This GGUF file is a direct conversion of Wan-AI/Wan2.2-T2V-A14B Since this is a quantized model, all original licensing terms and usage restrictions remain in effect. Usage The model can be used with the ComfyUI custom node ComfyUI-GGUF by city96 Place model files in ComfyUI/models/unet see the GitHub readme for further installation instructions.
vsqa-preview-14b-v2 is an open-weight model for text to video from Memset0, released under Apache License 2.0. Its published files total 84.6 GB.
Preview checkpoint, not yet evaluated. Final student of a 200-step DMD run (E0029 Variant V0403).
Parameters—
Context—
Weights84.6 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—
Model Card
By Memset0, published under apache-2.0, revision ced6cd88cfc0.
Preview checkpoint, not yet evaluated. Final student of a 200-step DMD run (E0029 Variant V0403). It differs from memset0/vsqa-preview-14b (V0362) in two ways: - critic (fake score) learning rate 2e-6 instead of 4e-7 (critic update ratio 1 unchanged); - the linear layers' NVFP4 QAT uses global scales fixed to 1.0 (nvfp4qattraings1) for activations and weights, instead of the dynamic 448·6/max|x| recomputed every forward. All at 77 frames, 1280×768, global batch 16, from the original Wan2.1-T2V-14B weights: 1. Dense NVFP4 per-layer reconstruction, 250 steps. 2. Plain ODE initialization (endpoint regression), 500 steps, VSA C256 cube (4,8,8), sparsity 0.9, NVFP4 attention + linear QAT. 3. DMD…
Read Memset0's full model card
VSQA preview 14B v2 — DMD 200 steps (critic LR 2e-6, linear NVFP4 global scale 1)
Preview checkpoint, not yet evaluated. Final student of a 200-step DMD run (E0029 Variant V0403). It differs from memset0/vsqa-preview-14b (V0362) in two ways:
- critic (fake score) learning rate 2e-6 instead of 4e-7 (critic update ratio 1 unchanged);
- the linear layers' NVFP4 QAT uses global scales fixed to 1.0 (
nvfp4_qat_train_gs1) for activations and weights, instead of the dynamic448·6/max|x|recomputed every forward.
Training lineage
All at 77 frames, 1280×768, global batch 16, from the original Wan2.1-T2V-14B weights:
- Dense NVFP4 per-layer reconstruction, 250 steps.
- Plain ODE initialization (endpoint regression), 500 steps, VSA C256 cube (4,8,8), sparsity 0.9, NVFP4 attention + linear QAT.
- DMD distillation, 200 steps from step 2: student LR 2e-6, critic LR 2e-6, generator updated every iteration, no EMA; teacher and critic are the original dense Wan2.1-T2V-14B.
Sampling contract
- 3 denoising steps at timesteps
[1000.0, 941.1763916015625, 800.0], no additional timestep warp. flow_shift = 8.0, guidance scale 1.0 (distilled), 77 frames at 1280×768.- Training attention backend
VSA_QAT_TRAIN_C256(sparsity 0.9); linear quantizationnvfp4_qat_train_gs1. Reproducing the trained numerics needs the FastVideo VSA/QAT kernels plus the global-scale-1 patch; plain dense BF16 inference is a different, untested configuration.
Files
Diffusers layout; transformer/ is the exported student, other components are unchanged from the base model.
metadata.json, artifact_manifest.json, resolved.yaml, source.yaml and export-verification.json record
provenance and the strict-reload export verification. License follows the base model (Apache-2.0).
Identity and Version
- Repository
- memset0/vsqa-preview-14b-v2
- Publisher
- Memset0
- Task
- Text to video
- Modality
- Video
- Library
- diffusers
- Parameters
- Not stated by the source
- Languages
- dmd
- Revision
- ced6cd88cfc047a630fca47c7cdcf63cf939676c
- First published
- 2026-09-25
- Last updated
- 2026-09-25
Files and Weights
34 files, 84.6 GB in total. The weights are 7 files totalling 84.6 GB in safetensors.
