<h1 id="dasiwa-wan-2.2-i2v-14b-tastysin-or-lightspeed-or-gguf">💀💋 DaSiWa-WAN 2.2 I2V 14B TastySin | Lightspeed | GGUF 💋💀</h1>This is a WAN 2.2 Model: You will need one pair of High + Low.Version overview: <a target="_blank" rel="ugc" href="https://civarchive.com/articles/23495/dasiwa-model-versions-and-timeline">https://civarchive.com/articles/23495/dasiwa-model-versions-and-timeline</a><hr /><h3 id="key-features:">🔮 Key Features:</h3><ul><li>🔥 LoRA-Free Generations Generate high-quality videos without stacking Wan 2.2 LoRAs (unless you want adding spacial styles/concepts).</li><li>☄️Fast: 4 step generation</li><li>💫Quality motions (less slowdowns, no pixelated hyper-motion)</li><li>🔞 NSFW and SFW + Extreme versatile (more build in concepts):<ul><li>Enhanced anatomy + poses + framing</li><li>Better understanding of sexual concepts</li></ul></li></ul><ul><li>🪄 Better prompt responsiveness</li><li>👘 Better understanding of anime/manga style composition</li><li>🪡 Q8 (FP16 base) precision</li><li>🚫 Do not use any extra speed-up (low step) LoRAs, this is baked in already</li></ul><h3 id="optimisations">✅ Optimisations</h3><ul><li>🌟 CFGZeroStar patch (better results and prompt adherence)</li><li>🍰 Baked Latest Distillation (r64-1022)</li><li>🍀Additional concept optimisation (compared to MidnightFlirt)</li><li>🎬 Reward Attention (more realistic movements)</li><li>✨ No or extreme low transformation of details with anime/realistic images (lips, eyes, ears, breasts, genitals, ...)</li><li>🌠 Refined motions (compared to MidnightFlirt)</li><li>💌 It can guesstimate better details out of frame (compared to MidnightFlirt)</li><li>💫 As close to RL motion-speeds as I could get with speed-up tech</li><li>🍌Even less tries for good results (compared to MidnightFlirt)</li><li>🧩Even better compatibility with LoRAs (compared to MidnightFlirt)</li><li>🖼️ Usable with your preferred/custom CLIP (if compatible)</li><li>😵‍💫Additional reduced hallucinations (compared to MidnightFlirt)</li><li>💠 Zero prompt results capable</li><li>🚫Excluded CLIP</li></ul><hr /><h3 id="workflow">🍒Workflow</h3>Make sure to checkout my easy to use <a target="_blank" rel="ugc" href="https://civarchive.com/models/1823089">Workflows</a>!<hr /><h3 id="lora's">🍄LoRA's</h3>Try first without additional LoRAs!But: This checkpoint is not meant to replace all LoRAs, it is meant to:<ul><li>Perform better overall at his own</li><li>As easy as possible to use</li><li>With LoRAs to be absolutely awesome</li></ul><hr /><h3 id="announcement">🪧Announcement</h3>⚠️ Read the corresponding <a target="_blank" rel="ugc" href="https://civarchive.com/articles/23271/">announcement</a>.📢 Make sure to check it out for in-depth information and a complex comparison!<hr />🆕 New to WAN 2.2 I2V? - Check out my <a target="_blank" rel="ugc" href="https://civarchive.com/articles/20293/darksidewalkers-wan-22-14b-i2v-usage-guide-definitive-edition">guide</a>.<hr /><h3 id="recommended-settings">🛠️ Recommended Settings</h3><ul><li>Steps: 4</li><li>CFG: 1</li><li>Sampler/Scheduler: Euler/Simple, UniPC_BH2/Simple</li><li>Resolution up to 720p (native quality).</li><li>Add other LoRAs with 0.3 - 0.6 at first</li><li>16 or 24 fps, 81 or 97 frames ~ 5s</li></ul><hr /><h3 id="dependencies">Dependencies</h3><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/main/split_files/vae/wan_2.1_vae.safetensors">Wan 2.1 VAE</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors">CLIP-Text-Encoder</a></li></ul><hr /><h3 id="speed-+-examples">💫 Speed + Examples</h3>Q8 checkpoint - On 16GB VRAM, 64GB RAM, 4 steps, cfg 1, 81 frames<ul><li>368p: 120 sec</li><li>480p: 160 sec</li><li>576p: 220 sec</li><li>608p: 340 sec</li><li>672p: 680 sec</li><li>720p: 730 sec</li><li>Most examples are without any additional LoRAs</li><li>With LoRAs are for testing the compatibility</li><li>Initial anime-like example images made by me are also made with my model 🗡️💀 <a target="_blank" rel="ugc" href="https://civarchive.com/models/1034922">DaSiWa-Illustrious-XL</a> 💀🗡️</li><li>Other models for realistic reasoning</li></ul><hr /><h3 id="known-issues">🩻 Known issues</h3><ul><li>🫦 The most delicious sin!</li><li>Tell me 🫵🫢</li></ul><hr /><h3 id="approximate-expected-quality-from-quantization">Approximate expected quality from quantization</h3>This are my tests compared to a full fp16 safetensor checkpoint taking prompt- and visual satisfaction into account on my DaSiWa checkpoints.Quantization - Quality estimation<ul><li>Q8 ⭐⭐⭐⭐⭐ ~ like FP16/FP8+, excellent results</li><li>Q6 ⭐⭐⭐⭐ ~ like FP8, very good results</li><li>Q5 ⭐⭐⭐ ~ good to very good results</li><li>Q4 ⭐⭐ ~ medium-good results</li><li>Q2 ⭐ ~ poor results, only use if you have to</li></ul>⚠️ Do not compare this with the unofficial quants of my checkpoint made by others, they are based on FP8 and not FP16 like my quants.<hr /><h3 id="fixes-and-feedback">🩺 Fixes &amp; Feedback</h3><ul><li>If you use LoRAs, try to respect the LoRA training triggers and try some versatile descriptions, most LoRAs will work with 0.3-0.6 (start with 0.3)</li><li>Raise LoRAs in little steps +0.1</li><li>Do not mass add LoRAs, just add 1 or 2 (x2 High+Low)</li><li>Negative prompting do not work with cfg 1, that's a limitation of speed-ups with cfg 1 (except you use NAG)</li><li>Low resolution (e.g.384x576) are for fast samples and will blur fine details, do a higher resolution if you want clear details</li></ul>🪧❗ Test your comfyui-backend with this absolute basic <a target="_blank" rel="ugc" href="https://civarchive.com/models/1823089?modelVersionId=2405252">test-workflow</a> before asking about errors.<hr /><h3 id="why-i-made-this">🖤 Why I Made This</h3>I was tired of using all these massive list of LoRAs, just to get a remotely good result after 10 generations, consuming hours of time. So I can just make my videos with 1 or 2 concept LoRAs without pushing 6 till 10 LoRAs (Low/High) into a generation. This checkpoint is also my personal playground.<hr /><h3 id="closing-words">Closing words</h3>🤩 I want to thank all the fantastic other creators who made super nice LoRAs and concepts to play with! Support that awesome creators by using their LoRAs and post to their gallery and share the meta-data! ⚠️ I made all this with permissions or open-source resources (the time it is incorporated).I share as much insights as I can without compromising my work. I'm doing this for fun as my hobby and just do not want my hobby to be destroyed.More details can be obtained in the corresponding announcements!<hr />If you would like to contribute in my awesome (😉) checkpoint or willing to share resources I'll gladly give credit! Just contact me!<hr /><span style="background-color:rgb(26, 27, 30);color:rgb(141, 141, 143);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">YOU are responsible for outputs as always! If you make <a target="_blank" rel="ugc" href="https://civarchive.com/content/tos"><span style="background-color:rgb(26, 27, 30);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">ToS</a><span style="background-color:rgb(26, 27, 30);color:rgb(141, 141, 143);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px"> violating content and I get aware I WILL report this.<hr /><h3 id="disclaimer">Disclaimer</h3>This models are shared without warranties and with the condition that it is used in a lawful and responsible way. I do not support or take responsibility for illegal, harmful, or harassing uses. By downloading or using it, you accept that you are solely responsible for how it is used.<hr /><h3 id="custom-license-addendum:-distribution-restriction">Custom License Addendum: Distribution Restriction</h3>Notice: Notwithstanding the base license selected for this model, the following restrictive terms apply:<ol><li>No Redistribution: You are not permitted to host, mirror, or redistribute this model (checkpoint, LoRA, or Safetensors files) on any other platform, website, or service (including but not limited to Hugging Face, Tensor.art, or SeaArt) without explicit written permission from the creator.</li><li>Attribution &amp; Source: This model is officially maintained only on Civitai or other platforms where I explicitly own the repository. To ensure users receive the correct version, updates, and safety metadata, please point users to the original URL.</li><li>Usage: All other rights regarding the use of the model for image generation remain as per the terms and the restrictions provided per model.</li></ol>

💀💋 DaSiWa-WAN 2.2 I2V 14B TastySin v8 | Lightspeed | GGUF 💋💀

DasiwaWAN22I2V14BTastysinV8_q2Low.gguf

Q2 Low

Q2 High

Q4 High

Q4 Low

Q5 Low

Q5 High

Q6 Low

Q6 High

Q8 Low

Q8 High

Files with SHA256: f196e9c244b6467e4a9bb0433b8e2f1e0c67e3899f75caba4c52f0fba0719cc3