Resource
BERT is a method of pre-training language representations which obtains state-of-the-art results on a wide array of NLP tasks.
Use the NGC CLI to download:
Copied!
| Name | Size | Updated | Actions |
|---|---|---|---|
.dockerignore | 22 B | October 2, 2019 UTC | |
bert_config.json | 314 B | October 2, 2019 UTC | |
create_pretraining_data.py | 16.96 KB | October 2, 2019 UTC | |
Dockerfile | 773 B | October 2, 2019 UTC | |
extract_features.py | 11.87 KB | October 2, 2019 UTC | |
file_utils.py | 8.09 KB | October 2, 2019 UTC | |
fused_adam_local.py | 8.97 KB | October 2, 2019 UTC | |
LICENSE | 11.09 KB | October 2, 2019 UTC | |
modeling.py | 60.1 KB | October 2, 2019 UTC | |
NOTICE | 148 B | October 2, 2019 UTC | |
optimization.py | 9.5 KB | October 2, 2019 UTC | |
README.md | 44.01 KB | October 2, 2019 UTC | |
requirements.txt | 208 B | October 2, 2019 UTC | |
run_glue.py | 27.51 KB | October 2, 2019 UTC | |
run_pretraining_inference.py | 12.58 KB | October 2, 2019 UTC | |
run_pretraining.py | 18.25 KB | October 2, 2019 UTC | |
run_squad.py | 51.63 KB | October 2, 2019 UTC | |
run_swag.py | 24.04 KB | October 2, 2019 UTC | |
schedulers.py | 3.28 KB | October 2, 2019 UTC | |
tokenization.py | 14.73 KB | October 2, 2019 UTC |