Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

ImportError : REPO_ID_SEPARATOR is deprecated and private in transformers 4.20.0 #1599

Closed
ierezell opened this issue Jun 17, 2022 · 6 comments
Closed

ImportError : REPO_ID_SEPARATOR is deprecated and private in transformers 4.20.0 #1599

ierezell opened this issue Jun 17, 2022 · 6 comments

Comments

@ierezell
Copy link

ierezell commented Jun 17, 2022 •
edited
Loading

In sentence-transformers/utils.py the line from huggingface_hub import snapshot_download import REPO_ID_SEPARATOR

  1. the REPO_ID_SEPARATOR is not used in this file
  2. REPO_ID_SEPARATOR is not exported anymore (in latest transformer 4.20.0) and raise a ImportError at runtime which make all the code crash.

Solution :
Just remove the line from huggingface_hub import snapshot_download import REPO_ID_SEPARATOR
works on my side, I can make a PR but I guess it will be faster for you to respect your processes.

Thanks in advance,
Have a great day.
@ierezell ierezell changed the title REPO_ID_SEPARATOR is deprecated and private in transformers 4.20.0 ImportError : REPO_ID_SEPARATOR is deprecated and private in transformers 4.20.0 Jun 17, 2022
@nreimers
Copy link
Member

Thanks, it is already fixed and will soon be released

@MaartenGr MaartenGr mentioned this issue Jun 19, 2022
Closed
@GodsDusk
Copy link

now there is a import error:

from huggingface_hub.snapshot_download import REPO_ID_SEPARATOR
ImportError: cannot import name 'REPO_ID_SEPARATOR' from 'huggingface_hub.snapshot_download'
  • sentence-transformers 2.2.0
  • transformers 4.20.0
  • python 3.8 and 3.9

@rachelsng
Copy link

Hi, any resolution on this? Thanks!

@CaoHaiNam
Copy link

CaoHaiNam commented Jun 21, 2022 •
edited
Loading

I commented this line and it worked well.

@nreimers
Copy link
Member

I will push a new version today afternoon. Until then you can install it from source or downgrade huggingface_hub package

kdutia added a commit to climatepolicyradar/navigator-search-indexer that referenced this issue Jun 21, 2022
@kdutia
get tests working
ee2b553 
had to freeze huggingface_hub dependency. see UKPLab/sentence-transformers#1599
@nreimers
Copy link
Member

I uploaded the newest version 2.2.1 to Pypi. It should now work.

kdutia added a commit to climatepolicyradar/navigator-search-indexer that referenced this issue Sep 26, 2022
@kdutia
get tests working
6a50d05 
had to freeze huggingface_hub dependency. see UKPLab/sentence-transformers#1599
@Wauplin Wauplin mentioned this issue Nov 22, 2022
Closed
tomaarsen added a commit to tomaarsen/setfit that referenced this issue Dec 13, 2022
@tomaarsen
sentence-transformer must at least be version 2.2.1
3cbf541 
UKPLab/sentence-transformers#1599
lewtun pushed a commit to huggingface/setfit that referenced this issue Dec 14, 2022
@tomaarsen
Expand CI tests using matrix; make dependencies less restrictive; fix…
273244b 
… ONNX tests (#233)

* Add matrix to run CI on Python 3.7 through 3.10, Ubuntu & Windows

And to run tests both on the most recent dependencies and on the oldest legal dependencies

* Add dependency caching, install the right deps

and set fail-fast to false

* Upgrade CI actions

https://github.com/actions/checkout#checkout-v3
https://github.com/actions/setup-python

* Use single quotes in hashFiles call

* Use single quotes throughout the entire workflow

* sentence-transformer must at least be version 2.2.1

UKPLab/sentence-transformers#1599

* Move ONNX dummy inputs to the correct device

If not done, then the setfit model will be placed on the GPU while the inputs remain on the CPU. This caused test failures for me locally

* Resolve INVALID_ARGUMENT : Unexpected input data type. Actual: (tensor(int32)) , expected: (tensor(int64))

* Increment lowest datasets version to 2.3.0

As the sst2 dataset, commonly used in tests in setfit, was introduced in that version
(https://github.com/huggingface/datasets/releases/tag/2.3.0)
neofob pushed a commit to neofob/gamechanger-ml that referenced this issue Aug 26, 2023
Fix required package version
e932548 
lock.requirements.txt:
    * REPO_ID_SEPARATOR is deprecated so transformers and
      sentence-transformers need to be upgraded.

Reference: UKPLab/sentence-transformers#1599

Signed-off-by: Tuan T. Pham <[email protected]>
@neofob neofob mentioned this issue Aug 26, 2023
Closed
12 tasks
neofob pushed a commit to neofob/gamechanger-data that referenced this issue Aug 29, 2023
Fix pip package version with new gamechange-ml
a9ff530 
dev_tools/requirements/rhel8.locked.requirements.txt:
    * Update gamechanger-ml version to the latest v1.8.0
    * Update python package version for
        * gensim, huggingface-hub, protobuf, sentence-transformers,
          tokenizers, transformers

References:
    * UKPLab/sentence-transformers#1599
    * dod-advana/gamechanger-ml#153

Signed-off-by: Tuan T. Pham <[email protected]>
neofob pushed a commit to neofob/gamechanger-data that referenced this issue Aug 29, 2023
Fix pip package version with new gamechange-ml
93bb8e4 
dev_tools/requirements/rhel8.locked.requirements.txt:
    * Update gamechanger-ml version to the latest v1.8.0
    * Update python package version for
        * gensim, huggingface-hub, protobuf, sentence-transformers,
          tokenizers, transformers

References:
    * UKPLab/sentence-transformers#1599
    * dod-advana/gamechanger-ml#153

Signed-off-by: Tuan T. Pham <[email protected]>
neofob pushed a commit to neofob/gamechanger-data that referenced this issue Aug 31, 2023
Fix pip package version with new gamechange-ml
424fcd2 
dev_tools/requirements/rhel8.locked.requirements.txt:
    * Update gamechanger-ml version to the latest v1.8.0
    * Update python package version for
        * gensim, huggingface-hub, protobuf, sentence-transformers,
          tokenizers, transformers

References:
    * UKPLab/sentence-transformers#1599
    * dod-advana/gamechanger-ml#153

Signed-off-by: Tuan T. Pham <[email protected]>
neofob pushed a commit to neofob/gamechanger-data that referenced this issue Dec 7, 2023
Fix pip package version with new gamechange-ml
884d4f3 
dev_tools/requirements/rhel8.locked.requirements.txt:
    * Update gamechanger-ml version to the latest v1.8.0
    * Update python package version for
        * gensim, huggingface-hub, protobuf, sentence-transformers,
          tokenizers, transformers

References:
    * UKPLab/sentence-transformers#1599
    * dod-advana/gamechanger-ml#153

Optimize package building and use rhel8 requirements

* Update txtai pip version to solve conflict with gamechaner-ml

Note: gamechanger-ml v1.8.0 from github but it identifies itself
as v1.7.0.

```
570.8 The conflict is caused by:
570.8     The user requested sentence-transformers==2.2.2
570.8     txtai 2.0.0 depends on sentence-transformers>=0.4.0
570.8     gamechangerml 1.7.0 depends on sentence-transformers~=2.0.0
```

* Update torch version
* Update package versioning

Signed-off-by: Tuan T. Pham <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
5 participants
@nreimers @GodsDusk @ierezell @CaoHaiNam @rachelsng