Commit Graph

Select branches

Hide Pull Requests

JetRunner-patch-1

adapt_t5_for_covid_19_3b

adapters

add_inference_caching_reformer

add_mp_net

add_sampling_and_training_to_transfo_xl_models

add_tf_train_to_benchmark

bart-before-3140

bart-mem-script

bart_onnx

batch-encoding-pickable

batch-parity-cleaner

benchmark

benchmark_for_2_11

bert-optim

better-tf-examples

blenderbot

branch_to_save_trax_integration_tests

broken-fp16-test

ci_fix-all-tests

ci_test-tf-slow

ci_test_cuda_v

ci_test_v350_release_slow_tests

circular-reference-schedulers

clean_encoder_decocer_modeling

clean_scripting

cleaner-pt-tf-conversion

conda

conv_bert

conversation_stateless

copy

copy_old_rag

ctrl-tokenizer-standardize

default-more-to-fast

dependabot/pip/examples/lxmert/notebook-6.1.5

deprecation-warning

develop

distil-update

documentation

dropper-celoss

electra-trainer

encoder_decoder_eli5

extract-dir

fast-tokenizer-warnings

fast-tokenizers

fill-mask-pipeline

finalize_rag

first_version_tf_benchmark

fix-1.3

fix-2042

fix-3091

fix-automodel-doc

fix-documentation-warnings

fix-gpt2-finetuning-memory

fix-model-output

fix-nan-values

fix-tf-xlm

fix-tf-xlnet

fix-tfroberta

fix-tok-pipe

fix-tokenizers-pretrained

fix-tpu-glue

fix_head_masking

fix_plotting_tool_benchmark

fix_pt_tf_loading

fixing_warnings_in_deberta

generation_sampler

gpt2_inference

gradient-checkpointing

gradient-checkpointing-v2

ignored-index-coherence

improve_shorter_than_chunk_length_usage_reformer

improved-generation

improved-testing

label-smooth-regression

longformer_tf

make_style

master

mem-benchmark

mem-prof-bart

memory_benchmark

metrics

mixing_args_and_kwargs

model-links

model-links-clean

model-parallelism

model-templates

model_parallel_tests

more_general_trainer_metric

ner_pipeline_offsets

new_flax_design

nlp

nlp-examples

notebooks-cron

onnx_export

pad_to_multiple_of

path-save-vocabulary-fast

patrickvonplaten-patch-2

performer

performer-experiments

pipeline-v2

pipelines-issues

pipelines_input_too_long

pipelines_tf_auto

pl-checkpoint-bug

position_ids_error

pre-style

previous-torch

prophetnet_develop

prophetnet_develop_parallel

qa-pipeline-handle-impossible

question-answering

quickstart-model2model

rbert

reformer_masked_lm

reformer_trax_tests

regression-test-bart

rename_ex_file

rename_keys_inference

roberta-no-token-types

save_old_prophetnet_model_structure

scripting-rebased

separate_generation_code

serving_improvements

shape_list

simple-examples

split-fast-slow

split_pipelines

ss-skip-failing-test

stas00-patch-1

stas00-patch-2

stas00-patch-3

stas00-patch-4

t5-parallelism

t5_documentation

tagging-architectures

tapas

tapas-final

tapas-style

tapas-v4-fix-truncation

tapas-v4-fixes

tapas-v4-lysandre

tapas_v4_final

temporary_fix_of_model_max_length

test-tf-load-slow-tests

text-classification-pipeline-raw-output

text-generation-pipeline-xlnet-transfo-padding

tf

tf2-extended

tf2-glue

tf_qa_pipelines_test

theseus

theseus-bart

tokenizers-config

tokenizers_save_pretrained

torch1-compat-testing

tpu-experiments

trainer-collator-device

transfoXL-StopIteration-bugfix

trigger-scheduled

try_if_works_for_longformer_mult_gpu

uncompressed_storage

update-deps

update_notebook

upgrade-run-generation

v3.0.2-docs

v3.5.1hotfix

v4.0.0-release

