NielsRogge/transformers

Pull Requests Commits

changed do_output_past function to check for self.config.output_past instead of self.output_past

patrickvonplaten committed 6 years ago

7e0c5c73

rename comments

patrickvonplaten committed 6 years ago

eeaa402c

remove ipdb debugging statements

patrickvonplaten committed 6 years ago

7bb42712

add and improve comments

patrickvonplaten committed 6 years ago

267587c2

add past hidden key states for more efficient language generation & add prepare_inputs for gpt2 and ctrl model

patrickvonplaten committed 6 years ago

d891fd0a

Merge pull request #2303 from patrickvonplaten/fix_error_with_repetition_penalty

thomwolf committed 6 years ago

Verified aeef4823

Merge pull request #2291 from aaugustin/fix-flake8-F841

thomwolf committed 6 years ago

Verified 0412f3d9

Merge pull request #2289 from patrickvonplaten/fix_effective_batch_size_lang_gen_xlm

thomwolf committed 6 years ago

Verified 8742c954

Merge pull request #2312 from vitaliyradchenko/fix_special_and_add_tokens_loading

thomwolf committed 6 years ago

Verified 1240be3e

add special tokens to unique_added_tokens_encoder

vitaliyradchenko committed 6 years ago

b262577d

fixed lack of added and special tokens

vitaliyradchenko committed 6 years ago

83a23479

Merge pull request #2310 from ShnitzelKiller/scatter-unfix

thomwolf committed 6 years ago

Verified cea04a24

use positional arguments due to inconsistent API

ShnitzelKiller committed 6 years ago

e1844d9a

revert erroneous fix

ShnitzelKiller committed 6 years ago

9fb7addd

tokenizers is now a real dependency

n1t0 committed 6 years ago

734d29b0

Add tests for fast tokenizers

n1t0 committed 6 years ago

2818e505

n1t0 committed 6 years ago

31c56f2e

BertTokenizerFast

n1t0 committed 6 years ago

951ae99b

GPT2TokenizerFast

n1t0 committed 6 years ago

041eac2d

FastPreTrainedTokenizer

n1t0 committed 6 years ago

3471ff0d

fix repetition penalty error in modeling_utils.py

patrickvonplaten committed 6 years ago

18e5bdbe

fix sequence length for prepare_inputs for xlnet

patrickvonplaten committed 6 years ago

f18ac4c2

fix effective batch_size error in prepare_inputs also for xlnet

patrickvonplaten committed 6 years ago

359dc438

fix bug in prepare inputs for language generation for xlm for effective batch_size > 1

patrickvonplaten committed 6 years ago

d98a384c

adding back last dropout in TF 2.0 T5

thomwolf committed 6 years ago

3e0cf495

adding back final dropout in T5

thomwolf committed 6 years ago

35d32308

Merge pull request #2271 from aaugustin/improve-setup-and-requirements

thomwolf committed 6 years ago

Verified 81db12c3

Run the slow tests every Monday morning.

aaugustin committed 6 years ago

10724a81

Remove [--editable] in install instructions.

aaugustin committed 6 years ago

a8d34e53

Enable F841 warning in flake8.

aaugustin committed 6 years ago

e74c73a8

Newer Older