chat Dialogs](https://github.com/project-baize/baize)
|
| 34 |
+
- [Medical chat Dialogs](https://github.com/project-baize/baize)
|
| 35 |
+
|
| 36 |
+
More details can be found in the Ask2Democracy [GitHub](https://github.com/jorge-henao/ask2democracy)
|