Downloads · 30 days
9
1% of all-time downloads
Sinhala-RoBERTa
d42kw01f/Sinhala-RoBERTa
Sinhala-RoBERTa is a fill-mask model from d42kw01f. Use it when you need the model to fill a missing word. It is set up for transformers.
This is a smaller per-trained model on Sinhalese Language using Masked Language Modeling(MLM). And the model is trained on Oscar Sinhala dataset.
All-time downloads
1.2K
Public
Repo size
536 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin268 MB · 100%
At a glance
- Task
- Fill-Mask
- Library
- transformers
- Model type
- roberta
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Nov 6, 2021
- SHA
- 5b3c8e5d
Try a prompt
Model card
From the Hugging Face model README
Description:
This is a smaller per-trained model on Sinhalese Language using Masked Language Modeling(MLM). And the model is trained on Oscar Sinhala dataset.
How to Use:
The model can be used directly with a pipeline for masked language modeling:
>>> from transformers import AutoTokenizer, AutoModelForMaskedLM, pipeline
>>> tokenizer = AutoTokenizer.from_pretrained("d42kw01f/Sinhala-RoBERTa")
>>> model = AutoModelForMaskedLM.from_pretrained("d42kw01f/Sinhala-RoBERTa")
>>> fill_mask = pipeline('fill-mask', model=model, tokenizer=tokenizer)
>>> fill_mask("මම ගෙදර <mask>.")
[{'score': 0.1822454035282135,
'sequence': 'මම ගෙදර ආව.',
'token': 701,
'token_str': ' ආව'},
{'score': 0.10513380169868469,
'sequence': 'මම ගෙදර ය.',
'token': 310,
'token_str': ' ය'},
{'score': 0.06417194753885269,
'sequence': 'මම ගෙදර එක.',
'token': 328,
'token_str': ' එක'},
{'score': 0.05026362091302872,
'sequence': 'මම ගෙදර ඇත.',
'token': 330,
'token_str': ' ඇත'},
{'score': 0.029960114508867264,
'sequence': 'මම ගෙදර යනව.',
'token': 834,
'token_str': ' යනව'}]
- Task
- Fill-Mask
- Library
- transformers
- Type
- roberta
- Created
- Mar 2, 2022
- Updated
- Nov 6, 2021