一、引言
在上一篇文章从144G到64G的极致压缩：K100_AI单卡部署MiniMax-H3全模态模型的实战（含详细部署步骤）_minimax h3 服务器部署-CSDN博客，我们成功在64GB显存的海光K100_AI单卡上跑通了MiniMax-H3的全流程。然而，受限于开源版本，H3的原生输出分辨率被限制在短边768px，距离其宣称的2K还有差距。本文将作为续篇，介绍如何通过"LoRA加速生成 + FlashVSR视频放大 "的组合方案，将768P视频高效升级至2K高清。核心思路是：用480p/768p进行快速创意验证，再对选中的镜头进行2K放大。
二、工作流方案设计
整体流程 ：文本提示词 → LoRA加速的MiniMax-H3生成 (输出768P视频) → FlashVSR视频放大 (输出2K视频)
方案一：LoRA加速生成工作流（源头提速）
-
核心目标 ：在保证质量的前提下，大幅缩短768P视频的生成时间，解决开源H3生成速度慢的痛点。
-
关键技术 ：引入 MiniMax-H3 Turbo LoRA ，将采样步数从约20步减少到4-8步 。配合双时钟Euler采样器 解决低步数下的音频问题。同时，可叠加使用SageAttention等加速节点。
-
工作流节点 ：
UNETLoader→LoraLoaderModelOnly(加载Turbo LoRA) →BasicGuider&BasicScheduler→SamplerCustomAdvanced(连接Turbo Sampler)。
方案二：FlashVSR视频放大工作流（画质升级）
-
核心目标 ：将已生成的768P视频高效、保真地放大至2K。
-
技术选型 ：FlashVSR是一个基于扩散模型的视频超分辨率 方案，能实现2倍或4倍放大。实测在类似配置下，处理124帧（约5秒）视频仅需56秒。
-
工作流节点 ：
VHS_LoadVideoFFmpeg(加载视频) →AILab_FlashVSR_Advanced(放大) →VHS_VideoCombine(合成输出)。
Lora加速生成流程图如下：
FlashVSR视频放大流程图如下：
三、具体部署实施过程
3.1 基础环境
-
硬件：海光K100_AI DCU显卡（64GB显存）
-
软件：ComfyUI（建议v0.30.0+）
3.2 模型与插件下载
模型下载（建议使用镜像站加速）：
- MiniMax-H3基础模型（来自
Comfy-Org/MiniMax-H3）:
bash
diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors (约19.5GB)
text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors (约14.6GB)
vae/minimax_h3_video_vae_fp16.safetensors
vae/minimax_h3_audio_vae_fp32.safetensorsMiniMax-H3 Turbo LoRA（来自larryvrh/MiniMax-H3-Turbo-Lora）:
bash
minimax_h3_turbo_v4_step600_ema.safetensors (推荐v4版本)FlashVSR模型（来自1038lab/FlashVSR）:
bash
首次运行AILab_FlashVSR_Advanced节点时会自动下载到models/FlashVSR目录。
也可以手工下载到ComfyUI/models/FlashVSR/目录里，如下：
modelscope download --model 1038lab/FlashVSR --local_dir ./models/FlashVSRCustom Nodes安装（在ComfyUI/custom_nodes/目录下操作）：
bash
# LoRA加速相关
git clone https://github.com/larryvrh/ComfyUI-MiniMax-H3-Turbo.git
git clone https://github.com/shuaixn/ComfyUI-MiniMaxH3DualClockSampler.git[reference:31]
git clone https://github.com/linjian-ufo/comfyui-speed-minimaxH3.git[reference:32]
# FlashVSR放大相关
git clone https://github.com/1038lab/ComfyUI-FlashVSR.git3.3 工作流导入与配置
1、导入LoRA加速工作流 ：将你提供的JSON文件拖入ComfyUI界面，并且通过Manager → Install Missing Custom Nodes安装缺失节点。
2、配置LoRA加速节点 ：在LoraLoaderModelOnly节点中，选择minimax_h3_turbo_v4_step600_ema.safetensors，强度设为1.0。确保SamplerCustomAdvanced节点连接了MiniMaxH3DualClockEulerSampler节点。将BasicScheduler的步数设为4-8步。
3、导入FlashVSR放大工作流：拖入对应的JSON文件。
4、配置FlashVSR节点：关键参数参考和我实践的经验：
bash
model_version: Full (Best Quality)
scale: 2
enable_tiling: false (64GB显存足够)
sageattention: enable
stability_level: 115、Lora加速工作流如下：
bash
{
"id": "e3f2b845-8f2c-4b5a-9caf-eac1029d3e7e",
"revision": 0,
"last_node_id": 122,
"last_link_id": 231,
"nodes": [
{
"id": 92,
"type": "SaveVideo",
"pos": [
-532.2764838467319,
4771.015641008128
],
"size": [
1070,
125.8125
],
"flags": {},
"order": 5,
"mode": 0,
"inputs": [
{
"name": "video",
"type": "VIDEO",
"link": 194
}
],
"outputs": [
{
"name": "video",
"type": "VIDEO",
"links": null
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "SaveVideo"
},
"widgets_values": [
"video/MiniMax_H3",
"auto",
"auto"
]
},
{
"id": 115,
"type": "ResolutionSelector",
"pos": [
-1550,
5220
],
"size": [
270,
170
],
"flags": {},
"order": 0,
"mode": 0,
"showAdvanced": true,
"inputs": [],
"outputs": [
{
"name": "width",
"type": "INT",
"links": [
219
]
},
{
"name": "height",
"type": "INT",
"links": [
220
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "ResolutionSelector"
},
"widgets_values": [
"16:9 (Widescreen)",
0.4,
32
]
},
{
"id": 105,
"type": "4c314f31-ecda-4b08-ae98-faaba1bf613f",
"pos": [
-1287.4035853353148,
4360.494136491905
],
"size": [
480,
630
],
"flags": {},
"order": 4,
"mode": 0,
"inputs": [
{
"name": "first_frame",
"shape": 7,
"type": "IMAGE",
"link": null
},
{
"name": "last_frame",
"shape": 7,
"type": "IMAGE",
"link": null
},
{
"name": "width",
"type": "INT",
"widget": {
"name": "width"
},
