gsarti/covid-papers-browser

Browse Covid-19 & SARS-CoV-2 Scientific Papers with Transformers 🦠 📖

covid-papers-browser's Issues

Wrong variable reference in embeddings dump to pickle

https://github.com/gsarti/covid-papers-browser/blob/master/scripts/interactive_search.py#L138

error in download_scibert.py (os.rmtree)

Hello, great library! I am getting the error AttributeError: module 'os' has no attribute 'rmtree' using python 3.7.

This is fixed if os.rmtree(path) is replaced with shutil.rmtree(path)

Add argparse parametrization for the finetuning script

Similar to what is currently available in download_model.py, add Argparse with parameters in finetune_nli.py for parameters:

• model_name, default 'models/scibert', type str

• batch_size, default 64, type int

• model_save_path, default 'models/scibert_nli', type str

• num_epochs, default 2, type int

• warmup_steps, default None, not required

• do_mean_pooling with action='store_true'

• do_cls_pooling with action='store_true'

• do_max_pooling with action='store_true'

Then:

• Add check for only one of the pooling condition to be verified (raise an AttributeError if more than one is). If none is specified, we use mean pooling strategy.

• Check if the warmup_step parameter is set before setting it to 10% of training: if it is, keep the user-defined value.

Open Source Helps!

Thanks for your work to help the people in need! Your site has been added! I currently maintain the Open-Source-COVID-19 page, which collects all open source projects related to COVID-19, including maps, data, news, api, analysis, medical and supply information, etc. Please share to anyone who might need the information in the list, or will possibly contribute to some of those projects. You are also welcome to recommend more projects.

http://open-source-covid-19.weileizeng.com/

Cheers!

Unable to download scibert-nli after latest changes

Trying:

python scripts/download_model.py --model scibert-nli

gets:

Traceback (most recent call last):
  File "scripts/download_model.py", line 50, in <module>
    tokenizer = AutoTokenizer.from_pretrained(MODELS_PRETRAINED[args.model])
KeyError: 'scibert-nli'

Error in pip install of requirements, but still worked?

I'm on ubuntu, made a fresh conda environment and conda installed pip within it, then ran pip install -r requirements.txt

I got this

ERROR: sentence-transformers 0.2.5.1 has requirement transformers==2.3.0, but you'll have transformers 2.5.1 which is incompatible.

But then the pip install continued and apparently succeeded:

rmtree instead of shutil.rmtree

covid-papers-browser/scripts/download_scibert.py

Line 42 in 7f7ec8c

shutil.rmtree(path)

What is the hardware used to fine tune SciBERT using NLI datasets?

Its a nice work. I would like to know the following details @gsarti
a)Which GPU you have used to fine-tune SciBERT?
b) How much time did it take as a whole to fine SciBERT using NLI and MNLI datasets?

Source code for SciBERT NLI

Hi,
I am a novice in the domain of fine-tuning my own dataset on any Transformer models. I want to fine-tune my dataset on the SciBERT NLI model(https://huggingface.co/gsarti/scibert-nli). The dataset formation is
S1 [SEP] S2 [SEP] Inference

I am not sure how to fine-tune my dataset on the SciBERT NLI model(https://huggingface.co/gsarti/scibert-nli).

I am sorry for asking very simple question. But any suggestion or link helps me. Thanks in advance.

unable to download models using command

I cant download models using python scripts/download_model.py --model scibert-nli
Script is stuck. is there a manual way I can download the models?

Elastic Search

Switch from mongo to elastic search.

Train transformer models on MedNLI

Great work!

Heres some data that might be more domain related. Its not open but it might be helpful?
https://jgc128.github.io/mednli/