Update README.md
Browse files
README.md
CHANGED
|
@@ -10,12 +10,14 @@ pipeline_tag: text-generation
|
|
| 10 |
|
| 11 |
# Bespoke-MiniChart-7B
|
| 12 |
|
| 13 |
-
|
|
|
|
|
|
|
| 14 |
|
| 15 |
This is an open‑source chart‑understanding Vision‑Language Model (VLM) developed at [Bespoke Labs](https://www.bespokelabs.ai/) and maintained by [Liyan Tang](https://www.tangliyan.com/) and Bespoke Labs. It sets a new state‑of‑the‑art in chart question‑answering (Chart‑QA) for 7 billion‑parameter models, outperforming much larger closed models such as Gemini‑1.5‑Pro and Claude‑3.5 on seven public benchmarks.
|
| 16 |
|
| 17 |
1. **Blog Post**: https://www.bespokelabs.ai/blog/bespoke-minichart-7b
|
| 18 |
-
2. **
|
| 19 |
---
|
| 20 |
|
| 21 |
# Example Outputs
|
|
@@ -36,13 +38,23 @@ The examples below showcase how Bespoke-MiniChart-7B can perform both visual per
|
|
| 36 |
|
| 37 |
Bespoke-MiniChart-7B achieves state-of-the-art performance on chart understanding among models with similar sizes. In addition to that, the model can even surpass closed-models such as Gemini-1.5-Pro and Claude-3.5.
|
| 38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
<p align="left">
|
| 41 |
-
<img src="https://cdn-uploads.huggingface.co/production/uploads/6444e4417a7b94ddc2d14e1d/
|
| 42 |
</p>
|
| 43 |
|
|
|
|
| 44 |
# Model Use:
|
| 45 |
|
|
|
|
|
|
|
| 46 |
The model is available on the playground here: https://playground.bespokelabs.ai/minichart
|
| 47 |
|
| 48 |
You can also run the model with the following snippet:
|
|
|
|
| 10 |
|
| 11 |
# Bespoke-MiniChart-7B
|
| 12 |
|
| 13 |
+
<a href="https://playground.bespokelabs.ai/minichart">
|
| 14 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/6444e4417a7b94ddc2d14e1d/g-QaXrmPLYk5m3Hq5vFtr.png" width="200px" />
|
| 15 |
+
</a>
|
| 16 |
|
| 17 |
This is an open‑source chart‑understanding Vision‑Language Model (VLM) developed at [Bespoke Labs](https://www.bespokelabs.ai/) and maintained by [Liyan Tang](https://www.tangliyan.com/) and Bespoke Labs. It sets a new state‑of‑the‑art in chart question‑answering (Chart‑QA) for 7 billion‑parameter models, outperforming much larger closed models such as Gemini‑1.5‑Pro and Claude‑3.5 on seven public benchmarks.
|
| 18 |
|
| 19 |
1. **Blog Post**: https://www.bespokelabs.ai/blog/bespoke-minichart-7b
|
| 20 |
+
2. **Playground**: https://playground.bespokelabs.ai/minichart
|
| 21 |
---
|
| 22 |
|
| 23 |
# Example Outputs
|
|
|
|
| 38 |
|
| 39 |
Bespoke-MiniChart-7B achieves state-of-the-art performance on chart understanding among models with similar sizes. In addition to that, the model can even surpass closed-models such as Gemini-1.5-Pro and Claude-3.5.
|
| 40 |
|
| 41 |
+
<p align="left">
|
| 42 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/6444e4417a7b94ddc2d14e1d/5pejAyzPG_tRBU6FwH7PA.png" width="700">
|
| 43 |
+
</p>
|
| 44 |
+
|
| 45 |
+
We also compare the performance of our model finetuned using SFT+DPO vs SFT only.
|
| 46 |
+
|
| 47 |
+
In the table below, M1 and M2 are finetuned models with 270K and 1M SFT examples respsectively, and Bespoke-MiniChart-7B is the model finetuned using SFT+DPO.
|
| 48 |
|
| 49 |
<p align="left">
|
| 50 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/6444e4417a7b94ddc2d14e1d/WRsPs437niUrXmYtkRajG.png" width="700">
|
| 51 |
</p>
|
| 52 |
|
| 53 |
+
|
| 54 |
# Model Use:
|
| 55 |
|
| 56 |
+
[](https://colab.research.google.com/drive/1FEmlwGgn9209iQO-rs2-9UHPLoytwZMH?usp=sharing)
|
| 57 |
+
|
| 58 |
The model is available on the playground here: https://playground.bespokelabs.ai/minichart
|
| 59 |
|
| 60 |
You can also run the model with the following snippet:
|