Zhaoyang Wang

Average the logits of 16 demonstrations

1

Hi! Really thanks for your work. And I have a question about prediction at inference with demonstrations. As the paper mentioned, the final prediction logits is by averaging the the...

enable truncate_head param when using normal prompt (w/o demo)

1

The ``truncate_head`` param of ``tokenize_multipart_input`` function is not enabled, when call it for normal prompt tuning (i.e., w/o demonstration). In fact, for some extreme long dataset, we have to manually...

Because return the not actually converted sentence may confuse someone that is careless about the supported locales. And I think raise an error early can stop future big error for...

Add one book and one course in English

2

## What does this PR do? Add resource(s) ## For resources ### Description The first one is a book for high-dimensional data analysis. ~~The second one is a course for...

Environment: VS2019 My code: BigInteger p = BigInteger(pp); vector fibo(t + 10); fibo[1] = 1; fibo[2] = 1; for (LL i = 3; i

resume_from_checkpoint is not working with Deepspeed

6

### System Info - `transformers` version: 4.26.1 - Platform: Linux-5.4.0-139-generic-x86_64-with-glibc2.17 - Python version: 3.8.16 - Huggingface_hub version: 0.12.1 - PyTorch version (GPU?): 1.13.1 (True) - Tensorflow version (GPU?): 2.7.0 (True)...

Fix the typing error of ``max_seq_length: Optional[str]`` of ``PromptDataLoader`` and ``ErniePromptDataLoader`` both of which should be ``int``.

It seems like the official development and maintaining of OpenAttack have stalled. Here are some useful resources including personal implemented function for collecting generated adversarial examples, and backup of some...