Alan Lee

@nateraw OK my fault. I use model_id `google/vit-base-patch16-224` instead of `google/vit-base-patch16-224-in21k`. Cound you please tell me the difference between them? I read the doc and still can't tell the difference.

@Zhang21 your solution works fine with me

Same. Here is my workaround. ```http POST /tokens/OAuth/2 HTTP/1.1 Host: https://accounts.accesscontrol.windows.net Content-Type: application/json { "grant_type":"client_credentials", "client_id":"", "client_secret":"", "resource":"" } ```

> @secsilm what's your problem when you use your original way? Backend got empty request body. But now it worked using my original way. Maybe there were some mistakes. Thanks.

I think it's ok to call "pooler" layer. This layer transforms the output shape of the Transformer from `[batch_size, seq_length, hidden_size]` to `[batch_size, hidden_size]`. This is similar to [GlobalMaxPool1D](https://www.tensorflow.org/api_docs/python/tf/keras/layers/GlobalMaxPool1D), but...

> @secsilm I understand that might be doing some kind of GlobalMaxPool1D, however do you know what exact algorithm they are using to reduce the dimension, I am afraid they...

> Hi, > > thanks for reporting. This is actually the expected behavior for `loky`. The `loky` backend rely on `spawn` to start the new processes for the pool of...

@Abhishek17w After changing css you should clean your browser cache. For example, you can press `F12` and select `Disable cache` in `Network`, then press `F5` refresh the page.

I have trained ner model with my data and got pretty good result. Not even using `[CLS]`.

看了看源码，加载的哪个模型要看 `seg_only` 的值了（默认 `False`）： - `True` -> `cws_model.bin`； - `False` -> `model_c_model.bin` 见 https://github.com/thunlp/THULAC-Python/blob/3f1f126cd92c3d2aebdf4ab4850de3c9428a3b66/thulac/__init__.py#L55-L66