"link": 219
},
{
"name": "height",
"type": "INT",
"widget": {
"name": "height"
},
"link": 220
},
{
"label": "duration",
"name": "value_1",
"type": "FLOAT",
"widget": {
"name": "value_1"
},
"link": null
},
{
"label": "audio_vae",
"name": "vae_name_1",
"type": "COMBO",
"widget": {
"name": "vae_name_1"
},
"link": null
}
],
"outputs": [
{
"name": "VIDEO",
"type": "VIDEO",
"links": [
194
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"previewExposures": []
},
"widgets_values": [
"PROMPT FOR MINIMAX-H3 (Text-to-Video + Audio)\n[VISUAL DESCRIPTION -- Cinematic & Scene Setting]\nGenre: High-fantasy Xianxia (Immortal Cultivation), epic cinematic style.\nDuration: 15 seconds total.\nCamera Movement: Slow, sweeping drone-style tracking shot starting from a low-angle wide establishing view, gradually pushing forward into a medium two-shot, then settling into a gentle orbiting close-up.\nScene Environment (0s -- 15s):\nAn ethereal, celestial immortal realm at twilight. Massive jade-green floating mountains hover amid a sea of swirling, iridescent golden clouds. A narrow, ancient bluestone bridge, covered in glowing moss, connects two colossal peaked pagodas. To the left, a luminous waterfall of liquid starlight cascades silently into an abyss of mist. Thousands of firefly-like spirit particles (glowing cyan and magenta) drift lazily through the air, casting dynamic volumetric god-rays through the humid atmosphere.\nCharacters & Appearance:\nCharacter A (Han Li - 韩立): A stoic male cultivator in his 30s. Wears a tattered but elegant dark-green robe with silver runic embroidery. Has a calm, weathered face with sharp eyes. Carries a small, unassuming bamboo tube on his back.\nCharacter B (Nangong Wan - 南宫婉): A graceful female cultivator in her late 20s. Wears a pristine, flowing white silk dress with icy-blue phoenix patterns. Her long black hair is pinned up with a jade hairpin, and she exudes an aloof, ethereal aura.\nAction & Movement (Timeline):\n0s -- 5s: Both characters walk slowly side-by-side across the stone bridge. Their immortal robes and sleeves flutter gently backward, defying gravity in the slow breeze. They move in perfect synchronicity, stepping lightly as if floating.\n6s -- 10s: They suddenly pause and turn their heads slightly to face each other. The camera orbits 45 degrees around them. Their lips move naturally to deliver the dialogue. The surrounding spirit particles swirl faster around them as they speak.\n11s -- 15s: After speaking, they resume walking forward, and the camera slowly pulls back to a wide shot, revealing the vast, boundless scale of the divine mountains, ending with a soft cinematic fade into the glowing horizon.\nVisual Style: 8K resolution, hyper-detailed textures, shallow depth of field (background softly blurred), lens flare from the magical light, high dynamic range (HDR), slow-motion effect applied to the fabric movement, award-winning fantasy cinematography.\n[AUDIO & DIALOGUE INSTRUCTIONS -- Mandarin Chinese]\nAudio Type: Synchronized character voice-overs with ambient environmental audio.\nBackground Ambience (0s -- 15s):\nSoft, rumbling bass of distant waterfalls, mixed with high-frequency crystalline chimes and the gentle whoosh of wind passing through ancient pines. Maintain a mystic, serene soundscape underneath the dialogue.\nDialogue Sync (Exact timing at 7s to 10s):\nSpeaker A (Han Li -- Deep, calm, resonant male voice):\n\"此处的灵气，竟比外界浓郁数倍。\"\n(Translation: The spiritual energy here is several times denser than the outside world.)\nSpeaker B (Nangong Wan -- Clear, cool, ethereal female voice, slightly softer tone):\n\"看来你我的机缘，便在此处了。\"\n(Translation: It seems our destined opportunity lies right here.)