v4.0.1-release

warning_for_too_long_input

xnli

0.1.2

0.5.0

1.0

1.1.0

1.2.0

3.0.1

v0.1.2

v0.2.0

v0.3.0

v0.4.0

v0.5.0

v0.5.1

v0.6.0

v0.6.1

v0.6.2

v1.0.0

v2.0.0

v2.1.0

v2.1.1

v2.10.0

v2.11.0

v2.2.0

v2.2.1

v2.2.2

v2.3.0

v2.4.0

v2.4.1

v2.5.0

v2.5.1

v2.6.0

v2.7.0

v2.8.0

v2.9.0

v2.9.1

v3.0.0

v3.0.1

v3.0.2

v3.1.0

v3.2.0

v3.3.0

v3.3.1

v3.4.0

v3.5.0

v3.5.1

v4.0.0

v4.0.0-rc-1

v4.0.1

v4.1.0

v4.1.1

Mono Color

• ae333d04b2 torch.cuda.is_available() is redundant as apex handles that internally (#9350) master Stas Bekman 2020-12-30 01:09:51 -08:00
• e79c56fc5b [examples/language-modeling] Add dataset download instructions stas00-patch-4 Stas Bekman 2020-12-29 22:49:34 -08:00
• 429c676f95 [trainer] parametrize default output_dir stas00-patch-3 Stas Bekman 2020-12-29 15:00:38 -08:00
• 8217d4e37f [prophetnet] wrong import (#9349) Stas Bekman 2020-12-29 13:32:07 -08:00
• 912f6881d2 add import math (#9346) Patrick von Platen 2020-12-29 19:35:06 +01:00
• 785e52cd30 improve templates (#9342) Patrick von Platen 2020-12-29 16:48:44 +01:00
• 64103fb6be Fix TransfoXL (#9302) Julien Plu 2020-12-28 20:52:18 +01:00
• d97d06d05f Fix TF T5 (#9301) Julien Plu 2020-12-28 20:51:40 +01:00
• 83fdd252f6 [Seq2Seq Templates] Correct some TF-serving errors and add gradient checkpointing to PT by default. (#9334) Patrick von Platen 2020-12-28 17:51:04 +01:00
• 4ba7b0f6cc Torch import guard. conversation_stateless Nicolas Patry 2020-12-28 16:45:38 +01:00
• f9a057392f Fixing TranslationPipeline Hierarchy + return_name Nicolas Patry 2020-12-28 16:38:55 +01:00
• c145a9e87b Bugfix. Nicolas Patry 2020-12-28 14:39:32 +01:00
• 0e3e30779c Bumping isort version. Nicolas Patry 2020-12-28 14:20:22 +01:00
• ba6633e9a5 Making tests non flaky (by fixing generation length) Nicolas Patry 2020-12-28 14:18:11 +01:00
• 1b9edfee3e Adding tests that don't require downloading models + conversation can be Nicolas Patry 2020-12-28 14:00:36 +01:00
• 8e74eca7f2 push (#9320) Patrick von Platen 2020-12-27 21:57:50 +01:00
• 61443cd7d9 [GPT2] Correct gradient checkpointing (#9308) Patrick von Platen 2020-12-25 23:28:12 +01:00
• ba0252b7f9 Cleaning up conversation tests. Nicolas Patry 2020-12-25 14:20:35 +01:00
• 41e55cb75a Merging all duplicated codes for Text2TextPipeline while preserving backward compat. Nicolas Patry 2020-12-23 14:19:09 +01:00
• e043021108 Enabling pipelines users to override truncation tokenizer argument. Nicolas Patry 2020-12-23 13:49:22 +01:00
• fc10e4bcf8 Enable TruncationStrategy override for pipelines Nicolas Patry 2020-12-23 13:47:40 +01:00
• 21fc676645 add translation example (#9303) Vasudev GuptaandSuraj Patil 2020-12-25 14:47:49 +05:30
• 52b3a05e83 [Bart doc] Fix outdated statement (#9299) Patrick von Platen 2020-12-24 14:47:53 +01:00
• 7777db159f Update tokenization_utils_base.py (#9293) Bram Vanroy 2020-12-24 14:43:14 +01:00
• 71963a6633 fix typo in modeling_encoder_decoder.py (#9297) Daniele SartianoandSuraj Patil 2020-12-24 14:38:08 +01:00
• f3a3b91d6f Proposed Fix : [RagSequenceForGeneration] generate "without" input_ids (#9220) 3 people 2020-12-24 19:38:00 +07:00
• 2a18b70998 enable cache by default (#9296) Suraj Patil 2020-12-24 17:47:36 +05:30
• 0dc6fe8ecb Relative imports. Nicolas Patry 2020-12-24 10:53:12 +01:00
• 6189ae9960 Fix typo in file_utils.py (#9289) Jungwhan 2020-12-24 17:18:33 +09:00
