Transformers 4.44 support #1996
Conversation
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
IlyasMoutawwakil
force-pushed
the
transformers-4.44
branch
from
41c81d5 to
ea84dde
Compare
IlyasMoutawwakil
requested review from
dacorvo,
tengomucho and
michaelbenayoun
| @@ -28,7 +28,7 @@ | |||
|
|
|||
| class TokenClassificationProcessing(TaskProcessor): | |||
| ACCEPTED_PREPROCESSOR_CLASSES = (PreTrainedTokenizerBase,) | |||
| DEFAULT_DATASET_ARGS = "conll2003" | |||
| DEFAULT_DATASET_ARGS = {"path": "conll2003", "trust_remote_code": True} | |||
There was a problem hiding this comment.
we shouldn't set trust_remote_code to True by default here, where is this needed ?
There was a problem hiding this comment.
this dataset was being used with "remote code" (it generates the dataset) for as long as it existed, but the datasets package introduced the new argument trust_remote_code lately, this preprocessing is used in in tflite tests, we can either find a similar dataset that doesn't use code or allow remote code (that was already being executed).
https://huggingface.co/datasets/eriktks/conll2003/blob/main/conll2003.py
There was a problem hiding this comment.
I think we need to change the dataset to something that doesn't require trust_remote_code to be set to True by default then (something that should be explicitly set by users)
What does this PR do?
Adding support for transformers 4.44. Biggest issue is with the changes in the bloom model attention block and its kv cache changes. Aslo fixing the failing TFLite tests (all green).
Before submitting
Who can review?