bitsandbytes-foundation/bitsandbytes · Discussions · GitHub

X Tutup

Discussions

You must be logged in to vote

LARS and LAMB issues related to str2optimizer32bit

RasmusHoier asked Nov 7, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Multi-backend refactor: Alpha release ( AMD ROCm ONLY )

Titus-von-Koeller started Aug 29, 2024 in Refactorings

15
You must be logged in to vote

Any plans to implement mxfp4 and/or nvfp4?

davidmezzetti asked Oct 16, 2025 in Q&A · Answered

2
You must be logged in to vote

Does low bit quantization affect convergence?

micklexqg asked Aug 12, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Frequently Asked Questions

Titus-von-Koeller asked Feb 1, 2024 in Q&A · Answered

18
You must be logged in to vote

[Feature Idea] Support for 1.58-bit Ternary Quantization (BitNet)

ratchanonth60 asked Jun 10, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Multi-backend support: Apple Silicon / Mac (call for contributors + help fleshing out the details)

Titus-von-Koeller started Aug 29, 2024 in Refactorings

20
You must be logged in to vote

BitsAndBytes with CUDA 12.8 support on Windows

rand55 asked May 7, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Efficient Permute/Transpose Operations on 4-Bit Quantized Tensors

yhyang201 asked Apr 11, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Alternatives to BitsAndBytes for HF models

Timelessprod asked Aug 28, 2024 in Q&A · Unanswered

1
You must be logged in to vote

FP4 capabilities Blackwell GPUs

Settek-git asked Mar 10, 2025 in Q&A · Unanswered

0
You must be logged in to vote

Is it possible to support nvfp4 in blackwell ?
Feature Request
edisonchan started Feb 27, 2025 in General

1
You must be logged in to vote

Building on NVidia GH-200 - Is this a supported platform?

simusid started Feb 19, 2025 in General

1
You must be logged in to vote

Multi-backend refactor: Alpha release ( INTEL ONLY )

Titus-von-Koeller started Aug 29, 2024 in Refactorings

27
You must be logged in to vote

The confusion surrounding bnb_4bit_compute_dtype, torch_dtype, and prepare_model_for_kbit_training.

xiaobingbuhuitou asked Feb 14, 2025 in Q&A · Unanswered

0
You must be logged in to vote

CUDA Capabilities Selected & CUDA Targets - issue

LDec-AlmostHuman started Feb 9, 2025 in Dev Corner

0
You must be logged in to vote

warning inputs will be cast from X to float16 during quantization

mluerig asked Nov 22, 2024 in Q&A · Unanswered

1
You must be logged in to vote

Improvement suggestions for the multi-backend-refactor installation instructions

Titus-von-Koeller asked May 24, 2024 in CATCH-ALL: alpha testing the `multi-backend-refactor` · Unanswered

26
You must be logged in to vote

bitsandbytes is not playing well with kohya_ss

joshgura started Dec 12, 2024 in General

0
You must be logged in to vote

cuda 12.6

eljandoubi asked Oct 23, 2024 in Q&A · Answered

2
You must be logged in to vote

Bitsandbytes quantization and QLORA fine-tuning

nikoladj99 started Oct 13, 2024 in General

0
You must be logged in to vote

v0.44.0: New AdEMAMix optimizer, Embeddings quantization, and more!

Titus-von-Koeller announced Sep 29, 2024 in Announcements

0
You must be logged in to vote

How can i correctly use a CPU to perform inference of a quantized model

neavo asked Sep 27, 2024 in CATCH-ALL: alpha testing the `multi-backend-refactor` · Unanswered

0
You must be logged in to vote

Welcome to bitsandbytes Discussions!

TimDettmers announced Jan 26, 2024 in Announcements

1
You must be logged in to vote

Quantize a single tensor obtained from a float32 model

Boltzmachine asked Sep 18, 2024 in Q&A · Unanswered

0

X Tutup