Full-text search
Search in
Scope to owner or repo
227 results
furusu / ControlNet
README.md
model
1 matches
ND911 / EclecticEuphoria_NoobIll_vpred
README.md
model
1 matches
Runware / NoobAI-XL-VPred-1.0
README.md
model
4 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-xl, art, not-for-all-audiences, text-to-image, en, base_model:Laxhar/noobai-XL-Vpred-0.75, base_model:finetune:Laxhar/noobai-XL-Vpred-0.75, license:other, autotrain_compatible, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
26
NoobAI's v_prediction model is a separate model from [Terminus XL](https://huggingface.co/collections/bghira/terminus-xl...
⋯
32
Implemented as a v-prediction model (distinct from eps-prediction), it requires specific parameter configurations - deta...
⋯
98
scheduler_args = {"prediction_type": "v_prediction", "rescale_betas_zero_snr": True}
⋯
281
- **V-Prediction**: Thanks to the following individuals for their detailed instructions and experiments.
justinchuby / onnx-genai-cogvideox-2b
README.md
model
1 matches
Old-Fisherman / SDXL_Finetune_GGUF_Files
README.md
model
4 matches
tags: gguf, base_model:stabilityai/stable-diffusion-xl-base-1.0, base_model:quantized:stabilityai/stable-diffusion-xl-base-1.0, license:other, region:us
29
## V_Prediction Custom Node
30
31
If you are using a v_prediction model, the quantized model won't be automatically detected as such in ComfyUI. I made a ...
32
V_Prediction Node Link: https://github.com/magekinnarus/ComfyUI-V-Prediction-Node
This file contains 1 more match not shown. See all 4
matches in the full file.
terminusresearch / terminus-xl-velocity-training
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, text-to-image, full, base_model:terminusresearch/terminus-xl-velocity-v2, base_model:finetune:terminusresearch/terminus-xl-velocity-v2, license:creativeml-openrail-m, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
1010
- Micro-batch size: 32
1011
- Gradient accumulation steps: 2
1012
- Number of GPUs: 8
1013
- Prediction type: v_prediction
1014
- Rescaled betas zero SNR: True
1015
- Optimizer: AdamW, stochastic bf16
1016
- Precision: Pure BF16
RareConcepts / pseudo-flex-v2
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, text-to-image, full, base_model:stabilityai/stable-diffusion-2-1, base_model:finetune:stabilityai/stable-diffusion-2-1, license:creativeml-openrail-m, region:us
49
- Effective batch size: 32
50
- Micro-batch size: 4
51
- Gradient accumulation steps: 8
52
- Prediction type: v_prediction
53
- Rescaled betas zero SNR: True
54
- Optimizer: AdamW, stochastic bf16
55
- Precision: Pure BF16
Luo-Yihong / yoso_pixart512
README.md
model
1 matches
tags: diffusers, safetensors, text-to-image, en, arxiv:2403.12931, base_model:PixArt-alpha/PixArt-XL-2-512x512, base_model:finetune:PixArt-alpha/PixArt-XL-2-512x512, region:us
30
pipe = pipe.to('cuda')
31
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
32
pipe.scheduler.config.prediction_type = "v_prediction"
33
generator = torch.manual_seed(318)
34
imgs = pipe(prompt="Pirate ship trapped in a cosmic maelstrom nebula, rendered in cosmic beach whirlpool engine, volumet...
