release 0.43.3

CHANGELOG.md

+### 0.43.3
+
+#### Improvements:
+
+- FSDP: Enable loading prequantized weights with bf16/fp16/fp32 quant_storage
+    - Background: This update, linked to [Transformer PR #32276](https://github.com/huggingface/transformers/pull/32276), allows loading prequantized weights with alternative storage formats. Metadata is tracked similarly to `Params4bit.__new__` post PR #970. It supports models exported with non-default `quant_storage`, such as [this NF4 model with BF16 storage](https://huggingface.co/hugging-quants/Meta-Llama-3.1-405B-BNB-NF4-BF16).
+    - Special thanks to @winglian and @matthewdouglas for enabling FSDP+QLoRA finetuning of Llama 3.1 405B on a single 8xH100 or 8xA100 node with as little as 256GB system RAM.
+
+
 ### 0.43.2
 
 This release is quite significant as the QLoRA bug fix big implications for higher `seqlen` and batch sizes.

bitsandbytes/__init__.py

@@ -21,4 +21,4 @@ __pdoc__ = {
     "optim.optimizer.MockArgs": False,
 }
 
-__version__ = "0.43.3.dev"
+__version__ = "0.43.3"

setup.py

@@ -25,7 +25,7 @@ class BinaryDistribution(Distribution):
 
 setup(
     name="bitsandbytes",
-    version="0.43.3.dev",
+    version="0.43.3",
     author="Tim Dettmers",
     author_email="dettmers@cs.washington.edu",
     description="k-bit optimizers and matrix multiplication routines.",