\nVoice Directing:\nEnsure the emotional delivery matches the solemnity of the scene. Han Li's line should sound contemplative and slightly awed. Nangong Wan's line should sound confident and decisive, yet gentle. The voices must naturally match the lip-sync timing of the 2-second pause (7s-9s for Han Li, 9s-10s for Nangong Wan) and reverberate slightly to match the vast, open mountain environment.\nNegative Prompts (Avoid):\nNo distorted faces, no extra limbs, no modern objects, no cyberpunk elements, no cartoonish animation style, no rapid camera shaking, no out-of-sync dialogue, no muffled audio. Avoid low-light darkness; keep the scene brightly luminous.\nFinal output required: A seamless 15-second MP4 file containing the described cinematic walking sequence combined with the embedded Mandarin dialogue audio track perfectly synced to the characters' lip movements.",
1344,
768,
15,
593205691719292,
"minimax_h3_fl2va_int8_convrot.safetensors",
"qwen3vl_32b_minimax_h3_int8_convrot.safetensors",
"minimax_h3_video_vae_fp16.safetensors",
"minimax_h3_audio_vae_fp32.safetensors"
]
},
{
"id": 116,
"type": "MarkdownNote",
"pos": [
-2080,
4840
],
"size": [
450,
740
],
"flags": {},
"order": 1,
"mode": 0,
"inputs": [],
"outputs": [],
"title": "Note: MiniMax H3",
"properties": {},
"widgets_values": [
"## MiniMax H3\n\n[MiniMax H3](https://www.minimax.io/blog/minimax-h3) is MiniMax's general-purpose, omni-modal generation model. It jointly understands text, image, video, and audio, and generates video with **native stereo audio**: voice, sound effects, and music are modeled jointly in a single forward pass, not layered on afterward. Output is up to 2K resolution, 24fps, and up to about 15 seconds.\n\n## ComfyUI links\n- [ComfyUI#15224](https://github.com/Comfy-Org/ComfyUI/pull/15224)\n- [🤗 Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n\n## About this workflow\n\n**Key inputs**\n\n- **prompt**: describe the shots, camera moves, and the accompanying audio (dialogue, SFX, music) in one block\n- **width / height**: set via Resolution Selector. H3's native canvas is a 768px short edge, capped at 768x1344 pixels, rounded to a multiple of 32\n- **duration (seconds)**: converted to a valid frame `length` by the Math Expression node, snapping up to the model's 17-frame-per-block (17k+5) grid at 24fps\n"
],
"color": "#222",
"bgcolor": "#000"
},
{
"id": 117,
"type": "MarkdownNote",
"pos": [
-2550,
4840
],
"size": [
440,
740
],
"flags": {},
"order": 2,
"mode": 0,
"inputs": [],
"outputs": [],
"title": "Note: Model Links",
"properties": {},
"widgets_values": [
"## Model Links\n\n**vae**\n\n- [minimax_h3_video_vae_fp16.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors)\n- [minimax_h3_audio_vae_fp32.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors)\n\n**diffusion_models**\n\n- [minimax_h3_fl2va_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors)\n\n**text_encoders**\n\n- [qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 vae/\n│ │ ├── minimax_h3_video_vae_fp16.safetensors\n│ │ └── minimax_h3_audio_vae_fp32.safetensors\n│ ├── 📂 diffusion_models/\n│ │ └── minimax_h3_fl2va_pruned_int8_convrot.safetensors\n│ └── 📂 text_encoders/\n│ └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n"
],
"color": "#222",
"bgcolor": "#000"
},
{
"id": 118,
"type": "MarkdownNote",
"pos": [
-1570,
5460
],
"size": [
300,
520
],
"flags": {},
"order": 3,
"mode": 0,
"inputs": [],
"outputs": [],
"title": "Note: Size Settings Reference",
"properties": {},
"widgets_values": [
"| megapixels | Aspect | Output (multiple=32) |\n|---|---|---|\n| 0.2 | 16:9 | 608 x 352 |\n| 0.3 | 16:9 | 736 x 416 |\n| 0.4 | 16:9 | 864 x 480 |\n| 0.5 | 16:9 | 960 x 544 |\n| 0.6 | 16:9 | 1056 x 608 |\n| 0.7 | 16:9 | 1152 x 640 |\n| 0.8 | 16:9 | 1216 x 672 |\n| 0.9 | 16:9 | 1280 x 736 |\n| 0.98 | 16:9 | 1344 x 768 |\n| 1.0 | 16:9 | 1376 x 768 |\n| 1.2 | 16:9 | 1504 x 832 |\n| 1.5 | 16:9 | 1664 x 928 |\n| 1.8 | 16:9 | 1824 x 1024 |\n| 2.0 | 16:9 | 1920 x 1088 |\n"
