NielsRogge/transformers

Pull Requests Commits

potpurri of small fixes (#8807)

stas00 committed 5 years ago

Verified ddf3c646

Fix PPLM (#8779)

chutaklee committed 5 years ago

Verified 52708d26

Revert "finetune.py: specifying generation min_length (#8478)" (#8805)

patrickvonplaten committed 5 years ago

Verified 8f07f5c4

Create README.md (#8760)

mrm8488 committed 5 years ago

Verified 66e9608b

finetune.py: specifying generation min_length (#8478)

Daniel Khashabi committed 5 years ago

Verified 5aa361f3

Create README.md (#8752)

joangines committed 5 years ago

Verified 30e7f7e5

[XLNet] Fix mems behavior (#8567)

patrickvonplaten committed 5 years ago

Verified 2a6fbe6a

Return correct Bart hidden state tensors (#8747)

joeddav committed 5 years ago

Verified 369f1d77

Fix QA argument handler (#8765)

LysandreJik committed 5 years ago

Verified 138f45c1

Big model table (#8774)

sgugger committed 5 years ago

Verified 4821ea5a

Create README.md (#8761)

mrm8488 committed 5 years ago

Verified 90d5ab3b

New TF model inputs (#8602)

jplu committed 5 years ago

Verified 29d49924

[core] implement support for run-time dependency version checking (#8645)

stas00 committed 5 years ago

Verified 82d443a7

fix rag index names in eval_rag.py example (#8730)

lhoestq committed 5 years ago

Verified a7d73cfd

added instructions for syncing upstream master with forked master via PR (#8745)

bdalal committed 5 years ago

Verified 8d4ed7e9

MT5 should have an autotokenizer (#8743)

LysandreJik committed 5 years ago

Verified e09e54fd

Fix slow tests v2 (#8746)

LysandreJik committed 5 years ago

Verified 6fdd0bb2

Support various BERT relative position embeddings (2nd) (#8276)

zhiheng-huang committed 5 years ago

Verified 2c83b3c3

[EsperBERTo] Fix URLs to assets

julien-c committed 5 years ago

9e71aa2f

Model parallel documentation (#8741)

LysandreJik committed 5 years ago

Verified 02f48b9b

TF BERT test update

LysandreJik committed 5 years ago

7f2c0091

Update TF BERT test

LysandreJik committed 5 years ago

e1b7e10d

Add early stopping callback to pytorch trainer (#8581)

Colin Brochtrup committed 5 years ago

Verified 8ffc01a7

Fix max length in run_plm script (#8738)

sgugger committed 5 years ago

Verified 367f497d

consistent ignore keys + make private (#8737)

stas00 committed 5 years ago

Verified e84786aa

Document new training argument

sgugger committed 5 years ago

49759c0c

gpt2 and t5 parallel modeling (#8696)

alexorona committed 5 years ago

Verified 1cd9be2a

[trainer] make generate work with multigpu (#8716)

stas00 committed 5 years ago

Verified 1e45bef0

Change default cache path (#8734)

sgugger committed 5 years ago

Verified 90002427

Improve bert-japanese tokenizer handling (#8659)

julien-c committed 5 years ago

Verified 0cc5ab13

Newer Older