Pull Requests huggingface/optimum-habana

Use environment variable for doc build token

#2360 opened 2026-01-06 09:35 by paulinebm

Refactor slow_tests_gaudi2.yml workflow configuration

#2354 opened 2025-12-09 09:34 by glegendre01

#2352 opened 2025-12-01 21:39 by mkpatel3-github

add flux kontext related files

#2336 opened 2025-11-04 11:02 by tianyuan0211

Temporarily remove broken examples for OH 1.20.0 release

#2329 opened 2025-10-29 13:16 by gplutop7

Enable Int64 support for trl-sft workload.

#2311 opened 2025-10-17 05:05 by AKloniecki

Backport transpose_for_scores function in modeling xlm roberta.

#2300 opened 2025-10-08 12:14 by AKloniecki

transformer qwen3 Porting to optimum-habana

#2234 opened 2025-08-31 16:53 by rkumar2patel

Enable Mamba to use PScan Kernel

#2177 opened 2025-07-29 03:32 by slokesha

Model phi performance fix

#2165 opened 2025-07-22 00:26 by slokesha

Update datasets that do not work with `datasets > 4.0.0`

#2164 opened 2025-07-21 13:15 by regisss

Upgrade to TRL 0.17.0 version

#2130 opened 2025-07-10 14:42 by alekseyfa

Fix a version check of pytorch when fsdp plugin is used

#1777 opened 2025-02-12 23:19 by jojivk73

Multi node Multi card MPIJob

#1702 opened 2025-01-17 18:35 by sramakintel

[llama] Store KV Cache on CPU and Use PyTorch `SPDA` for Next token generation

#1182 opened 2024-08-02 02:23 by zhentaoyu