NielsRogge/transformers

Pull Requests Commits

Don't use fast llama tokenizer for now

NielsRogge committed 2 years ago

a68bf0b2

NielsRogge committed 2 years ago

efe69230

sgugger committed 2 years ago

Verified ba695c1e

Add AzureOpenAiAgent (#24058)

sgugger committed 2 years ago

Verified c3572e6b

Up pinned accelerate version (#24089)

muellerzr committed 2 years ago

Verified 5eb3d3c7

fix accelerator prepare during eval only mode (#24014)

pacman100 committed 2 years ago

Verified d1c039e3

Do not prepare lr scheduler as it as the right number of steps (#24088)

sgugger committed 2 years ago

Verified 2c887cf8

fix executable batch size issue (#24067)

pacman100 committed 2 years ago

Verified 12298cb6

Update delete_doc_comment_trigger.yml (#24084)

mishig25 committed 2 years ago

Verified ef010071

Fix expected value in tests of the test fetcher (#24077)

sgugger committed 2 years ago

Verified 89b00eef

[doc build] Use secrets (#24079)

mishig25 committed 2 years ago

Verified 5c9394b5

Make the TF dummies even smaller (#24071)

Rocketknight1 committed 2 years ago

Verified 1fc832b4

Be nice to TF (#24076)

ydshieh committed 2 years ago

Verified 092c14c3

[`bnb`] Fix bnb skip modules (#24043)

younesbelkada committed 2 years ago

Verified 47952192

Fix `is_optimum_neuron_available` (#23961)

michaelbenayoun committed 2 years ago

Verified a1160185

[`Hub`] Add `safe_serialization` in push_to_hub (#24074)

younesbelkada committed 2 years ago

Verified 6b548129

Support PEFT models when saving the model using trainer (#24073)

younesbelkada committed 2 years ago

Verified 6daf7c31

Add support for non-rust implemented tokenization for `__getitem__` method. (#24039)

jacklanda committed 2 years ago

Verified 1e4a7737

[Wav2Vec2] Fix torch srcipt (#24062)

patrickvonplaten committed 2 years ago

Verified 52972e70

Generate: increase left-padding test atol (#23448)

gante committed 2 years ago

Verified 612b2a1a

Remote code improvements (#23959)

sgugger committed 2 years ago

Verified f1660d7e

Fix device placement for model-parallelism in generate for encoder/de… (#24025)

sgugger committed 2 years ago

Verified 60825f2c

bring back `filtered_test_list_cross_tests.txt` (#24055)

ydshieh committed 2 years ago

Verified 02d255db

Use new parametrization based weight norm if available (#24030)

ezyang committed 2 years ago

Verified bc9ecef9

Move TF building to an actual build() method (#23760)

Rocketknight1 committed 2 years ago

Verified 4a55e478

Oops, missed one (#24054)

muellerzr committed 2 years ago

Verified cbf6bc23

Reduce memory usage in TF building (#24046)

Rocketknight1 committed 2 years ago

Verified 7203ea67

Act on deprecations in Accelerate no_trainer examples (#24053)

muellerzr committed 2 years ago

Verified 072188d6

Tiny fix for `check_self_hosted_runner.py` (#24052)

ydshieh committed 2 years ago

Verified ff4c0fc7

Add TimmBackbone model (#22619)

amyeroberts committed 2 years ago

Verified a717e031

Older