huggingface/transformers

Pull Requests Commits

Let's use a smaller set of files first

Rocketknight1 committed 278 days ago

c02b2157

Revert changes to focus on smaller model set

Rocketknight1 committed 278 days ago

a311acf3

Rocketknight1 committed 278 days ago

5f96f727

try folding transpose into einsum

Rocketknight1 committed 278 days ago

97f17261

Rocketknight1 committed 278 days ago

c2313a35

why not more einsums

Rocketknight1 committed 278 days ago

60f2dafe

Rocketknight1 committed 278 days ago

be0cbe06

Fix looping in torch guard decorator (#42260)

Cyrilvallez committed 278 days ago

Verified a5c903f8

Fix bnb for the weights refactor (#42043)

SunMarc committed 278 days ago

Verified 67302b04

delete already deprecated models (#42235)

ydshieh committed 279 days ago

Verified 9f311047

Cleanup reference to TFBertTokenizer and TFGPT2Tokenizer (#42182)

Rocketknight1 committed 279 days ago

Verified d372b827

Revert "Make tests run in less time by reducing `batch_size`" (#42258)

ydshieh committed 279 days ago

Verified b2feaa21

Make tests run in less time by reducing `batch_size` (#42213)

ydshieh committed 279 days ago

Verified 1acbd0b3

Allow VLMs to have a correct `base_model` (#41589)

zucchini-nlp committed 279 days ago

Verified c40b370b

🚨 Delete generation params from model config (#41695)

zucchini-nlp committed 279 days ago

Verified b1bdf9cb

Reduce timing on CircleCI - part 1 (Use @slow for IntegrationTests) (#42206)

ydshieh committed 279 days ago

Verified cd416f3c

[loading] Fix device when source and target are different (#42246)

Cyrilvallez committed 279 days ago

Verified 1742d119

Stop inheriting tests (again) (#42247)

Rocketknight1 committed 279 days ago

Verified 16924cd3

Fix UnboundLocalError in RT-DETR loss computation (#42224)

yashwantbezawada committed 279 days ago

Verified 266d3b05

Fix processor test for glm (#42233)

molbap committed 279 days ago

Verified 8d6c4583

FIX Broken PEFT adapter loading (#42187)

BenjaminBossan committed 280 days ago

Verified 2cc9152d

calls `AttentionMaskConverter._unmask_unattended` for xpu device before (#42230)

kaixuanliu committed 280 days ago

Verified 8637f6e7

Loading optimization (#42239)

Cyrilvallez committed 280 days ago

Verified 0e74a71c

Add prefix sharing to continuous batching (#42094)

remi-or committed 280 days ago

Verified 47227f47

Fix TP plans for MoE models (#42236)

Cyrilvallez committed 280 days ago

Verified 7f9f4d9c

Fix initialization guard for pytest (#42234)

Cyrilvallez committed 280 days ago

Verified 462beff5

GLM-V update with new processor (#42122)

zRzRzRzRzRzRzR committed 282 days ago

Verified 66d57110

Much more efficient and clear weight initialization and tie weights (#42191)

Cyrilvallez committed 282 days ago

Verified 8598421b

Update `test_dynamic_cache_exportability_multiple_run` (failing on torch 2.10 nightly) (#42212)

ydshieh committed 283 days ago

Verified 16c7afd0

[`BLT`] Fix cache usage (#42188)

vasqu committed 283 days ago

Verified 309180f9

Newer Older