huggingface/transformers

Pull Requests Commits

Flash-attn performance: remove cuda sync during inference (#33570)

Cyrilvallez committed 1 year ago

Verified 1f33023c

Add position ids in forward pass to opt model (#33121)

avishaiElmakies committed 1 year ago

Verified 4953ddf0

[WIP] Add Tokenizer for MyT5 Model (#31286)

tomlimi committed 1 year ago

Verified 1bd604d1

[`TF`] Fix Tensorflow XLA Generation on limited seq_len models (#33903)

vasqu committed 1 year ago

Verified 5ef432e4

Bug fix gguf qwen2moe (#33940)

VladOS95-cyber committed 1 year ago

Verified 22e102ad

add test for Jamba with new model jamba-tiny-dev (#33863)

yecohn committed 1 year ago

Verified 56be9f19

Updating `char_to_token` documentation to note behaviour when `trim_offsets` is True (#33919)

Craigacp committed 1 year ago

Verified a7e4e1a7

Paligemma: fix static cache test (#33941)

zucchini-nlp committed 1 year ago

Verified 612065ef

Cache: revert DynamicCache init for BC (#33861)

gante committed 1 year ago

Verified 38f9f10d

fix red check-copies (#33964)

ArthurZucker committed 1 year ago

Verified f92d3548

Add Zamba (#30950)

pglorio committed 1 year ago

Verified f319ba16

PhiMoE (#33363)

garg-amit committed 1 year ago

Verified e3775539

hot fix `self.position_embeddings->self.position_embedding` (#33958)

ArthurZucker committed 1 year ago

Verified 46579c0e

Fix attn mask ignore logic in training-time trace (#32613)

zhenglongjiepheonix committed 1 year ago

Verified 0d1692a4

Removed unnecessary transpose in Switch Transformer Routing (#33582)

karan-uppal3 committed 1 year ago

Verified 614660fd

🔴 🚨 Resizing tokens embeddings: initialize from old embeddings' normal distribution. (#33325)

abuelnasr0 committed 1 year ago

Verified 78ef5832

Enables CPU AWQ model with IPEX version. (#33460)

jiqing-feng committed 1 year ago

Verified b916efcb

Add a section on writing tool templates to the chat template docs (#33924)

Rocketknight1 committed 1 year ago

Verified de4112e4

[`PR run-slow`] (#33939)

ArthurZucker committed 1 year ago

Verified 2e719e35

Ignore keys on `validate_rope` (#33753)

zucchini-nlp committed 1 year ago

Verified 061c2c4c

[i18n-ru] Fixes typo in the README_ru.md (#33882)

Artanias committed 1 year ago

Verified 4a173b88

[Doc]: Broken link in Kubernetes doc (#33879)

saldanhad committed 1 year ago

Verified b6a01df6

Fix distil whisper segment computation (#33920)

ylacombe committed 1 year ago

Verified 124713c3

Minor error condition bug fix (#33781)

htahboub committed 1 year ago

Verified 2bd4d589

Remove `logits.float()` (#33902)

Matthew Hoffman committed 1 year ago

Verified 550673a7

Uniformize kwargs for Idefics/2 processors (#32568)

yonigozlan committed 1 year ago

Verified 074aa3b3

Config: lower `save_pretrained` exception to warning (#33906)

gante committed 1 year ago

Verified b0c5660e

Add support for `weights_only` flag when loading state_dict (#32481)

jerryzh168 committed 1 year ago

Verified 15a4d248

add setter for trainer processor (#33911)

ArthurZucker committed 1 year ago

Verified a220c5b9

[PEFT] Support low_cpu_mem_usage option for PEFT loading adapters (#33725)

BenjaminBossan committed 1 year ago

Verified 6500f78c

Newer Older