Author: Fantashit

Posted in Uncategorized

Please publish face-landmarks-detection 0.0.3 to npm

It looks like the face-landmarks-detection model was updated to work with TFJS 3.0 a few weeks ago https://github.com/tensorflow/tfjs-models/blob/2af288d98609b32df2075cf7226c694989a5d476/face-landmarks-detection/package.json#L3 But npm still only has the previous…

Continue Reading
Posted in Uncategorized

model.summary() output shape is “multiple” when defining input_shape via model.build()

When I define a model and pass the input_shape to the first layer, the Output Shape is well-defined after I call model.summary(). However, if I…

Continue Reading
Posted in Uncategorized

TFBartForConditionalGeneration with labels padded with -100 gives Nan loss.

I am pretraining T5 and Bart. I noticed that the padding token for labels of these models should be -100 for decoder_input_ids. I change the…

Continue Reading
Posted in Uncategorized

python utils/check_repo.py fails

on master after making sure I got all the deps updated (from make style/quality/fixup) No library .py files were modified running deps_table_update updating src/transformers/dependency_versions_table.py python…

Continue Reading
Posted in Uncategorized

run_ner.py raised error

Environment info transformers version: 4.3.0.dev0 Platform: MacOS Python version: 3.6 PyTorch version (GPU?): 1.7.1 Tensorflow version (GPU?): 2.4.1 Using GPU in script?: No Using distributed…

Continue Reading
Posted in Uncategorized

[License info] Longformer SQuAD finetuned model

Hello @patil-suraj , would it be possible to provide licensing information for the pretrained model weights shared at: https://huggingface.co/valhalla/longformer-base-4096-finetuned-squadv1 I would be interested in offering…

Continue Reading
Posted in Uncategorized

[DeepSpeed] [success] trained t5-11b on 1x 40GB gpu

Managed to train t5-11b on 1x 40GB gpu w/ Deepspeed (A100-SXM4-40GB) Thank you, @PeterAJansen for letting me use your hardware! Thank you, @jeffra and @samyam,…

Continue Reading
Posted in Uncategorized

🚀 Faster batch translation with FSMT model

🚀 Faster batch translation with FSMT model Currently, generating translations for multiple inputs at once is very slow using Transformers’ FSMTForConditionalGeneration implementation. In fact it’s…

Continue Reading
Posted in Uncategorized

1.3GB dataset creates over 107GB of cache file!

Environment info transformers version: 4.4.0 dev0 Platform: Google Colab Python version: 3.6 PyTorch version (GPU?): 1.7 Tensorflow version (GPU?): None Using GPU in script?: None….

Continue Reading
Posted in Uncategorized

Uploaded a new model but is not found on the hub.

🌟 New model addition I recently added this model: https://huggingface.co/flexudy/t5-small-wav2vec2-grammar-fixer However, I get this error whilst trying to download it. Can’t load tokenizer for ‘flexudy/t5-small-wav2vec2-grammar-fixer’…

Continue Reading