Commit History

Bug fixes
ab0523b

PeteBleackley commited on

Script for preparing training datasets
d63afc6

PeteBleackley commited on

Decided to use pandas rather than datasets
9ca9d81

PeteBleackley commited on

Additions to requirements.txt and .gitignore
5f8e115

PeteBleackley commited on

Components for managing training corpora
4f8366b

PeteBleackley commited on

Coreference Resolution for WikiQA dataset
e149b0f

PeteBleackley commited on

Data preparation for WikiQA
c106121

PeteBleackley commited on

CorpusRepeater
985ef96

PeteBleackley commited on

More work on models
f16a715

PeteBleackley commited on

Added transformers to requirements.txt
eef3bd6

PeteBleackley commited on

Encoder, Decoder and Trainer models (assuming RoBERTa base models)
f9c0522

PeteBleackley commited on

GlobalAttentionPoolingHead layer
8f1745b

PeteBleackley commited on

Minor formatting change
df051eb

PeteBleackley commited on

Datasets documentation
cc0d6a0

PeteBleackley commited on

Datasets documentation
58e8b0b

PeteBleackley commited on

Code for building and training base modkels.
66ccfdf

PeteBleackley commited on

Code for building and training base modkels.
83d5adb

PeteBleackley commited on

Corpus iterator for BNC
75ef467

PeteBleackley commited on

Updates to README.md and requirements.txt
47a7fc3

PeteBleackley commited on

HierarchicalLogits layer and function to create base models
064f030

PeteBleackley commited on

Hyena layer and HierarchicalSoftMax layer
646cf92

PeteBleackley commited on

Initial commit
eb23991
unverified

Pete Bleackley commited on