• 222dbdb203 allow integer device for BatchEncoding (#9271) Jethro KuanandJethro Kuan 2020-12-24 16:01:56 +08:00
• 6c091abef2 [Templates] Adapt Bert (#9284) Patrick von Platen 2020-12-24 01:44:33 +01:00
• 88ef8893cd Add caching mechanism to BERT, RoBERTa (#9183) Suraj Patil 2020-12-23 23:01:32 +05:30
• a1cb6e9866 Adapt to new name of label_smoothing_factor training arg (#9282) Sylvain Gugger 2020-12-23 11:05:21 -05:00
• bcc87c639f Minor documentation revisions from copyediting (#9266) Connor Brinton 2020-12-23 10:15:49 -05:00
• 17021e54a1 Typo import fix. Nicolas Patry 2020-12-23 13:05:36 +01:00
• a67016442c Text2TextGenerationPipeline, TranslationPipeline, SummarizationPipeline Nicolas Patry 2020-12-23 13:02:42 +01:00
• 0cd2ed7521 ConversationnalPipeline Nicolas Patry 2020-12-23 12:56:35 +01:00
• 974aab8bb0 TableQuestionAnsweringPipeline Nicolas Patry 2020-12-23 12:49:22 +01:00
• 78f8e10cf3 QuestionAnsweringPipeline Nicolas Patry 2020-12-23 12:44:08 +01:00
• 338f224936 NerPipeline + TokenClassificationPipeline Nicolas Patry 2020-12-23 12:38:52 +01:00
• 68dc37d599 FillMaskPipeline split_pipelines Nicolas Patry 2020-12-23 12:34:18 +01:00
• 6175976506 ZeroShot + get_framework import. Nicolas Patry 2020-12-23 12:28:10 +01:00
• e9343aa36c TextClassifictionPipeline Nicolas Patry 2020-12-23 12:24:27 +01:00
• 59182f841c TextGenerationPipeline. Nicolas Patry 2020-12-23 12:18:50 +01:00
• 775eebb39d Moving FeatureExtractionPipeline into its own file. Nicolas Patry 2020-12-23 11:55:52 +01:00
• d5db6c37d4 [Seq2Seq Templates] Fix check_repo.py templates file (#9277) Patrick von Platen 2020-12-23 11:40:20 +01:00
• 329ec23b1c Moving everything into base.py Nicolas Patry 2020-12-23 11:40:15 +01:00
• 4bafc43b0e Fix param error (#9273) Xu Song 2020-12-23 18:34:57 +08:00
• 58e8a7611f Fix gpt2 document (#9272) Xu Song 2020-12-23 18:34:15 +08:00
• fdc819729d Splitting pipelines into its own module. Nicolas Patry 2020-12-23 11:28:57 +01:00
• cbe63949d7 Model Templates for Seq2Seq (#9251) label-smooth-regression Patrick von Platen 2020-12-22 23:41:20 +01:00
• e6c1f1cad8 Revert renaming in finetune_trainer (#9262) Sylvain Gugger 2020-12-22 15:42:34 -05:00
• ab17758874 Add speed metrics to all example scripts + template (#9260) Sylvain Gugger 2020-12-22 14:02:26 -05:00
• 5b5f7dd09c [hf_api] Fix incorrect typing Julien Chaumond 2020-12-22 19:52:47 +01:00
• 1558d191e6 Fix TF BART for saved model creation (#9252) 3 people 2020-12-22 18:07:04 +01:00
• 37d6fb5d04 Fix link to bertabs/README.md (#9255) Manuel Romero 2020-12-22 17:41:23 +01:00
• 189c1b91a6 Fix link to old language modeling script (#9254) Manuel Romero 2020-12-22 17:40:47 +01:00
• 490b39e614 Seq2seq trainer (#9241) 3 people 2020-12-22 11:33:44 -05:00
• 1fc7119181 Fix script that check objects are documented (#9259) Sylvain Gugger 2020-12-22 11:12:58 -05:00
• e9d77ccd5a [EncoderDecoder] Make tests more aggressive (#9256) Patrick von Platen 2020-12-22 17:00:04 +01:00
• ec07da65e2 Update the README of the text classification example (#9237) Sylvain GuggerandPatrick von Platen 2020-12-21 15:23:40 -05:00
• 4eef5889ac Adding performer fine-tuning research exampke (#9239) Teven 2020-12-21 21:19:41 +01:00
• 9a12b9696f [MPNet] Add slow to fast tokenizer converter (#9233) Patrick von Platen 2020-12-21 15:41:34 +01:00
• f4432b7e01 add base model classes to bart subclassed models (#9230) Suraj Patil 2020-12-21 19:56:46 +05:30
