liger-kernel-nightly 0.6.0.dev20250709042125__py3-none-any.whl → 0.6.0.dev20250714030807__py3-none-any.whl

liger_kernel/transformers/model/gemma3.py

@@ -255,7 +255,7 @@ def multimodal_forward(
         shift_labels = shift_labels.view(-1).to(hidden_device)
 
         lce = LigerFusedLinearCrossEntropyLoss()
-        loss = lce(self.language_model.lm_head.weight, shift_hidden_states, shift_labels)
+        loss = lce(self.lm_head.weight, shift_hidden_states, shift_labels)
     else:
         logits = self.lm_head(kept_hidden_states)
         if labels is not None:

{liger_kernel_nightly-0.6.0.dev20250709042125.dist-info → liger_kernel_nightly-0.6.0.dev20250714030807.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: liger_kernel_nightly
-Version: 0.6.0.dev20250709042125
+Version: 0.6.0.dev20250714030807
 Summary: Efficient Triton kernels for LLM Training
 License: BSD 2-CLAUSE LICENSE
         Copyright 2024 LinkedIn Corporation 
@@ -411,7 +411,7 @@ loss.backward()
 
 - For issues, create a Github ticket in this repository
 - For open discussion, join [our discord channel on GPUMode](https://discord.com/channels/1189498204333543425/1275130785933951039)
-- For formal collaboration, send an email to yannchen@linkedin.com and hning@linkedin.com
+- For formal collaboration, send an email to yannchen@linkedin.com and zhipwang@linkedin.com
 
 ## Cite this work
 
@@ -419,13 +419,13 @@ Biblatex entry:
 ```bib
 @article{hsu2024ligerkernelefficienttriton,
       title={Liger Kernel: Efficient Triton Kernels for LLM Training},
-      author={Pin-Lun Hsu and Yun Dai and Vignesh Kothapalli and Qingquan Song and Shao Tang and Siyu Zhu and Steven Shimizu and Shivam Sahni and Haowen Ning and Yanning Chen},
+      author={Pin-Lun Hsu and Yun Dai and Vignesh Kothapalli and Qingquan Song and Shao Tang and Siyu Zhu and Steven Shimizu and Shivam Sahni and Haowen Ning and Yanning Chen and Zhipeng Wang},
       year={2024},
       eprint={2410.10989},
       archivePrefix={arXiv},
       primaryClass={cs.LG},
       url={https://arxiv.org/abs/2410.10989},
-      journal={arXiv preprint arXiv:2410.10989},
+      journal={Proceedings of the ICML 2025 Workshop on Championing Opensource Development in Machine Learning (CODEML ’25)},
 }
 ```
 

{liger_kernel_nightly-0.6.0.dev20250709042125.dist-info → liger_kernel_nightly-0.6.0.dev20250714030807.dist-info}/RECORD

@@ -68,7 +68,7 @@ liger_kernel/transformers/experimental/embedding.py,sha256=2P0QYdlFyFrG5OqTzTa1w
 liger_kernel/transformers/model/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 liger_kernel/transformers/model/gemma.py,sha256=mNX-mIwV6jI4zfbrUHp0C468pOmjzsL7mjXipGt-eS0,10007
 liger_kernel/transformers/model/gemma2.py,sha256=R_JFPyWTk7RyA7D05ZiIaNO5pX8gWcvfWf-6rdCRMxs,11296
-liger_kernel/transformers/model/gemma3.py,sha256=XbwoqOSPmtS0BPHgT8jZftTzplmiAicgBa6ocNcet8o,12800
+liger_kernel/transformers/model/gemma3.py,sha256=FKO4j3t4W_5uECRA1lhVnXC-It2GhirHm4tpCf9ApAc,12785
 liger_kernel/transformers/model/glm4.py,sha256=GlnEhdGJuDIqp2R9qC54biY3HwV1tWmfpJm6ijoAsrM,5257
 liger_kernel/transformers/model/llama.py,sha256=i8jJgyZsMKWQ-zKloETLugtwFpUOdaWxLDceciFXKd4,12832
 liger_kernel/transformers/model/llama4.py,sha256=IgbB8sTh3dlETQnaNNy1bZLuXy-Nt7qmeAjF27ydGpg,4210
@@ -89,9 +89,9 @@ liger_kernel/transformers/trainer/__init__.py,sha256=p7yQfklV8-467qSz_ZMimkbDF7H
 liger_kernel/transformers/trainer/orpo_trainer.py,sha256=tX0h63aOFe3rNqTmk6JpMf75UPo981yzEa6TghnjS0Q,5370
 liger_kernel/triton/__init__.py,sha256=qCiCamzCRv6lpV8IqpAc9YMdNKC7GKurClWceQPnlis,92
 liger_kernel/triton/monkey_patch.py,sha256=Rd0hUHAzDkFfHvnX7-PBaNK5EKnZhtfM_h-fgQH9HPY,1568
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/LICENSE,sha256=OhzLDHJ0to4a8sodVLELZiCFylZ1NAAYLs-HrjPy0ag,1312
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/METADATA,sha256=wPo1-0a5kmFfk7NvN28x3Sk_NV0g4AEzrOnbB4fJHMk,24535
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/NOTICE,sha256=njwnoPZLh9AN8SJQzxvCGLHi-8X__AvWRze6joNXIY8,2066
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/WHEEL,sha256=iAkIy5fosb7FzIOwONchHf19Qu7_1wCWyFNR5gu9nU0,91
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/top_level.txt,sha256=2eghu4hA3LnkM7ElW92tQ8zegWKgSbeo-k-aGe1YnvY,13
-liger_kernel_nightly-0.6.0.dev20250709042125.dist-info/RECORD,,
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/LICENSE,sha256=OhzLDHJ0to4a8sodVLELZiCFylZ1NAAYLs-HrjPy0ag,1312
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/METADATA,sha256=OSGl46Gr7NnD-8gSkMhIczRaOmpJFM_W6pc28IbIHm0,24634
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/NOTICE,sha256=njwnoPZLh9AN8SJQzxvCGLHi-8X__AvWRze6joNXIY8,2066
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/WHEEL,sha256=iAkIy5fosb7FzIOwONchHf19Qu7_1wCWyFNR5gu9nU0,91
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/top_level.txt,sha256=2eghu4hA3LnkM7ElW92tQ8zegWKgSbeo-k-aGe1YnvY,13
+liger_kernel_nightly-0.6.0.dev20250714030807.dist-info/RECORD,,