SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2304.13013 · NeurIPS · 2023

Stable and low-precision training for large-scale vision-language models

Mitchell Wortsman, Ali Farhadi, Ari Morcos, Luke Zettlemoyer, Ludwig Schmidt, Tim Dettmers

arXiv · PDF · Open in the Atlas

Code that ran

None of the 3 samples (functions and classes) lifted from this paper's linked repositories has a recorded run on a synthesized input (3 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2304.13013.json · badge: /api/ran/2304.13013.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
TimDettmers/bitsandbytes community
Papers with Code archive
0 ran · 3 no recorded run
FunctionStatusWhere it lives
Int8Params Not yet run
unverified (harvested, no recorded run)
TimDettmers/bitsandbytes/bitsandbytes/nn/modules.py
code served (permissive licence) · get_code("419fc400d6ba95b4")
How it ran

no invocation recorded.

Linear8bitLt Not yet run
unverified (harvested, no recorded run)
TimDettmers/bitsandbytes/bitsandbytes/nn/modules.py
code served (permissive licence) · get_code("610984316a762af9")
How it ran

no invocation recorded.

maybe_rearrange_weight Not yet run
unverified (harvested, no recorded run)
TimDettmers/bitsandbytes/bitsandbytes/nn/modules.py
code served (permissive licence) · get_code("5bd645c0fa9375ce")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 3 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We introduce new methods for 1) accelerating and 2) stabilizing training for large language-vision models. 1) For acceleration, we introduce SwitchBack, a linear layer for int8 quantized training which provides a speed-up of 13-25% while matching the performance of bfloat16 training within 0.1 percentage points for the 1B parameter CLIP ViT-Huge-the largest int8 training to date. Our main focus is int8 as GPU support for float8 is rare, though we also analyze float8 training through simulation. While SwitchBack proves effective for float8, we show that standard techniques are also successful if the network is trained and initialized so that large feature magnitudes are discouraged, which we accomplish via layer-scale initialized with zeros. 2) For stabilization, we analyze loss spikes and find they consistently occur 1-8 iterations after the squared gradients become under-estimated by their AdamW second moment estimator. As a result, we recommend an AdamW-Adafactor hybrid which avoids loss spikes when training a CLIP ViT-Huge model and outperforms gradient clipping at the scales we test.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2304.13013")
get_code_for_paper("2304.13013")
have("2304.13013")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2304.13013.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2304.13013.svg)](https://syntology.ai/paper/2304.13013)

Connect an agent — have() is free.