File size: 4,027 Bytes
0d3cf39
 
 
b7ea01b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
---
license: apache-2.0
---

**English** | [中文](./README_zh.md)

## Code implementation of new GTE embeddings

This model is a BERT-like encoder with the following optimizations implemented:

1. Replacing absolute position embeddings with RoPE [^1].
2. Substituting the conventional activation functions with Gated Linear Units (GLU) [^2].
3. Setting attention dropout to 0 to use `xformers` and `flash_attn`.
4. Using unpadding to eliminate the needless computations for padding tokens [^3]. (this is off by default and should be used in conjunction with `xformers` for optimal acceleration).
5. Setting `vocab_size` as a multiple of 64.

### Recommendation: Enable Unpadding and Acceleration with `xformers`

This code supports the acceleration of attention computations using `xformers`, which can automatically choose the optimal implementation based on the type of device, such as `flash_attn`. Therefore, we can also achieve significant acceleration on old devices like the V100.


Firstly, install `xformers` (with `pytorch` pre-installed):
```
if pytorch is installed using conda:
    conda install xformers -c xformers
elif pytorch is installed using pip:
    # cuda 11.8 version
    pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu118
    # cuda 12.1 version
    pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu121
```
For more information, refer to [Installing xformers](https://github.com/facebookresearch/xformers?tab=readme-ov-file#installing-xformers).

Then, when loading the model, set `unpad_inputs` and `use_memory_efficient_attention` to `true`, and enable `fp16` mixed precision computation to achieve the fastest acceleration.

```python
import torch
from transformers import AutoModel, AutoTokenizer

path = 'Alibaba-NLP/gte-base-en-v1.5'
device = torch.device('cuda')
tokenzier = AutoTokenizer.from_pretrained(path)
model = AutoModel.from_pretrained(
    path,
    trust_remote_code=True,
    unpad_inputs=True,
    use_memory_efficient_attention=True,
).to(device)

with torch.autocast(device_type=device.type, dtype=torch.float16):  # or bfloat16
    with torch.inference_mode():
        outputs = model(**inputs.to(device))

```

Alternatively, you can directly modify the `unpad_inputs` and `use_memory_efficient_attention` settings to `true` in the model's `config.json`, eliminating the need to set them in the code.


---

<details>
  <summary> Clarification of Relationship with nomic-embed and nomicBERT </summary>

One may question the originality of our work and consider it a mere replication of `nomicBERT`. To clarify, our work is parallel but stems from the same idea as `nomicBERT`.

Applying RoPE and GLU to BERT to support longer texts is a straightforward idea. Our exploration of the transformer++ encoder (i.e., BERT + RoPE + GLU) began in August 2023.
And by November 2023, we had completed the `gte-base-en-v1.1`. Then, I went on to prepare for the ACL submission of the other project...

The release of `nomic-embed` [^4] brought to our attention the pressure, as well as provided us with more resources, which allowed us to continue with this project.
Without the outstanding work of `nomicai`, the release of `gte-v1.5` could have been delayed much longer. Thanks!

</details>

---

[^1]: Su, Jianlin, Murtadha Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. "Roformer: Enhanced transformer with rotary position embedding." Neurocomputing 568 (2024): 127063.

[^2]: Shazeer, Noam. "Glu variants improve transformer." arXiv preprint arXiv:2002.05202 (2020).

[^3]: Portes, Jacob, Alexander Trott, Sam Havens, Daniel King, Abhinav Venigalla, Moin Nadeem, Nikhil Sardana, Daya Khudia, and Jonathan Frankle. "MosaicBERT: A Bidirectional Encoder Optimized for Fast Pretraining." Advances in Neural Information Processing Systems 36 (2024).

[^4]: Nussbaum, Zach, John X. Morris, Brandon Duderstadt, and Andriy Mulyar. "Nomic Embed: Training a Reproducible Long Context Text Embedder." arXiv preprint arXiv:2402.01613 (2024).