35
num_inference_steps=1,
bghira / terminus-xl-refiner
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, text-to-image, full, base_model:segmind/SSD-1B, base_model:finetune:segmind/SSD-1B, license:creativeml-openrail-m, autotrain_compatible, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
50
- Micro-batch size: 4
51
- Gradient accumulation steps: 4
52
- Number of GPUs: 1
53
- Prediction type: v_prediction
54
- Rescaled betas zero SNR: True
55
- Optimizer: AdamW, stochastic bf16
56
- Precision: Pure BF16
bghira / terminus-velocity-v2-reddit
README.md
model
1 matches
tags: diffusers, stable-diffusion, stable-diffusion-diffusers, text-to-image, lora, template:sd-lora, base_model:stabilityai/stable-diffusion-3-medium-diffusers, base_model:adapter:stabilityai/stable-diffusion-3-medium-diffusers, license:creativeml-openrail-m, region:us
50
- Micro-batch size: 1
51
- Gradient accumulation steps: 1
52
- Number of GPUs: 1
53
- Prediction type: v_prediction
54
- Rescaled betas zero SNR: True
55
- Optimizer: AdamW, stochastic bf16
56
- Precision: Pure BF16
RedRayz / illumina-noob-vpd
README.md
model
4 matches
tags: stable-diffusion, sdxl, en, base_model:Laxhar/noobai-xl-EarlyAccess, base_model:finetune:Laxhar/noobai-xl-EarlyAccess, license:other, region:us
20
Fine-tuned NoobAI-XL(v-prediction) and merged SPO
21
22
NoobAI-XL(v-prediction)をファインチューンし、SPOをマージしました。
⋯
42
5. Select `v_prediction` in Discrete Sampling Type
⋯
116
5. Discrete Sampling Typeを`v_prediction`にする
RedRayz / abydos-noobvpd-0.5
README.md
model
3 matches
tags: stable-diffusion, sdxl, base_model:RedRayz/illumina-noob-vpd, base_model:finetune:RedRayz/illumina-noob-vpd, license:other, region:us
18
Finetuned NoobAI-XL(v-prediction) + SPO with Blue Archive style
⋯
38
4. Enable “Enable Advanced Model Sampling”
39
5. Select `v_prediction` in Discrete Sampling Type
⋯
91
4. “Enable Advanced Model Sampling”を有効にする
92
5. Discrete Sampling Typeを`v_prediction`にする
hanzogak / comradeshipXL-v14VT
README.md
model
1 matches
tags: diffusers, safetensors, license:other, autotrain_compatible, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
17
use_safetensors=True,
18
add_watermarker=False
19
)
20
scheduler_args = {"prediction_type": "v_prediction", "rescale_betas_zero_snr": True}
21
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config, timestep_spacing="trailing", **sched...
22
23
pipe.to('cuda')
yurman / uncond-sd2-base-complex
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, base_model:yurman/uncond_sd2-base, base_model:finetune:yurman/uncond_sd2-base, license:creativeml-openrail-m, diffusers:StableDiffusionUnconditionalPipeline, region:us
46
* Image resolution: 256
47
* Mixed-precision: no
48
* Max rotation degree: 10
49
* Prediction Type: v_prediction
50
* SNR Gamma: 5.0
yurman / uncond-sd2-base-complex-4
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, base_model:yurman/uncond_sd2-base, base_model:finetune:yurman/uncond_sd2-base, license:creativeml-openrail-m, diffusers:StableDiffusionUnconditionalPipeline, region:us
46
* Image resolution: 256
47
* Mixed-precision: no
48
* Max rotation degree: 10
49
* Prediction Type: v_prediction
50
* SNR Gamma: 5.0
yurman / uncond-sd2-base-complex-8
README.md
model
1 matches
tags: diffusers, safetensors, stable-diffusion, stable-diffusion-diffusers, base_model:yurman/uncond_sd2-base, base_model:finetune:yurman/uncond_sd2-base, license:creativeml-openrail-m, diffusers:StableDiffusionUnconditionalPipeline, region:us
46
* Image resolution: 256
47
* Mixed-precision: no
48
* Max rotation degree: 10
49
* Prediction Type: v_prediction
50
* SNR Gamma: 5.0
Koolchh / AnimeBoysZeroXL
README.md
model
3 matches
tags: diffusers, text-to-image, stable-diffusion, license:openrail++, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
43
...️ Important**: This model uses Zero Terminal SNR with V-prediction. Please ensure you are using the correct settings dur...
44
...Discrete` node into your workflow. Set `sampling` to `v_prediction`, `zsnr` to `true`.
⋯
130
- 5% caption dropout for unconditional guidance.
131
- Trained with Zero Terminal SNR and V-prediction.
132
- Implemented adaptive loss weighting.
Koolchh / AnimeBoysNabla
README.md
model
2 matches
tags: diffusers, text-to-image, stable-diffusion, license:openrail++, endpoints_compatible, diffusers:StableDiffusionXLPipeline, region:us
45
...️ Important**: This model uses Zero Terminal SNR with V-prediction. Please ensure you are using the correct settings dur...
46
...Discrete` node into your workflow. Set `sampling` to `v_prediction`, `zsnr` to `true`.
wy891 / vis2tac-latent-sparsh-r38
README.md
model
1 matches
tags: region:us
23
- Frozen vision encoder: facebook/dinov2-small
24
- Frozen tactile target encoder: facebook/dinov2-base (Sparsh-style)
25
- Denoiser: ResidualMLPDenoiser (8 res blocks, hidden=2048, latent=1536)
26
- DDPM scheduler with v_prediction
27
- Loss: MSE(v_pred, v_target) + 0.1 x cosine(x0_pred, target)
28
- EMA 0.999, validation by probe Delta (NOT MSE)