],
"color": "#222",
"bgcolor": "#000"
}
],
"links": [
[
194,
105,
0,
92,
0,
"VIDEO"
],
[
219,
115,
0,
105,
2,
"INT"
],
[
220,
115,
1,
105,
3,
"INT"
]
],
"groups": [],
"definitions": {
"subgraphs": [
{
"id": "4c314f31-ecda-4b08-ae98-faaba1bf613f",
"version": 1,
"state": {
"lastGroupId": 4,
"lastNodeId": 122,
"lastLinkId": 231,
"lastRerouteId": 0
},
"revision": 0,
"config": {},
"name": "Image to Video (MiniMax H3)",
"inputNode": {
"id": -10,
"bounding": [
-2560,
4720,
128,
268
]
},
"outputNode": {
"id": -20,
"bounding": [
670,
4780,
128,
68
]
},
"inputs": [
{
"id": "d6eaa195-2266-4016-b7ed-b2d17a3e53c2",
"name": "first_frame",
"type": "IMAGE",
"linkIds": [
195
],
"pos": [
-2456,
4744
]
},
{
"id": "03b95567-f496-4279-9d38-989dd34fa882",
"name": "last_frame",
"type": "IMAGE",
"linkIds": [
196
],
"pos": [
-2456,
4764
]
},
{
"id": "d7302ca7-24ed-44c4-8ac9-736540dab7fb",
"name": "prompt",
"type": "STRING",
"linkIds": [
197
],
"pos": [
-2456,
4784
]
},
{
"id": "709e2d94-3172-496f-ad86-7d672f3df568",
"name": "width",
"type": "INT",
"linkIds": [
200
],
"pos": [
-2456,
4804
]
},
{
"id": "8205b85b-19bb-47f6-8418-89cd42940c1d",
"name": "height",
"type": "INT",
"linkIds": [
201
],
"pos": [
-2456,
4824
]
},
{
"id": "a40e5e96-4307-4a8e-a4f8-13a27d4bc9d3",
"name": "value_1",
"type": "FLOAT",
"linkIds": [
206
],
"label": "duration",
"pos": [
-2456,
4844
]
},
{
"id": "9f8734bc-c2b3-43b0-911e-e89560e8b777",
"name": "noise_seed",
"type": "INT",
"linkIds": [
207
],
"pos": [
-2456,
4864
]
},
{
"id": "9a6d2811-673c-47e0-aa94-dda7c83621e4",
"name": "unet_name",
"type": "COMBO",
"linkIds": [
221
],
"pos": [
-2456,
4884
]
},
{
"id": "7100a787-5536-4002-840b-1d2fd071010a",
"name": "clip_name",
"type": "COMBO",
"linkIds": [
227
],
"pos": [
-2456,
4904
]
},
{
"id": "d40632fd-e0ae-4319-b85d-5479db4d6e11",
"name": "vae_name",
"type": "COMBO",
"linkIds": [
223
],
"pos": [
-2456,
4924
]
},
{
"id": "2992852c-4b20-439a-8771-c866ec1996e6",
"name": "vae_name_1",
"type": "COMBO",
"linkIds": [
224
],
"label": "audio_vae",
"pos": [
-2456,
4944
]
}
],
"outputs": [
{
"id": "adb2611c-e490-4fca-8067-b718c992f8cc",
"name": "VIDEO",
"type": "VIDEO",
"linkIds": [
168
],
"localized_name": "VIDEO",
"pos": [
694,
4804
]
}
],
"widgets": [],
"nodes": [
{
"id": 11,
"type": "VAELoader",
"pos": [
-2020,
4970
],
"size": [
640,
70
],
"flags": {},
"order": 4,
"mode": 0,
"inputs": [
{
"localized_name": "vae名称",
"name": "vae_name",
"type": "COMBO",
"widget": {
"name": "vae_name"
},
"link": 223
}
],
"outputs": [
{
"localized_name": "VAE",
"name": "VAE",
"type": "VAE",
"links": [
8,
190
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "VAELoader",
"models": [
{
"name": "minimax_h3_video_vae_fp16.safetensors",
"url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors",
"directory": "vae"
}
]
},
"widgets_values": [
"minimax_h3_video_vae_fp16.safetensors"
]
},
{
"id": 24,
"type": "VAELoader",
"pos": [
-2020,
5100
],
"size": [
650,
70
],
"flags": {},
"order": 10,
"mode": 0,
"inputs": [
{
"localized_name": "vae名称",
"name": "vae_name",
"type": "COMBO",
"widget": {
"name": "vae_name"
},
"link": 224
}
],
"outputs": [
{
"localized_name": "VAE",
"name": "VAE",
"type": "VAE",
"links": [
23
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "VAELoader",
"models": [
{
"name": "minimax_h3_audio_vae_fp32.safetensors",
"url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors",
"directory": "vae"
}
]
},
"widgets_values": [
"minimax_h3_audio_vae_fp32.safetensors"
]
},
{
"id": 23,
"type": "VAEDecodeAudio",
"pos": [
-50,
4880
],
"size": [
230,
60
],
"flags": {
"collapsed": false
},
"order": 9,
"mode": 0,
"inputs": [
{
"localized_name": "Latent",