• 08abdabda1 Fixed beam search generation for GPT2 and T5 (#9219) TobiasNorlund 2020-12-21 14:05:23 +01:00
• 161a6461db Fix TF template (#9234) Julien Plu 2020-12-21 13:52:16 +01:00
• 5a8a4eb187 Improve BERT-like models performance with better self attention (#9124) Julien Plu 2020-12-21 13:10:15 +01:00
• 6b034309ca fix warning (#9231) Patrick von Platen 2020-12-21 10:41:34 +01:00
• a4b21cdd20 [RAG] Add Ray implementation for distributed retrieval (#9197) 3 people 2020-12-21 01:39:30 -08:00
• f38c4ad302 better logging and help (#9203) Stas Bekman 2020-12-20 10:28:28 -08:00
• e0e255be1f Added TF TransfoXL Sequence Classification (#9169) sandip 2020-12-19 19:14:04 +05:30
• 6b850b671d [run_glue] add speed metrics (#9198) Stas Bekman 2020-12-18 17:09:30 -08:00
• 3ff5e8955a [t5 doc] typos (#9199) Stas Bekman 2020-12-18 16:03:26 -08:00
• 291974c65c GPT-model attention heads pruning example (#9189) Aleksey Tikhonov 2020-12-18 22:32:10 +01:00
• 1198ba8fba Add timing inside Trainer (#9196) Sylvain Gugger 2020-12-18 15:10:39 -05:00
• 9a25c5bd3a Add new run_swag example (#9175) Sylvain GuggerandLysandre Debut 2020-12-18 14:19:24 -05:00
• 3e56e2ce04 Fix typo Sylvain Gugger 2020-12-18 10:11:07 -05:00
• 077a5dce32 Fix link to old SQUAD fine-tuning script (#9181) Manuel Romero 2020-12-18 15:12:10 +01:00
• 84d5879eaf [setup] correct transformers version format (#9176) Stas Bekman 2020-12-18 05:55:55 -08:00
• fd7b6a5274 fixed JSON error in run_qa with fp16 (#9186) Wissam Antoun 2020-12-18 14:53:23 +02:00
• 66a14a2f6f Fix link to old NER fine-tuning script (#9182) Manuel Romero 2020-12-18 01:50:01 +01:00
• f06d0fadc9 [trainer] apex fixes and tests (#9180) Stas Bekman 2020-12-17 16:49:11 -08:00
• 467e9158b4 Added TF CTRL Sequence Classification (#9151) sandip 2020-12-18 04:40:57 +05:30
• 63841c559b add tests for the new sharded ddp fairscale integration (#9177) Stas Bekman 2020-12-17 14:24:03 -08:00
• bf713cdec7 setup.py development version Lysandre 2020-12-17 11:29:31 -05:00
• bd40345d3e v4.1.1 docs Lysandre 2020-12-17 11:28:38 -05:00
• bfa4ccf77d Release: v4.1.1 v4.1.1 Lysandre 2020-12-17 11:25:49 -05:00
• e0790cca78 Fix TAPAS doc Lysandre 2020-12-17 11:24:23 -05:00
• 6d2e864db7 Put all models in the constants (#9170) Sylvain Gugger 2020-12-17 11:23:21 -05:00
• f83d9c8da7 v4.1.0 docs Lysandre 2020-12-17 10:16:07 -05:00
• f5438ab8a2 Release: v4.1.0 v4.1.0 Lysandre 2020-12-17 10:04:55 -05:00
• ac2c7e398f Remove erroneous character Lysandre 2020-12-17 09:47:19 -05:00
• 77d6941e64 Fix gradient clipping for Sharded DDP (#9168) Sylvain Gugger 2020-12-17 09:44:24 -05:00
• 1aca3d6afa Add disclaimer to TAPAS rst file (#9167) Lysandre Debutandsgugger 2020-12-17 09:34:06 -05:00
• dc9f245442 Torch scatter with torch 1.7.0 Lysandre 2020-12-16 13:48:52 -05:00
• 9a67185344 Experimental support for fairscale ShardedDDP (#9139) Sylvain Gugger 2020-12-16 13:47:48 -05:00
• 1c1a2ffbff TableQuestionAnsweringPipeline (#9145) Lysandre DebutandPatrick von Platen 2020-12-16 12:31:50 -05:00
• 07384baf7a AutoModelForTableQuestionAnswering (#9154) Lysandre Debut 2020-12-16 12:14:33 -05:00
• 34334662df Add message to documentation that longformer doesn't support token_type_ids (#9152) Hayden Housen 2020-12-16 11:06:14 -05:00
• 2f918defa8 hotfix torch scatter version Lysandre 2020-12-16 10:26:13 -05:00