v0.5.9: Adds XPU Setup, GLM-4 & Qwen3 Model Support, Key Bugfixes

What's Changed

• update setup.py for installation on xpu by @faaany in #668
• update XPU CI yaml file to use docker container by @faaany in #669
• Add average_log_prob as an init param for LigerFusedLinearDPOLoss by @vaibhavjindal in #676
• add shift label change by @shivam15s in #683
• remove tests that can pass on XPU by @faaany in #686
• Update mkdocs.yml by @shivam15s in #691
• Fix LigerCrossEntropy reduction='none' by @Tcc0403 in #680
• Support GLM-4 models by @intervitens in #685
• Import glm4_lce_forward locally in function by @vaibhavjindal in #695
• Qwen3 model support by @vaibhavjindal in #692
• Use logits_to_keep logic for training runs by @vaibhavjindal in #696
• increase gemma3 multimodal convergence test loss atol by @shivam15s in #697
• Update pyproject.toml by @shivam15s in #700

New Contributors

• @intervitens made their first contribution in #685

Full Changelog: v0.5.8...v0.5.9