Papers › Scaling Diffusion Language Models via Adaptation from Autoregressive Models

Scaling Diffusion Language Models via Adaptation from Autoregressive Models

23 Oct 2024arXiv:2410.17891archive 2025-07-28

Shansan Gong, Shivam Agarwal, Yizhe Zhang, Jiacheng Ye, Lin Zheng, Mukai Li, Chenxin An, Peilin Zhao, Wei Bi, Jiawei Han, Hao Peng, Lingpeng Kong

Diffusion Language Models (DLMs) have emerged as a promising new paradigm for text generative modeling, potentially addressing limitations of autoregressive (AR) models. However, current DLMs have been studied at a smaller scale compared to their AR counterparts and lack fair comparison on language modeling benchmarks. Additionally, training diffusion models from scratch at scale remains challenging. Given the prevalence of open-source AR language models, we propose adapting these models to build text diffusion models. We demonstrate connections between AR and diffusion modeling objectives and introduce a simple continual pre-training approach for training diffusion models. Through systematic evaluation on language modeling, reasoning, and commonsense benchmarks, we show that we can convert AR models ranging from 127M to 7B parameters (GPT2 and LLaMA) into diffusion models DiffuGPT and DiffuLLaMA, using less than 200B tokens for training. Our experimental results reveal that these models outperform earlier DLMs and are competitive with their AR counterparts. We release a suite of DLMs (with 127M, 355M, and 7B parameters) capable of generating fluent text, performing in-context learning, filling in the middle without prompt re-ordering, and following instructions https://github.com/HKUNLP/DiffuLLaMA.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 22 code samples harvested from 1 repository linked to this paper: 16 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 2 honoured, 0 violated, 6 with no contract checked; 8 where Syntology's instrument failed) · 6 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; 3 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 22 samples from 1 repository, 16 ran (8 with no instrument failure; 8 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 3 of the 16 samples that ran have no recorded date.

hkunlp/diffullama officialmentioned in papermentioned on GitHubpytorchSyntology: 22 samples, 16 ran (8 instrument-failed) · 6 unverified 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

22 samples harvested: 16 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 2 honoured, 0 violated, 6 with no contract checked; 8 where Syntology's instrument failed) · 6 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · honoured contract
5ran · our draft was wrong
3ran; the contract check could not run
6ran
6unverified

Licence: 22 of the 22 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 hkunlp/diffullama. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. For 1 sample that ran, the graph records a signature this page does not show: an argument's recorded type or element kind is a printout of an object in Syntology's harness (such as a mock object, a class or a pandas printout), not the type of an input, so it does not describe the call's shape. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

