search

jalammar / ecco

Explain, analyze, and visualize NLP language models. Ecco creates interactive visualizations directly in Jupyter notebooks explaining the behavior of Transformer-based language models (like GPT2, BERT, RoBERTA, T5, and T0).
https://ecco.readthedocs.io
BSD 3-Clause "New" or "Revised" License
1.98k stars 168 forks source link

Error message of run_nmf when using a model with activations=False #15

Closed ricvolpe closed 2 years ago

ricvolpe commented 3 years ago

If I load a model like this

lm = ecco.from_pretrained('gpt2')

without specifically setting activations=True

The error I get when trying to run NFM is not very clear, specifically, it shows

/usr/local/lib/python3.6/dist-packages/ecco/output.py in __init__(self, activations, n_input_tokens, token_ids, _path, n_components, tokens, **kwargs)
    498         from_layer = kwargs['from_layer'] if 'from_layer' in kwargs else None
    499         to_layer = kwargs['to_layer'] if 'to_layer' in kwargs else None
--> 500         if len(activations.shape) != 3:
    501             raise ValueError(f"The 'activations' parameter should have three dimensions: (layers, neurons, positions). "
    502                              f"Supplied dimensions: {activations.shape}", 'activations')

AttributeError: 'list' object has no attribute 'shape'

This seems an easy mistake that a beginner user could make and could easily be improved by checking earlier that output object has properties required for doing NFM.

jalammar commented 3 years ago

Good point @ricvolpe. I've added a clearer error message and will ship it to the next release. Thanks!