Bumped version for new release.
This commit is contained in:
parent
412fd0e717
commit
a06a0f6a08
27
CHANGELOG.md
27
CHANGELOG.md
|
@ -283,3 +283,30 @@ Bug fixes:
|
|||
- Removed outdated get_cuda_lib_handle calls that lead to errors. #595 Thank you @ihsanturk
|
||||
- Fixed bug where read-permission was assumed for a file. #497
|
||||
- Fixed a bug where prefetchAsync lead to errors on GPUs that do not support unified memory but not prefetching (Maxwell, SM52). #470 #451 #453 #477 Thank you @jllllll and @stoperro
|
||||
|
||||
|
||||
### 0.41.0
|
||||
|
||||
Features:
|
||||
- Added precompiled CUDA 11.8 binaries to support H100 GPUs without compilation #571
|
||||
- CUDA SETUP now no longer looks for libcuda and libcudart and relies PyTorch CUDA libraries. To manually override this behavior see: how_to_use_nonpytorch_cuda.md. Thank you @rapsealk
|
||||
|
||||
Bug fixes:
|
||||
- Fixed a bug where the default type of absmax was undefined which leads to errors if the default type is different than torch.float32. # 553
|
||||
- Fixed a missing scipy dependency in requirements.txt. #544
|
||||
- Fixed a bug, where a view operation could cause an error in 8-bit layers.
|
||||
- Fixed a bug where CPU bitsandbytes would during the import. #593 Thank you @bilelomrani
|
||||
- Fixed a but where a non-existent LD_LIBRARY_PATH variable led to a failure in python -m bitsandbytes #588
|
||||
- Removed outdated get_cuda_lib_handle calls that lead to errors. #595 Thank you @ihsanturk
|
||||
- Fixed bug where read-permission was assumed for a file. #497
|
||||
- Fixed a bug where prefetchAsync lead to errors on GPUs that do not support unified memory but not prefetching (Maxwell, SM52). #470 #451 #453 #477 Thank you @jllllll and @stoperro
|
||||
|
||||
Documentation:
|
||||
- Improved documentation for GPUs that do not support 8-bit matmul. #529
|
||||
- Added description and pointers for the NF4 data type. #543
|
||||
|
||||
User experience:
|
||||
- Improved handling of default compute_dtype for Linear4bit Layers, so that compute_dtype = input_dtype if the input data type is stable enough (float32, bfloat16, but not float16).
|
||||
|
||||
Performance:
|
||||
- improved 4-bit inference performance for A100 GPUs. This degraded performance for A40/RTX3090 and RTX 4090 GPUs slightly.
|
||||
|
|
Loading…
Reference in New Issue
Block a user