"name": "samples",
"type": "LATENT",
"link": 226
},
{
"localized_name": "vae",
"name": "vae",
"type": "VAE",
"link": 23
}
],
"outputs": [
{
"localized_name": "音频",
"name": "AUDIO",
"type": "AUDIO",
"links": [
166
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "VAEDecodeAudio"
},
"widgets_values": []
},
{
"id": 10,
"type": "VAEDecode",
"pos": [
-50,
4760
],
"size": [
230,
60
],
"flags": {
"collapsed": false
},
"order": 3,
"mode": 0,
"inputs": [
{
"localized_name": "Latent",
"name": "samples",
"type": "LATENT",
"link": 225
},
{
"localized_name": "vae",
"name": "vae",
"type": "VAE",
"link": 8
}
],
"outputs": [
{
"localized_name": "图像",
"name": "IMAGE",
"type": "IMAGE",
"links": [
167
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "VAEDecode"
},
"widgets_values": []
},
{
"id": 9,
"type": "BasicScheduler",
"pos": [
-790,
5030
],
"size": [
370,
130
],
"flags": {},
"order": 2,
"mode": 0,
"inputs": [
{
"localized_name": "模型",
"name": "model",
"type": "MODEL",
"link": 230
}
],
"outputs": [
{
"localized_name": "Sigmas",
"name": "SIGMAS",
"type": "SIGMAS",
"links": [
18
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "BasicScheduler"
},
"widgets_values": [
"simple",
8,
1
]
},
{
"id": 14,
"type": "SamplerCustomAdvanced",
"pos": [
-312.27272727272725,
4817.727272727273
],
"size": [
230,
140
],
"flags": {},
"order": 6,
"mode": 0,
"inputs": [
{
"localized_name": "噪波",
"name": "noise",
"type": "NOISE",
"link": 40
},
{
"localized_name": "引导器",
"name": "guider",
"type": "GUIDER",
"link": 12
},
{
"localized_name": "采样器",
"name": "sampler",
"type": "SAMPLER",
"link": 231
},
{
"localized_name": "西格玛",
"name": "sigmas",
"type": "SIGMAS",
"link": 18
},
{
"localized_name": "Latent图像",
"name": "latent_image",
"type": "LATENT",
"link": 188
}
],
"outputs": [
{
"localized_name": "Latent",
"name": "output",
"type": "LATENT",
"links": [
225,
226
]
},
{
"localized_name": "降噪Latent",
"name": "denoised_output",
"type": "LATENT",
"links": null
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "SamplerCustomAdvanced"
},
"widgets_values": []
},
{
"id": 16,
"type": "BasicGuider",
"pos": [
-790,
4800
],
"size": [
360,
60
],
"flags": {},
"order": 8,
"mode": 0,
"inputs": [
{
"localized_name": "模型",
"name": "model",
"type": "MODEL",
"link": 229
},
{
"localized_name": "条件",
"name": "conditioning",
"type": "CONDITIONING",
"link": 187
}
],
"outputs": [
{
"localized_name": "引导器",
"name": "GUIDER",
"type": "GUIDER",
"links": [
12
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "BasicGuider"
},
"widgets_values": []
},
{
"id": 6,
"type": "UNETLoader",
"pos": [
-2020,
4630
],
"size": [
640,
90
],
"flags": {},
"order": 1,
"mode": 0,
"inputs": [
{
"localized_name": "UNet名称",
"name": "unet_name",
"type": "COMBO",
"widget": {
"name": "unet_name"
},
"link": 221
}
],
"outputs": [
{
"localized_name": "模型",
"name": "MODEL",
"type": "MODEL",
"links": [
228
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "UNETLoader",
"models": [
{
"name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
"url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
"directory": "diffusion_models"
}
]
},
"widgets_values": [
"minimax_h3_fl2va_pruned_int8_convrot.safetensors",
"default"
]
},
{
"id": 13,
"type": "CLIPLoader",
"pos": [
-2020,
4780
],
"size": [
640,
120
],
"flags": {},
"order": 5,
"mode": 0,
"inputs": [
{
"localized_name": "CLIP名称",
"name": "clip_name",
"type": "COMBO",
"widget": {
"name": "clip_name"
},
"link": 227
}
],
"outputs": [
{
"localized_name": "CLIP",
"name": "CLIP",
"type": "CLIP",
"links": [
189
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "CLIPLoader",
"models": [
{
"name": "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
"url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
"directory": "text_encoders"
}
]
},
"widgets_values": [
"qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
"minimax",
"default"
]
},
{
"id": 15,
"type": "RandomNoise",
"pos": [
-790,
4660
],
"size": [
360,
90
],
"flags": {},
"order": 7,
"mode": 0,
"inputs": [
{
"localized_name": "噪波随机种",
"name": "noise_seed",
"type": "INT",
"widget": {
"name": "noise_seed"
},
"link": 207
}
],
"outputs": [
{
"localized_name": "噪波",
"name": "NOISE",
"type": "NOISE",
"links": [
40
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "RandomNoise"
},
"widgets_values": [
1,
"randomize"
]
},
{
"id": 91,
"type": "CreateVideo",
"pos": [
260,
4790
],
"size": [
270,
110
],
"flags": {},
"order": 11,
"mode": 0,
"inputs": [
{
"localized_name": "图像",
"name": "images",
"type": "IMAGE",
"link": 167
},
{
"localized_name": "音频",
"name": "audio",
"shape": 7,
"type": "AUDIO",
"link": 166
}
],
"outputs": [
{
"localized_name": "视频",
"name": "VIDEO",
"type": "VIDEO",
"links": [
168
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "CreateVideo"
},
"widgets_values": [
24,
8
]
},
{
"id": 104,
"type": "MiniMaxH3ImageToVideo",
"pos": [
-1303.8429752066131,
5222.210743801651
],
"size": [
410,
510
],
"flags": {},
"order": 12,
"mode": 0,
"inputs": [
{
"localized_name": "clip",
"name": "clip",
"type": "CLIP",
"link": 189
},
{
"localized_name": "vae",
"name": "vae",
"type": "VAE",
"link": 190
},
{
"localized_name": "first_frame",
"name": "first_frame",
"shape": 7,
"type": "IMAGE",
"link": 195
},
{
"localized_name": "last_frame",
"name": "last_frame",
"shape": 7,
"type": "IMAGE",
"link": 196
},
{
"localized_name": "prompt",
"name": "prompt",
"type": "STRING",
"widget": {
"name": "prompt"
},
"link": 197
},
{
"localized_name": "width",
"name": "width",
"type": "INT",
"widget": {
"name": "width"
},
"link": 200
},
{
"localized_name": "height",
"name": "height",
"type": "INT",
"widget": {
"name": "height"
},
"link": 201
},
{
"localized_name": "length",
"name": "length",
"type": "INT",
"widget": {
"name": "length"
},
"link": 199
}
],
"outputs": [
{
"localized_name": "positive",
"name": "positive",
"type": "CONDITIONING",
"links": [
187
]
},
{
"localized_name": "Latent",
"name": "LATENT",
"type": "LATENT",
"links": [
188
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "MiniMaxH3ImageToVideo"
},
"widgets_values": [
"Vaporwave title sequence look: pink and blue gradient palette, VHS tracking artifacts, Greek statue motifs, chrome palm trees, RGB chromatic aberration, lo-fi retro atmosphere, mood languid and nostalgic.\n\nTimeline:\n[0s-1s] VHS static opens the frame, the title \"COMFYUI\" appears with RGB split and a slight horizontal jitter.\n[1s-2.5s] Hard cut, a Greek plaster bust close-up, pink-purple gradient sky, a pixelated sun.\n[2.5s-4s] Clean \"STARRING\" credits appear, \"LATENT\" and \"CONTROLNET\" each shown exactly once.\n[4s-5s] Final card \"DIRECTED BY COMFYUI\" holds, one VHS tracking glitch settling into stability.\n\nHard cuts only, transitions landing with tape jumps, no push-ins, no dissolves.\n\nAudio: lo-fi vaporwave score, slow drum machine with soft bass, VHS tape-noise sample joins at 2.5s, melody fading for the last 1s.\n\nAll text must be clearly legible, do not misspell English, no Chinese characters, do not repeat names or job titles, no soft dissolves, no subtitle bars.",
1344,
768,
73
]
},
{
"id": 107,
"type": "ComfyMathExpression",
"pos": [
-1710,
5300
],
"size": [
360,
160
],
"flags": {
"collapsed": false
},
"order": 13,
"mode": 0,
"inputs": [
{
"label": "a",
"localized_name": "values.a",
"name": "values.a",
"type": "FLOAT,INT,BOOLEAN",
"link": 205
},
{
"label": "b",
"localized_name": "values.b",
"name": "values.b",
"shape": 7,
"type": "FLOAT,INT,BOOLEAN",
"link": null
}
],
"outputs": [
{
"localized_name": "浮点",
"name": "FLOAT",
"type": "FLOAT",
"links": null
},
{
"localized_name": "整数",
