Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
As an example, let's use WhisperModel.forward's docstring:
thon
r"""
Returns:
Example:
thon
>>> import torch
>>> from transformers import WhisperModel, WhisperFeatureExtractor
>>> from datasets import load_dataset
>>> model = WhisperModel.from_pretrained("openai/whisper-base")
>>> feature_extractor = WhisperFeatureExtractor.from_pretrained("openai/whisper-base")
>>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
>>> inputs = feature_extractor(ds[0]["audio"]["array"], return_tensors="pt")
>>> input_features = inputs.input_features
>>> decoder_input_ids = torch.tensor([[1, 1]]) * model.config.decoder_start_token_id
>>> last_hidden_state = model(input_features, decoder_input_ids=decoder_input_ids).last_hidden_state
>>> list(last_hidden_state.shape)
[1, 2, 512]
```"""
Just run the following line to automatically test every docstring example in the desired file:
pytest --doctest-modules <path_to_file_or_dir>
If the file has a markdown extention, you should add the --doctest-glob="*.md" argument.