NielsRogge/transformers

Pull Requests Commits

Allow passing encoder_ouputs as tuple to EncoderDecoder Models (#16814)

jsnfly committed 3 years ago

Verified 51e0ebed

use base_version to check torch version in torch_less_than_1_11 (#16806)

nbroad1881 committed 3 years ago

Verified 51fa7191

[Data2Vec] Add data2vec vision (#16760)

patrickvonplaten committed 3 years ago

Verified 8d3f952a

fix megatron bert convert state dict naming (#15820)

Codle committed 3 years ago

Verified 33cd4be5

[Quicktour Audio] Improve && remove ffmpeg dependency (#16723)

patrickvonplaten committed 3 years ago

Verified 9a2995ee

[ViT, BEiT, DeiT, DPT] Improve code (#16799)

NielsRogge committed 3 years ago

Verified d3c9d0e5

Fix syntax error in TorchHub workflow

sgugger committed 3 years ago

3785f466

Create empty venv on cache miss (#16816)

gante committed 3 years ago

Verified 6984848e

Raise error and suggestion when using custom optimizer with Fairscale or Deepspeed (#16786)

allanj committed 3 years ago

Verified 43814483

TF generate refactor - XLA sample (#16713)

gante committed 3 years ago

Verified b4ddd267

CI: non-remote GH Actions now use a python venv (#16789)

gante committed 3 years ago

Verified 02de7a8e

Pin Jax to last working release (#16808)

sgugger committed 3 years ago

Verified dee6f016

Update README.md (#16797)

NielsRogge committed 3 years ago

Verified 78f346c2

Fix PT TF ViTMAE (#16766)

ydshieh committed 3 years ago

Verified ee209d4d

[modeling utils] revamp `from_pretrained(..., low_cpu_mem_usage=True)` + tests (#16657)

stas00 committed 3 years ago

Verified 5da33f87

[trainer / deepspeed] fix hyperparameter_search (#16740)

stas00 committed 3 years ago

Verified ce2fef2a

Fix issue avoid-missing-comma found at https://codereview.doctor (#16768)

code-review-doctor committed 3 years ago

Verified 1b7de41a

[SpeechEncoderDecoderModel] Fix bug in reshaping labels (#16748)

sanchit-gandhi committed 3 years ago

Verified de8b06f9

Improve image classification example (#16585)

NielsRogge committed 3 years ago

Verified 048443db

Kill async pushes when calling push_to_hub with blocking=True (#16755)

sgugger committed 3 years ago

Verified 3e4eec47

[deepspeed / m2m_100] make deepspeed zero-3 work with layerdrop (#16717)

stas00 committed 3 years ago

Verified c21e1071

Make nightly install dev accelerate (#16783)

muellerzr committed 3 years ago

Verified 89293a0f

Fix batch size in evaluation loop (#16763)

sgugger committed 3 years ago

Verified b151ddb9

[Flax `.from_pretrained`] Raise a warning if model weights are not in float32 (#16762)

sanchit-gandhi committed 3 years ago

Verified d8269eb4

Enabling `Tapex` in table question answering pipeline. (#16663)

Narsil committed 3 years ago

Verified 195fbbb6

[Doctest] added doctest changes for electra (#16675)

bhadreshpsavani committed 3 years ago

Verified 442dc456

Fixup no_trainer examples scripts and add more tests (#16765)

muellerzr committed 3 years ago

Verified be752d12

[self-scheduled ci] explain where dependencies are (#16757)

stas00 committed 3 years ago

Verified 3a16ab25

Add self training code for text classification (#16738)

tuvuumass committed 3 years ago

Verified 34ef029d

Add defensive check for config num_labels and id2label (#16709)

sgugger committed 3 years ago

Verified 8e0d3b42

Newer Older