"name": "INT",
"type": "INT",
"links": [
199
]
},
{
"localized_name": "布尔值",
"name": "BOOL",
"type": "BOOLEAN",
"links": null
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "ComfyMathExpression"
},
"widgets_values": [
"max(5, round(a * 24)) + (5 - (max(5, round(a * 24)) % 17)) % 17"
]
},
{
"id": 111,
"type": "PrimitiveFloat",
"pos": [
-2020,
5300
],
"size": [
270,
70
],
"flags": {},
"order": 14,
"mode": 0,
"inputs": [
{
"localized_name": "值",
"name": "value",
"type": "FLOAT",
"widget": {
"name": "value"
},
"link": 206
}
],
"outputs": [
{
"localized_name": "浮点",
"name": "FLOAT",
"type": "FLOAT",
"links": [
205
]
}
],
"title": "Float (duration)",
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "PrimitiveFloat"
},
"widgets_values": [
2
]
},
{
"id": 121,
"type": "LoraLoaderModelOnly",
"pos": [
-1177.5675743948764,
4786.736042457064
],
"size": [
270,
89.875
],
"flags": {},
"order": 15,
"mode": 0,
"inputs": [
{
"localized_name": "模型",
"name": "model",
"type": "MODEL",
"link": 228
}
],
"outputs": [
{
"localized_name": "模型",
"name": "MODEL",
"type": "MODEL",
"links": [
229,
230
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.30.0",
"Node name for S&R": "LoraLoaderModelOnly"
},
"widgets_values": [
"minimax_h3_turbo_v4_step600_ema.safetensors",
1
]
},
{
"id": 122,
"type": "MiniMaxH3DualClockEulerSampler",
"pos": [
-356.0299456339103,
5067.791826578954
],
"size": [
326.265625,
57.921875
],
"flags": {},
"order": 0,
"mode": 0,
"inputs": [],
"outputs": [
{
"localized_name": "sampler",
"name": "sampler",
"type": "SAMPLER",
"links": [
231
]
}
],
"properties": {
"aux_id": "shuaixn/ComfyUI-MiniMaxH3DualClockSampler",
"ver": "986f8e9ebd4477caa6395eead23d968648cc7a7d",
"Node name for S&R": "MiniMaxH3DualClockEulerSampler"
},
"widgets_values": []
}
],
"groups": [
{
"id": 1,
"title": "Models",
"bounding": [
-2050,
4540,
700,
670
],
"color": "#3f789e",
"flags": {}
},
{
"id": 2,
"title": "Sampling",
"bounding": [
-810,
4540,
690,
670
],
"color": "#3f789e",
"flags": {}
},
{
"id": 3,
"title": "Conditioning",
"bounding": [
-1333.8429752066131,
5112.210743801651,
480,
670
],
"color": "#3f789e",
"flags": {}
},
{
"id": 4,
"title": "Decoding and create video",
"bounding": [
-90,
4540,
670,
670
],
"color": "#3f789e",
"flags": {}
}
],
"links": [
{
"id": 23,
"origin_id": 24,
"origin_slot": 0,
"target_id": 23,
"target_slot": 1,
"type": "VAE"
},
{
"id": 8,
"origin_id": 11,
"origin_slot": 0,
"target_id": 10,
"target_slot": 1,
"type": "VAE"
},
{
"id": 40,
"origin_id": 15,
"origin_slot": 0,
"target_id": 14,
"target_slot": 0,
"type": "NOISE"
},
{
"id": 12,
"origin_id": 16,
"origin_slot": 0,
"target_id": 14,
"target_slot": 1,
"type": "GUIDER"
},
{
"id": 18,
"origin_id": 9,
"origin_slot": 0,
"target_id": 14,
"target_slot": 3,
"type": "SIGMAS"
},
{
"id": 188,
"origin_id": 104,
"origin_slot": 1,
"target_id": 14,
"target_slot": 4,
"type": "LATENT"
},
{
"id": 187,
"origin_id": 104,
"origin_slot": 0,
"target_id": 16,
"target_slot": 1,
"type": "CONDITIONING"
},
{
"id": 167,
"origin_id": 10,
"origin_slot": 0,
"target_id": 91,
"target_slot": 0,
"type": "IMAGE"
},
{
"id": 166,
"origin_id": 23,
"origin_slot": 0,
"target_id": 91,
"target_slot": 1,
"type": "AUDIO"
},
{
"id": 189,
"origin_id": 13,
"origin_slot": 0,
"target_id": 104,
"target_slot": 0,
"type": "CLIP"
},
{
"id": 190,
"origin_id": 11,
"origin_slot": 0,
"target_id": 104,
"target_slot": 1,
"type": "VAE"
},
{
"id": 168,
"origin_id": 91,
"origin_slot": 0,
"target_id": -20,
"target_slot": 0,
"type": "VIDEO"
},
{
"id": 195,
"origin_id": -10,
"origin_slot": 0,
"target_id": 104,
"target_slot": 2,
"type": "IMAGE"
},
{
"id": 196,
"origin_id": -10,
"origin_slot": 1,
"target_id": 104,
"target_slot": 3,
"type": "IMAGE"
},
{
"id": 197,
"origin_id": -10,
"origin_slot": 2,
"target_id": 104,
