-
-
Thanks for your interesting work!
The [Reformer](https://arxiv.org/abs/2001.04451) uses RevNet in a clever way. They double the dimension of `x` such that for `x1,x2=split(x)` both `x1` and `x2` ha…
-
python: 3.7
transformers: 4.9.2
pytorch: 1.8.1
```python
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained("huawei-noah/TinyBERT_4L_zh")
model = AutoM…
-
像是缺失了文件
Unrecognized model in D:\LIUGEGE\ComfyUI\models\Joy_caption_alpha\text_model. Should have a `model_type` key in its config.json, or contain one of the following strings in its name: albert, a…
-
## 🚀 Feature
We can reduce the number of parameters by sharing the linear projection layers for queries and keys in the nn.Transformer and nn.activation.MultiheadAttention modules.
I think we …
-
hi, I don't find the implementation of the LSH attention. Is it in progress?
thx
-
Hey,
Thank you for open-sourcing this super useful work! 🤗
I have a question - [Longformer](https://github.com/allenai/longformer) by AI2 supports up to 16K tokens and achieves SOTA on many long…
-
Hi Eric-
Thanks for the awesome work. I noticed that inside lib/fluent/plugin are three plugins. I would say `in_gelftail` and `out_gelf2file` can be yanked out of the repo for the following reasons:…
-
Hi,
I found that different hyper-parameters (number of layers, dimension, etc.) are used for different models.
Can you clarify how the baselines are compared?
For example,
https://github.c…
mlpen updated
3 years ago
-
Split Germanic (Saxon) religion from Norse Germanic and make give it enough flavor to actually be interesting to play.
- [ ] Split Saxon into its own religion away from Norse
- [ ] Make Saxon re…