SAVRN
Search Contact SAVRN

Open-weight model · Text generation

dQwen3.5-0.8B-Base

by IFML UT-IFML/dQwen3.5-0.8B-Base

A masked diffusion language model adapted from Qwen3.5-0.8B. The backbone is hybrid: only its attention layers are made bidirectional, and the Gated DeltaNet layers stay causal. This is a base model, with no instruction tuning.

Parameters752M
Context262,144
Weights1.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads435

Runs On

What it takes to serve dQwen3.5-0.8B-Base (752M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 1.5 GB 1.8 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 0.8 GB 0.9 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 0.4 GB 0.5 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

Model Card

By IFML, published under apache-2.0, revision 6dbda9fa4b46.

A masked diffusion language model adapted from Qwen3.5-0.8B. The backbone is hybrid: only its attention layers are made bidirectional, and the Gated DeltaNet layers stay causal. This is a base model, with no instruction tuning. Paper: dQwen3.5: Hybrid-Attention Diffusion Language Models. Code: https://github.com/AntonXue/dQwen Needs a CUDA GPU and transformers>=5.13 (tested with torch 2.7.1+cu128, flash-linear-attention 0.5.1). generate decodes the whole canvas at once, committing positions above a confidence threshold (tau=0.9); pass blocklength=32 for left-to-right block decoding, or tau=None, stepsperblock=k for a fixed budget. The 50B-token checkpoint from the paper is…

Read IFML's full model card

A masked diffusion language model adapted from Qwen3.5-0.8B. The backbone is hybrid: only its attention layers are made bidirectional, and the Gated DeltaNet layers stay causal. This is a base model, with no instruction tuning.

Paper: dQwen3.5: Hybrid-Attention Diffusion Language Models. Code: https://github.com/AntonXue/dQwen

The dQwen3.5 family

model total params backbone
dQwen3.5-0.8B-Base (this repo) 0.75B hybrid
dQwen3.5-2B-Base 1.88B hybrid
dQwen3.5-4B-Base 4.21B hybrid
dQwen3.5-9B-Base 8.95B hybrid
dQwen3-1.7B-Base 1.72B full attention (control)

Quickstart

Needs a CUDA GPU and transformers>=5.13 (tested with torch 2.7.1+cu128, flash-linear-attention 0.5.1).

import torch
from transformers import AutoModel

model = AutoModel.from_pretrained("UT-IFML/dQwen3.5-0.8B-Base", trust_remote_code=True, dtype=torch.bfloat16).cuda().eval()
print(model.generate("def fibonacci(n):", gen_length=512, stop_strings=["\ndef "]).text)

generate decodes the whole canvas at once, committing positions above a confidence threshold (tau=0.9); pass block_length=32 for left-to-right block decoding, or tau=None, steps_per_block=k for a fixed budget. The 50B-token checkpoint from the paper is revision="step25000-swa".

Citation

@article{xue2026dqwen,
  title   = {dQwen3.5: Hybrid-Attention Diffusion Language Models},
  author  = {Xue, Anton and Rout, Litu and Akella, Aditya and Klivans, Adam and Sanghavi, Sujay and Shakkottai, Sanjay},
  journal = {arXiv preprint arXiv:2609.20751},
  year    = {2026}
}

Configuration

Architecture
DiffuQwen35
Context length (tokens)
262,144
Layers
24
Hidden size
1,024
Feed-forward size
3,584
Attention heads
8
Key/value heads
2
Head dimension
256
Vocabulary size
248,320
Model type
diffuqwen35

Identity and Version

Repository
UT-IFML/dQwen3.5-0.8B-Base
Publisher
IFML
Task
Text generation
Modality
Text
Library
transformers
Parameters
752M parameters
Languages
Not stated by the source
Revision
6dbda9fa4b46a39c70898f71b22cc35b09ebdba9
First published
2026-08-03
Last updated
2026-09-18

Files and Weights

11 files, 1.5 GB in total. The weights are 1 file totalling 1.5 GB in safetensors.

