Papers › BitDistiller: Unleashing the Potential of Sub-4-Bit LLMs via Self-Distillation

BitDistiller: Unleashing the Potential of Sub-4-Bit LLMs via Self-Distillation

16 Feb 2024arXiv:2402.10631archive 2025-07-28

Dayou Du, Yijia Zhang, Shijie Cao, Jiaqi Guo, Ting Cao, Xiaowen Chu, Ningyi Xu

The upscaling of Large Language Models (LLMs) has yielded impressive advances in natural language processing, yet it also poses significant deployment challenges. Weight quantization has emerged as a widely embraced solution to reduce memory and computational demands. This paper introduces BitDistiller, a framework that synergizes Quantization-Aware Training (QAT) with Knowledge Distillation (KD) to boost the performance of LLMs at ultra-low precisions (sub-4-bit). Specifically, BitDistiller first incorporates a tailored asymmetric quantization and clipping technique to maximally preserve the fidelity of quantized weights, and then proposes a novel Confidence-Aware Kullback-Leibler Divergence (CAKLD) objective, which is employed in a self-distillation manner to enable faster convergence and superior model performance. Empirical evaluations demonstrate that BitDistiller significantly surpasses existing methods in both 3-bit and 2-bit configurations on general language understanding and complex reasoning benchmarks. Notably, BitDistiller is shown to be more cost-effective, demanding fewer data and training resources. The code is available at https://github.com/DD-DuDa/BitDistiller.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2402.10631")

Code

Syntology Ran 7 of 15 code samples harvested from 1 repository linked to this paper; 8 have no recorded run. Of those that ran: 1 ran · violated contract; 1 ran · fixture could not drive it; 5 ran with no contract checked.

By repository: official repository: 15 samples from 1 repository, 7 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

dd-duda/bitdistiller officialmentioned in papermentioned on GitHubpytorchMIT report
microsoft/bitblas mentioned on GitHubpytorchMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

15 samples harvested; 7 ran; 0 honoured the contract we drafted; 8 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · violated contract
1ran · fixture could not drive it
5ran
8unverified

Licence: 0 of the 15 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from dd-duda/bitdistiller. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

build_alibi_bias dd-duda/bitdistiller/inference/models/mpt.py official repository ran MIT (permissive) · f9ef8eeb630a0972 · report
calculate_zeros_width dd-duda/bitdistiller/quantization/qmodule.py official repository ran MIT (permissive) · fd8719af3db2c5ae · report
gen_slopes dd-duda/bitdistiller/inference/models/mpt.py official repository ran fingerprinted MIT (permissive) · 5dfd0c2387dff573 · report
get_named_linears dd-duda/bitdistiller/quantization/pre_quant.py official repository ran MIT (permissive) · 0310065bdf9208bb · report
make_divisible dd-duda/bitdistiller/quantization/qmodule.py official repository ran fingerprinted MIT (permissive) · ad85b288e9acb9c2 · report
precompute_freqs_cis dd-duda/bitdistiller/inference/models/llama.py official repository ran · violated contract MIT (permissive) · 14a84c2cbfebc413 · report
reshape_for_broadcast dd-duda/bitdistiller/inference/models/llama.py official repository ran · fixture could not drive it fingerprinted MIT (permissive) · 70bf6ebaafd266c4 · report
apply_rotary_emb dd-duda/bitdistiller/inference/models/llama.py official repository unverified MIT (permissive) · 797af70a155dd94f · report
convertModelToQuant dd-duda/bitdistiller/quantization/qlinear.py official repository unverified MIT (permissive) · daff7fa3b7b944d2 · report
get_blocks dd-duda/bitdistiller/quantization/pre_quant.py official repository unverified MIT (permissive) · fa80a6587641fb46 · report
get_calib_dataset dd-duda/bitdistiller/quantization/clip_utils.py official repository unverified MIT (permissive) · a1ca17436132cd8f · report
get_calib_dataset_code dd-duda/bitdistiller/quantization/clip_utils.py official repository unverified MIT (permissive) · eca59c8a0da72a0c · report
get_pile_dataset dd-duda/bitdistiller/quantization/clip_utils.py official repository unverified MIT (permissive) · 05aeb596afd8012e · report
rotate_half dd-duda/bitdistiller/inference/models/falcon.py official repository unverified MIT (permissive) · 766ad0e51b7a8f09 · report
run_awq dd-duda/bitdistiller/quantization/pre_quant.py official repository unverified MIT (permissive) · fa755c29499d4073 · report

Tasks

Knowledge DistillationQuantization

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Knowledge Distillation

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections