<h2 id="generates-wan-2.1-videos-in-a-fraction-of-time.">Generates WAN 2.1 videos in a fraction of time.</h2>720p and 480p Version<h3 id="recommended-settings:">Recommended settings:</h3><ul><li>Sampler/Scheduler: Euler/Simple</li><li>Steps: 4</li><li>CFG: 1</li><li>Sigma-Shift 5</li></ul>Original Model from <a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Wan2.1-I2V-14B-480P-StepDistill-CfgDistill-Lightx2v">Lightx2v</a> converted to FP8 quantisation.☠️ Do not use any extra speed-up tricks or LoRAs or it may mess up your generations ... 🤬⚠️ Hint: Most of the time the model is taking you by word. If you write "white" it is white. "Translucent" is translucent... like for the fluids. 💦 Now you know! 🫵 translucent whitish 🤫⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️<h3 id="recommended-specs:">Recommended specs:</h3>8 GB VRAM, 32 GB RAMSample times: &lt;2 minutes for 81 frames, 4 steps on RTX 4070 Ti Super.<h3 id="compatible-with-14b-loras.">Compatible with 14B LoRAs.</h3>I normally use 0-2 LoRAs, strength at 0.4-1 depending on how much the effect should be. 0.7-0.9 works best most of the time, not overwriting the style of an image. At multiple LoRAs is seems best to tune the strength a bit down to 0.3-0.6.<h3 id="basic-workflow-example:">Basic workflow example:</h3>Here: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1811161?modelVersionId=2049602">https://civarchive.com/models/1811161?modelVersionId=2049602</a><h3 id="my-favourite-ui:">My favourite UI:</h3>SwarmUI <a target="_blank" rel="ugc" href="https://github.com/mcmonkeyprojects/SwarmUI">https://github.com/mcmonkeyprojects/SwarmUI</a><h3 id="testing-(my-specs):">Testing (my specs):</h3>I can go wild on setting with this full checkpoint, even with added LoRAs:<ul><li>121 frames possible: ~ 3 minutes</li><li>121 frames on 24 fps possible (more motion): ~ 3 minutes</li><li>128 frames on 24 fps possible (more motion and extended): ~ 3.5 minutes</li></ul><h3 id="dependencies:">Dependencies:</h3><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/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_2.1_ComfyUI_repackaged/resolve/main/split_files/clip_vision/clip_vision_h.safetensors">Clip-Vision</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors">Clip-T5-Encoder</a></li></ul><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.

DaSiWa WAN 2.1 Lightspeed

dasiwaWAN21_lightspeedI2v14B720p.safetensors

Lightspeed-i2v-14B-720p

Lightspeed-t2v-14B

Clip - Vision H

Clip - T5 Encoder FP8

Lightspeed-i2v-14B-480p

Generates Wan 2.1 videos in a fraction of time.Steps: 4CFG: 1Sigma-Shift 5Original Model from Lightx2v converted to FP8 quantisation.☠️ Do not use any extra speed-up tricks or LoRAs or it may mess up your generations ... 🤬⚠️ Hint: Most of the time the model is taking you by word. If you write "white" it is white. "Translucent" is translucent... like for the fluids. 💦 Now you know! 🫵 translucent whitish 🤫⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️Sage Attention, Flash Attention, Radial Attention, q8-kernel, sgl-kernel, vllm are built in, no need to enable it! Compatible with 14B LoRAs.I normally use 0-2 LoRAs, strength at 0.4-1 depending on how much the effect should be. 0.7-0.9 works best most of the time, not overwriting the style of an image.At multiple LoRAs is seems best to tune the strength a bit down to 0.3-0.6.Basic workflow example:Here: <a target="_blank" rel="noopener noreferrer nofollow" href="https://civitai.com/models/1811161?modelVersionId=2049602">https://civitai.com/models/1811161?modelVersionId=2049602</a>My favourite UI:SwarmUI <a target="_blank" rel="noopener noreferrer nofollow" href="https://github.com/mcmonkeyprojects/SwarmUI">https://github.com/mcmonkeyprojects/SwarmUI</a>Testing (my specs):I can go wild on setting with this full checkpoint, even with added LoRAs:121 frames possible: ~ 3 minutes121 frames on 24 fps possible (more motion): ~ 3 minutes128 frames on 24 fps possible (more motion and extended): ~ 3.5 minutesIf needed:Clip/T5-Encoder: umt5_xxl_fp8_e4m3fn_scaled.safetensors

Files with SHA256: 903e61344665a197ff2cf2300e4f4593c88535fc1961dc6b73ad51fdfc7b0952