Weights1 file · 1.5 GB
Configuration4 files · 17.0 KB
Tokenizer2 files · 20.0 MB
Documentation1 file · 2.1 KB
Other2 files · 1.1 MB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model.safetensorsWeights1.5 GB ef14cffbe345
argparse.jsonConfiguration868 B —
config.jsonConfiguration1.9 KB —
generation_config.jsonConfiguration179 B —
modeling_dqwen3_5.pyConfiguration14.0 KB —
README.mdDocumentation2.1 KB —
chat_template.jinjaOther7.8 KB —
training_log.jsonlOther1.1 MB —
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer20.0 MB 06b9509352d2
tokenizer_config.jsonTokenizer1.2 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
1.5 GB
Download from IFML

Released by IFML through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published1.5 GB
16-bit1.5 GB
8-bit0.8 GB
4-bit0.4 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About dQwen3.5-0.8B-Base

How much GPU memory does dQwen3.5-0.8B-Base need?

About 1.8 GB at 16-bit and 0.5 GB at 4-bit: the weights (752M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run dQwen3.5-0.8B-Base on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use dQwen3.5-0.8B-Base commercially?

Yes. dQwen3.5-0.8B-Base is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is dQwen3.5-0.8B-Base's context length?

262,144 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Text generation

Qwen3-0.6B

Qwen

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features: - Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios. - Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and…

Open weights apache-2.0 752M parameters 40,960 tokens transformers
View model

A 0.6B parameter model built in two stages: knowledge distillation from a 30B Thinking teacher to establish a structured reasoning backbone, then supervised fine-tuning on legal instruction data. 50x compression. Under 500MB quantized. Runs on a phone. The training order is the thesis: teach the model how to reason first (distillation from Thinking teacher), then teach it what to reason about (legal SFT). The Thinking teacher's extended deliberation traces transfer deeper reasoning structure than an Instruct teacher — critical when the student has only 0.6B parameters to work with. Qwen3-0.6B distilled from Qwen3-30B-A3B-Thinking-2507 — a Mixture-of-Experts model with 30B total parameters…

Open weights apache-2.0 752M parameters 40,960 tokens transformers
View model

A 0.6B parameter model distilled from Qwen3-30B-A3B-Thinking on 6,122 STEM chain-of-thought samples. 50x parameter compression. The Thinking variant teacher produces richer extended reasoning traces than the Instruct variant, transferring deeper deliberation structure into the smallest possible student. The result: a model under 500MB quantized that produces structured STEM derivations because a 30B thinking model showed it how to reason. Two key differences from standard small-model distillation: 1. Thinking teacher, not Instruct teacher. The Qwen3-30B-A3B-Thinking variant generates extended internal reasoning before committing to an answer. Its softmax distributions are higher-entropy…

Open weights apache-2.0 752M parameters 40,960 tokens transformers
View model

Model · Text generation

gelatwo-common-gen-gpt2-large

Meihua Dang

gpt2-large fine-tuned on CommonGen, used as the base language model for the CommonGen experiments in Mitigating Bias in Locally Constrained Decoding via Tractable Proposals (arXiv:2606.01926). This is a plain causal language model: it supplies the base distribution that GCD and P-GCD steer. The tractable proposal it is paired with is the HMM at which shares its 50257-token vocabulary. configs/common-gen.yaml in github.com/MhDang/gelatwo already points at this checkpoint, so the CommonGen runs need no override.

Open weights mit 774M parameters
View model

Model · Text generation

gpt2-large

OpenAI community

GPT-2 Large is the 774M parameter version of GPT-2, a transformer-based language model created and released by OpenAI. The model is a pretrained model on English language using a causal language modeling (CLM) objective. - Test the full generation capabilities here: https://transformer.huggingface.co/doc/gpt2-large Use the code below to get started with the model. You can use this model directly with a pipeline for text generation. Since the generation relies on some randomness, we Here is how to use this model to get the features of a given text in PyTorch: In their model card about GPT-2, OpenAI wrote: In their model card about GPT-2, OpenAI wrote: In their model card about GPT-2, OpenAI…

Open weights mit 812M parameters transformers
View model

Model · Text generation

Mini-oss-0.6b

Convergent Intelligence

This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019). This model is part of the Convergent Intelligence LLC: Research Division portfolio. All models in this portfolio are developed under the Discrepancy Calculus (DISC) framework — a measure-theoretic approach to understanding and controlling the gap between what a model should produce and what it actually produces. DISC treats training…

Open weights 664M parameters 131,072 tokens transformers
View model