hkunlp/diffullama official repository 22 samples, 16 ran (8 instrument-failed) · 6 unverified
repeat_kv in 244 papers
census: 15 implementations from 60 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 4, kind float, dtype float32
DiffuLLaMA-training/model_llama.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted no licence file found · pointer only · 30d7eec482ebf6b1 · report fixture: signature hidden_states:Tensor/4/float, n_rep:int/0/int · returned Tensor · level recorded, date not recorded
apply_rotary_pos_emb in 393 papers
census: 14 implementations from 15 papers, 8 distinct outputs across 7 buckets, one shared input per bucket
DiffuLLaMA-training/model_llama.py ran · our draft was wrong no licence file found · pointer only · bac65c3dafaec040 · report fixture: signature q:Tensor/4/float, k:Tensor/4/float, cos:Tensor/3/float, sin:Tensor/3/float, position_ids:NoneType/0/?, unsqueeze_dim:int/0/int · returned tuple · level recorded, date not recorded
code in 14 papers DiffuLLaMA-training/packed_dataset.py ran · honoured contract no licence file found · pointer only · fa298642043e1282 · report fixture: returned int · level recorded 2026-09-24
extract_local DiffuLLaMA-training/easy_context/dist_flash_attn/prepare_input.py ran no licence file found · pointer only · 64ec6a16fb572f44 · report fixture: signature value:Tensor/2/float, rank:int/0/int, world_size:int/0/int, device:str/0/str, dim:int/0/int · returned Tensor · level recorded 2026-09-26
extract_local DiffuLLaMA-training/easy_context/ulysses_attn/prepare_inputs.py ran no licence file found · pointer only · 5e09c24645115dd9 · report fixture: signature value:Tensor/3/float, rank:int/0/int, world_size:int/0/int, device:str/0/str, dim:int/0/int · returned Tensor · level recorded 2026-09-26
generate_samples in 16 papers model.py ran no licence file found · pointer only · 2d05c810fd8d785d · report fixture: signature model:callable/0/?, diff_args:DiffArgs/0/?, tokenizer:MinimalTokenizer/0/?, inputs:dict/0/?, verbose:bool/0/bool · returned Tensor · level recorded 2026-09-26
get_anneal_attn_mask evaluation/eval-diffullama.py ran no licence file found · pointer only · ee7bd71eac25f44d · report fixture: signature seq_len:int/0/int, bsz:int/0/int, dtype:dtype/0/?, device:device/0/?, attn_mask_ratio:float/0/float · returned Tensor · level recorded 2026-09-26
maybe_get_set_global_memory_buffer DiffuLLaMA-training/easy_context/dist_flash_attn/async_communication.py ran · honoured contract fingerprinted no licence file found · pointer only · ce7333f42d88b05d · report fixture: signature q:Tensor/3/float, k:Tensor/3/float, v:Tensor/3/float, m:Tensor/2/float, l:Tensor/2/float, o:Tensor/3/float · returned tuple · level recorded 2026-09-22
maybe_get_set_global_memory_buffer_bwd DiffuLLaMA-training/easy_context/dist_flash_attn/async_communication.py ran · our draft was wrong fingerprinted no licence file found · pointer only · ee03eab03d798c57 · report fixture: signature dq:Tensor/4/float, dk:Tensor/4/float, dv:Tensor/4/float, q:Tensor/4/float, L:Tensor/4/float, k:Tensor/4/float, v:Tensor/4/float, o:Tensor/4/float, do:Tensor/4/float · returned tuple · level recorded 2026-09-22
maybe_repeat_kv_bwd DiffuLLaMA-training/easy_context/dist_flash_attn/lightseq_async_attn.py ran; the contract check could not run (RAISES) no licence file found · pointer only · d7b82e7e46f9cee8 · report fixture: signature nqh:int/0/int, kv:Tensor/4/float · returned Tensor · level recorded 2026-09-26
maybe_repeat_kv_fwd DiffuLLaMA-training/easy_context/dist_flash_attn/lightseq_async_attn.py ran; the contract check could not run (RAISES) no licence file found · pointer only · 3d82586ab3888495 · report fixture: signature nqh:int/0/int, kv:Tensor/4/float · returned Tensor · level recorded 2026-09-26
prepare_dist_flash_attn_inputs DiffuLLaMA-training/easy_context/dist_flash_attn/prepare_input.py ran no licence file found · pointer only · 9daba25cd657ac06 · report fixture: signature input_ids:Tensor/2/int, position_ids:Tensor/2/int, target_ids:Tensor/2/int, rank:int/0/int, world_size:int/0/int, device:str/0/str · returned dict · level recorded 2026-09-26
prepare_ulysses_attn_inputs DiffuLLaMA-training/easy_context/ulysses_attn/prepare_inputs.py ran no licence file found · pointer only · 737366305cfbf977 · report fixture: signature input_ids:Tensor/2/int, position_ids:Tensor/2/int, target_ids:Tensor/2/int, rank:int/0/int, world_size:int/0/int, device:str/0/str · returned dict · level recorded 2026-09-26
preprocess in 284 papers
census: 11 implementations from 13 papers, 10 distinct outputs across 5 buckets, one shared input per bucket
evaluation/eval-diffullama.py ran · our draft was wrong fingerprinted no licence file found · pointer only · dd2d39c8d8eff4d5 · report fixture: signature text:str/0/str · returned str · level recorded 2026-09-26
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
DiffuLLaMA-training/model_llama.py ran · our draft was wrong fingerprinted no licence file found · pointer only · b99eea6376d1e212 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
top_p_logits in 13 papers evaluation/eval-diffullama.py ran · our draft was wrong fingerprinted no licence file found · pointer only · fd2fe76b57111989 · report fixture: signature logits:Tensor/2/float, p:float/0/float · returned Tensor · level recorded 2026-09-26
checkpoint_end_with_flash_attention DiffuLLaMA-training/easy_context/dist_flash_attn/monkey_patch.py unverified no licence file found · pointer only · 59adcbbcc9bf1cbc · report
checkpoint_last_module DiffuLLaMA-training/easy_context/dist_flash_attn/monkey_patch.py unverified no licence file found · pointer only · 7517e14a19d34e0f · report
get_flash_attn_out_from_global_buffer DiffuLLaMA-training/easy_context/dist_flash_attn/monkey_patch.py unverified no licence file found · pointer only · 89e4348b4cb177bd · report
is_last_time DiffuLLaMA-training/easy_context/dist_flash_attn/async_communication.py unverified no licence file found · pointer only · 4be6bef6c83e9c27 · report
new_decoder_forward DiffuLLaMA-training/easy_context/ulysses_attn/monkey_patch.py unverified no licence file found · pointer only · 2fcefe3ddd3fa8eb · report
new_flash_attn_forward DiffuLLaMA-training/easy_context/ulysses_attn/monkey_patch.py unverified no licence file found · pointer only · 9949ec185b470d6e · report

Tasks

In-Context LearningLanguage ModelingLanguage Modelling

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Diffusion

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