Weights7 files · 84.6 GB
Configuration12 files · 282.9 KB
Tokenizer3 files · 21.4 MB
Documentation1 file · 2.1 KB
Other10 files · 6.7 MB
Repository1 file · 2.2 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| text_encoder/model-00001-of-00005.safetensors | Weights | 5.0 GB | c0ef3a140898 |
| text_encoder/model-00002-of-00005.safetensors | Weights | 4.9 GB | 481c7b2b3977 |
| text_encoder/model-00003-of-00005.safetensors | Weights | 5.0 GB | f93148bcc040 |
| text_encoder/model-00004-of-00005.safetensors | Weights | 5.0 GB | a451792c739c |
| text_encoder/model-00005-of-00005.safetensors | Weights | 2.9 GB | 7e76e18d2245 |
| transformer/model.safetensors | Weights | 61.3 GB | ddf4c88b4a31 |
| vae/diffusion_pytorch_model.safetensors | Weights | 507.6 MB | d6e524b3fffe |
| artifact_manifest.json | Configuration | 618 B | — |
| export-verification.json | Configuration | 837 B | — |
| metadata.json | Configuration | 162.1 KB | — |
| model_index.json | Configuration | 400 B | — |
| resolved.yaml | Configuration | 83.0 KB | — |
| scheduler/scheduler_config.json | Configuration | 751 B | — |
| source.yaml | Configuration | 3.6 KB | — |
| text_encoder/config.json | Configuration | 854 B | — |
| text_encoder/model.safetensors.index.json | Configuration | 22.5 KB | — |
| tokenizer/special_tokens_map.json | Configuration | 7.1 KB | — |
| transformer/config.json | Configuration | 466 B | — |
| vae/config.json | Configuration | 724 B | — |
| README.md | Documentation | 2.1 KB | — |
| assets/comp_effic.png | Other | 1.8 MB | b0e225caffb4 |
| assets/data_for_diff_stage.jpg | Other | 528.3 KB | 59aec08409f2 |
| assets/i2v_res.png | Other | 891.7 KB | 6823b3206d8d |
| assets/logo.png | Other | 56.3 KB | — |
| assets/t2v_res.jpg | Other | 301.0 KB | 91db57909244 |
| assets/vben_1.3b_vs_sota.png | Other | 515.8 KB | b7705db79f2e |
| assets/vben_vs_sota.png | Other | 1.6 MB | 9a0e86ca8504 |
| assets/video_dit_arch.jpg | Other | 643.4 KB | 195dceec6570 |
| assets/video_vae_res.jpg | Other | 212.6 KB | d8f9e7f73538 |
| examples/i2v_input.JPG | Other | 250.6 KB | 077e3d965090 |
| .gitattributes | Repository | 2.2 KB | — |
| tokenizer/spiece.model | Tokenizer | 4.5 MB | e3909a67b780 |
| tokenizer/tokenizer.json | Tokenizer | 16.8 MB | 20a46ac25674 |
| tokenizer/tokenizer_config.json | Tokenizer | 61.8 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 84.6 GB
Download from Memset0
Released by Memset0 through its official repository on Hugging Face. Read the license.
Built From
- Derived from Wan-AI/Wan2.1-T2V-14B-Diffusers
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 84.6 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About vsqa-preview-14b-v2
Can I use vsqa-preview-14b-v2 commercially?
Yes. vsqa-preview-14b-v2 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.
Similar Models
A LoRA for MiniMax-H3 that renders joint video + synchronized stereo audio in as few as 4 sampling steps instead of the usual ~20 — a ~5× sampling speedup — and keeps getting better as you add steps. For most work, use minimaxh3turbov4step600ema.safetensors. It's the markedly better micro-detail (faces, fingers, fine texture), and the over-sharpening / plastic look of the earlier v1 (~850) line is fully resolved. v4 introduced a static-frame enhancement — a big win for static and small-motion content. The one trade-off shows up only at 4 steps with large, fast motion, where v4 can produce motion-smear / trailing ghosting (we're actively fixing this). Two things address it: - Use 6–8 steps.…
This repository contains MiniMax-H3 Turbo and few-step LoRAs converted and optimized for ComfyUI. These LoRAs accelerate MiniMax-H3 video and synchronized-audio generation by reducing the required number of sampling steps. Four HyperFlow 8-step MiniMax-H3 LoRAs are now included. They are derived from the original HyperFlow release by Video Rebirth: HyperFlow is an 8-step MiniMax-H3 LoRA designed for the MiniMax-H3 T2VA, FL2VA and Ref2VA workflows. The upstream HyperFlow implementation uses its own fixed 8-step sigma schedule. For these ComfyUI conversions, the recommended manual sigma sequence is: This is a nine-point sigma sequence defining eight denoising intervals. Use the non-pruned…
Sulphur 2 An uncensored video generation model based on LTX 2.3 supporting both t2v and i2v natively, as well as all of the other ltx 2.3 formats. Follow us on X Join our Discord Support the next version of the project, even just a few dollars would go a long way: Kofi To get started with the model, I recommend downloading either of the dev versions, (fp8mixed or bf16) and downloading the distill lora provided. By the way, I'm aware the workflows contain sulphurfinal right now, just use the lora or use the full models, don't use both at the same time. This model contains a prompt enhancer. The easiest way to get started with the prompt enhancer is by using it on lmstudio. The way to…
Wan2.1 is an open-source suite of video foundation models, compatible with consumer-grade GPUs, that excels in various video generation tasks like text-to-video, image-to-video, and video editing, even supporting visual text generation. Download models using huggingface-cli: You can also download directly from this page. This model is a derivative work of the original model licensed under the Apache 2.0 License, and is therefore distributed under the terms of the same license. Thanks to Patrick Gillespie for creating the ASCII text art tool used in this project https://patorjk.com/software/taag/ Wan-AI for the Wan model https://huggingface.co/Wan-AI/Wan2.1-VACE-1.3B…
Wan2.1 is an open-source suite of video foundation models, compatible with consumer-grade GPUs, that excels in various video generation tasks like text-to-video, image-to-video, and video editing, even supporting visual text generation. Download models using huggingface-cli: You can also download directly from this page. This model is a derivative work of the original model licensed under the Apache 2.0 License, and is therefore distributed under the terms of the same license. Thanks to Patrick Gillespie for creating the ASCII text art tool used in this project https://patorjk.com/software/taag/ Wan-AI for the Wan model https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B…