huggingface/transformers

Pull Requests Commits

[run-slow] gemma2

ydshieh committed 2 years ago

d501489c

[run-slow] gemma2 on ci

ydshieh committed 2 years ago

b843e694

[run-slow] gemma2

ydshieh committed 2 years ago

48e54556

[run-slow] gemma2

ydshieh committed 2 years ago

7c897584

[run_slow] gemma2

ydshieh committed 2 years ago

f4c39ea1

[run_slow] gemma2

ydshieh committed 2 years ago

7d31833a

use a single for loop (#33148)

ArthurZucker committed 2 years ago

Verified c409cd81

Add a static cache that offloads to the CPU or other device (#32161)

gerbenvv committed 2 years ago

Verified 51296712

Mamba2 conversion script for original models (#32580)

vasqu committed 2 years ago

Verified 92a75ff6

pass module to Params4bit.from_prequantized to ensure quant_state (#32524)

winglian committed 2 years ago

Verified 39bfb2f5

added quick clarification (#33166)

DuyguA committed 2 years ago

Verified 5c1027bf

update push CI workflow files for security (#33142)

ydshieh committed 2 years ago

Verified 3d79dcbd

Fix spell mistakes (#33149)

matsuo1234567 committed 2 years ago

Verified 74e19e81

Customise the separator used for splicing in DataCollatorWithFlattening (#33114)

beep-bebop committed 2 years ago

Verified 5c84682f

Zero-shot pipelines: minor doc changes (#33127)

pcuenca committed 2 years ago

Verified f4c86d04

Fix import paths for test_module (#32888)

rasmi committed 2 years ago

Verified f9ed05dd

[RoBERTa-based] Add support for sdpa (#30510)

hackyon committed 2 years ago

Verified f1a385b1

[whisper] pass attention_mask to generate_with_fallback() (#33145)

benniekiss committed 2 years ago

Verified e0b87b0f

Fix: Jamba batched generation (#32914)

vasqu committed 2 years ago

Verified 3bfd3e48

fix model name and copyright (#33152)

mayank31398 committed 2 years ago

Verified 386931d9

Granite language models (#31502)

mayank31398 committed 2 years ago

Verified c35d2ccf

🚨 Add Blip2ForImageTextRetrieval (#29261)

jpizarrom committed 2 years ago

Verified 7591ca5b

Very small change to one of the function parameters (#32548)

alisalamatian1 committed 2 years ago

Verified 27903de7

🌐 [i18n-KO] Translated `conversations.md` to Korean (#32468)

newfull5 committed 2 years ago

Verified 6101d934

update torch req for 4-bit optimizer (#33144)

SunMarc committed 2 years ago

Verified 7ee4363d

fix redundant checkpointing in example training scripts (#33131)

eminorhan committed 2 years ago

Verified d47a9e8c

Llama: make slow tests green 🟢 (#33138)

gante committed 2 years ago

Verified c6b23fda

Add a fix for custom code tokenizers in pipelines (#32300)

Rocketknight1 committed 2 years ago

Verified 9956c2bc

fix Idefics2VisionConfig type annotation (#33103)

chenzizhao committed 2 years ago

Verified 834ec7b1

Update stateful_callbacks state before saving checkpoint (#32115)

brs-pt committed 2 years ago

Verified d1f39c48

Older