Israel Abebe

Distributed inference on pretrained PyTorch model

3

I am trying to do inference on my pretrained pytorch model. I loaded my model successfully like the following. ``` torch_model = net # after initializing my pytorch model and...

I am testing paddle OCR and wanted to know if it has a column identifying feature instead of treating columns line by line and joining them. @tink2123 is there already...

@raamb Was working on deploying the service to sent platform and most of the tutorials in [this](https://dev.singularitynet.io/tutorials/publish/) do not cover if we are deploying it from certain IP. I tried...

How to use RobertaFeaturizer in data loading

10

## ❓ Questions & Help I was experimenting with different featurizers and have problem loading RobertaFeaturizer as other featurizers. I tried ``` task, datasets, _ = dc.molnet.load_clintox(featurizer= dc.feat.RobertaFeaturizer.from_pretrained("seyonec/SMILES_tokenized_PubChem_shard00_160k"),splitter='scaffold') ``` but...

Requirement file

1

while trying to replicate the training, i am having a problem that arise from pytorh lightning version issues, can you create a requirement file for the repository? thanks

[detectron2](https://github.com/facebookresearch/detectron2) has better segmentation module that we can use.

Possible models to experiment with - https://huggingface.co/Davlan/afro-xlmr-large - https://huggingface.co/castorini/afriberta_large

credentials issue

3

I was playing around with the opensource local deployment and was having a problem with the credentials what am i doing wrong? 

Empty --log_samples outputs

1

_I checked similar issues_ While working with this tool for machine translation `generate_until` task all calculation give correct output but after doing `--log_samples ` in wandb i see the following...

How can we use gemini api keys to evaluate gemini models similar to openai models.