Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,47 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
<p align="center" width="100%">
|
| 6 |
+
<img src="https://i.postimg.cc/MKmyP9wH/new-banner.png" width="80%" height="80%">
|
| 7 |
+
</p>
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
<div>
|
| 11 |
+
<div align="center">
|
| 12 |
+
<a href='https://brianboli.com/' target='_blank'>Bo Li*<sup>1</sup></a> 
|
| 13 |
+
<a href='https://zhangyuanhan-ai.github.io/' target='_blank'>Yuanhan Zhang*<sup>,1</sup></a> 
|
| 14 |
+
<a href='https://cliangyu.com/' target='_blank'>Liangyu Chen*<sup>,1</sup></a> 
|
| 15 |
+
<a href='https://king159.github.io/' target='_blank'>Jinghao Wang*<sup>,1</sup></a> 
|
| 16 |
+
<a href='https://pufanyi.github.io/' target='_blank'>Fanyi Pu*<sup>,1</sup></a> 
|
| 17 |
+
</br>
|
| 18 |
+
<a href='https://jingkang50.github.io/' target='_blank'>Jingkang Yang<sup>1</sup></a> 
|
| 19 |
+
<a href='https://chunyuan.li/' target='_blank'>Chunyuan Li<sup>2</sup></a> 
|
| 20 |
+
<a href='https://liuziwei7.github.io/' target='_blank'>Ziwei Liu<sup>1</sup></a>
|
| 21 |
+
</div>
|
| 22 |
+
<div>
|
| 23 |
+
<div align="center">
|
| 24 |
+
<sup>1</sup>S-Lab, Nanyang Technological University 
|
| 25 |
+
<sup>2</sup>Microsoft Research, Redmond
|
| 26 |
+
</div>
|
| 27 |
+
|
| 28 |
+
This weight is for **initilizing training for Otter-MPT7B**. It's directly converted from [Openflamingov2](https://huggingface.co/openflamingo/OpenFlamingo-9B-vitl-mpt7b), we added a `<answer>` token for Otter's downstream instruction tuning.
|
| 29 |
+
|
| 30 |
+
You can load and try this model using
|
| 31 |
+
```python
|
| 32 |
+
load_bit = "bf16"
|
| 33 |
+
precision = {}
|
| 34 |
+
if load_bit == "bf16":
|
| 35 |
+
precision["torch_dtype"] = torch.bfloat16
|
| 36 |
+
elif load_bit == "fp16":
|
| 37 |
+
precision["torch_dtype"] = torch.float16
|
| 38 |
+
elif load_bit == "fp32":
|
| 39 |
+
precision["torch_dtype"] = torch.float32
|
| 40 |
+
model = OtterForConditionalGeneration.from_pretrained("luodian/OTTER-9B-LA-InContext", device_map="sequential", **precision)
|
| 41 |
+
model.text_tokenizer.padding_side = "left"
|
| 42 |
+
tokenizer = model.text_tokenizer
|
| 43 |
+
image_processor = transformers.CLIPImageProcessor()
|
| 44 |
+
model.eval()
|
| 45 |
+
```
|
| 46 |
+
|
| 47 |
+
Leave us a message if you have any error or question. You can follow [Otter code](https://github.com/Luodian/Otter) (see training section) to further tune your model on top of it.
|