transformers-interpret

transformers-interpret copied to clipboard

Hi there, I am observing the following (strange) behavior when using `pipeline` from the `transformers` library and `transformer-interpret`: ``` text = "Now Accord networks is a company in video, and...

simonschoe

Max class probability too low with a multi-class classifier

6

Hi @cdpierse , very nice project, congrats! I am doing some experiments for bias in sentiment classifiers using our tool Rubrix together with `transformers-interpret` and I have encountered an issue....

dvsrepo

change tokenizer parameters

4

Hi, It is ok when I use SequenceClassificationExplainer with short texts but for long texts it throws an error like `RuntimeError: The expanded size of the tensor (583) must match...

yusufcakmakk

As seen from [transformers-interpret](https://github.com/cdpierse/transformers-interpret)/[notebooks](https://github.com/cdpierse/transformers-interpret/tree/master/notebooks)/multiclass_classification_example.ipynb, the input to [transformers-interpret](https://github.com/cdpierse/transformers-interpret) is the text of a single sentence **can we input a batch? or input token_ids?**

WenxiongLiao

In questionanswerexplainer, for context length greater than 2000(approx) is throwing an error(Index out of range self)

3

niponchanda

Support for list of sentences

2

The current version does not support function for getting attributions for list of sentences in `SequenceClassificationExplainer` . Is this feature in process of development? If not then I would like...

aka2910

I tried just passing 'input 1 [SEP] input 2' but kept getting the wrong predicted label from the SingleLabelClassifier -- I assume its because the tokenizer isn't assigning token_type_ids. Is...

JosephGatto

ZeroShotClassificationExplainer does not correctly explain ZeroShotClassificationPipeline results (single label)

2

In the case of a single label, the logic to calculate the classification probability with the `ZeroShotClassificationExplainer` (see [here](https://github.com/cdpierse/transformers-interpret/blob/v0.6.0/transformers_interpret/explainers/sequence_classification.py#L193-L194)) is different than the logic in the Huggingface `ZeroShotClassificationPipeline` (see [here](https://github.com/huggingface/transformers/blob/v4.15.0/src/transformers/pipelines/zero_shot_classification.py#L249-L251)):...

ArneBinder

Sequence Classification Explainer  MultiLabel (wrong mapping value)  but thanks for the package~!

jw0831

[RobertaForSequenceClassification] RAM memory leaks during retrieving word attributions

8

Hi folks, I'm experiencing some memory leaks when using the Transformers Interpret (TI) library which triggers out of memory errors and kills the model service process. ## Setup Here is...

roma-glushko