"target_slot": 4,
"type": "STRING"
},
{
"id": 199,
"origin_id": 107,
"origin_slot": 1,
"target_id": 104,
"target_slot": 7,
"type": "INT"
},
{
"id": 200,
"origin_id": -10,
"origin_slot": 3,
"target_id": 104,
"target_slot": 5,
"type": "INT"
},
{
"id": 201,
"origin_id": -10,
"origin_slot": 4,
"target_id": 104,
"target_slot": 6,
"type": "INT"
},
{
"id": 205,
"origin_id": 111,
"origin_slot": 0,
"target_id": 107,
"target_slot": 0,
"type": "FLOAT"
},
{
"id": 206,
"origin_id": -10,
"origin_slot": 5,
"target_id": 111,
"target_slot": 0,
"type": "FLOAT"
},
{
"id": 207,
"origin_id": -10,
"origin_slot": 6,
"target_id": 15,
"target_slot": 0,
"type": "INT"
},
{
"id": 221,
"origin_id": -10,
"origin_slot": 7,
"target_id": 6,
"target_slot": 0,
"type": "COMBO"
},
{
"id": 223,
"origin_id": -10,
"origin_slot": 9,
"target_id": 11,
"target_slot": 0,
"type": "COMBO"
},
{
"id": 224,
"origin_id": -10,
"origin_slot": 10,
"target_id": 24,
"target_slot": 0,
"type": "COMBO"
},
{
"id": 225,
"origin_id": 14,
"origin_slot": 0,
"target_id": 10,
"target_slot": 0,
"type": "LATENT"
},
{
"id": 226,
"origin_id": 14,
"origin_slot": 0,
"target_id": 23,
"target_slot": 0,
"type": "LATENT"
},
{
"id": 227,
"origin_id": -10,
"origin_slot": 8,
"target_id": 13,
"target_slot": 0,
"type": "COMBO"
},
{
"id": 228,
"origin_id": 6,
"origin_slot": 0,
"target_id": 121,
"target_slot": 0,
"type": "MODEL"
},
{
"id": 229,
"origin_id": 121,
"origin_slot": 0,
"target_id": 16,
"target_slot": 0,
"type": "MODEL"
},
{
"id": 230,
"origin_id": 121,
"origin_slot": 0,
"target_id": 9,
"target_slot": 0,
"type": "MODEL"
},
{
"id": 231,
"origin_id": 122,
"origin_slot": 0,
"target_id": 14,
"target_slot": 2,
"type": "SAMPLER"
}
],
"extra": {}
}
]
},
"config": {},
"extra": {
"frontendVersion": "1.48.6",
"VHS_latentpreview": false,
"VHS_latentpreviewrate": 0,
"VHS_MetadataImage": true,
"VHS_KeepIntermediate": true,
"ds": {
"scale": 0.9598932243248864,
"offset": [
1904.9493458710508,
-4228.384242491727
]
}
},
"version": 0.4
}6、FlashVSR放大工作流：
bash
{
"id": "a870e8d5-f195-429f-b978-ce9c8de30718",
"revision": 0,
"last_node_id": 4,
"last_link_id": 7,
"nodes": [
{
"id": 2,
"type": "AILab_FlashVSR_Advanced",
"pos": [
389.534414565678,
-138.91610176301108
],
"size": [
273.75,
497.453125
],
"flags": {
"collapsed": false
},
"order": 1,
"mode": 0,
"inputs": [
{
"name": "frames",
"type": "IMAGE",
"link": 6
},
{
"name": "audio",
"shape": 7,
"type": "AUDIO",
"link": 7
}
],
"outputs": [
{
"name": "frames",
"type": "IMAGE",
"links": [
3
]
},
{
"name": "audio",
"type": "AUDIO",
"links": [
4
]
}
],
"properties": {
"cnr_id": "ComfyUI-FlashVSR",
"ver": "8877fdd593ea93b27353956dc69edf423c561fee",
"Node name for S&R": "AILab_FlashVSR_Advanced"
},
"widgets_values": [
"Full (Best Quality)",
2,
false,
384,
24,
2,
3,
11,
true,
true,
false,
"enable",
"auto",
"bf16",
345607175730854,
"randomize"
]
},
{
"id": 3,
"type": "VHS_VideoCombine",
"pos": [
692.6316374020166,
-82.9834434401472
],
"size": [
225,
536.328125
],
"flags": {},
"order": 2,
"mode": 0,
"inputs": [
{
"name": "images",
"type": "IMAGE",
"link": 3
},
{
"name": "audio",
"shape": 7,
"type": "AUDIO",
"link": 4
},
{
"name": "meta_batch",
"shape": 7,
"type": "VHS_BatchManager",
"link": null
},
{
"name": "vae",
"shape": 7,
"type": "VAE",
"link": null
}
],
"outputs": [
{
"name": "Filenames",
"type": "VHS_FILENAMES",
"links": null
}
],
"properties": {
"cnr_id": "comfyui-videohelpersuite",
"ver": "1.7.9",
"Node name for S&R": "VHS_VideoCombine"
},
"widgets_values": {
"frame_rate": 24,
"loop_count": 0,
"filename_prefix": "AnimateDiff",
"format": "video/h265-mp4",
"pix_fmt": "yuv420p10le",
"crf": 18,
"save_metadata": true,
"pingpong": false,
"save_output": true,
"videopreview": {
"hidden": false,
"paused": false,
"params": {
"filename": "AnimateDiff_00002-audio.mp4",
"subfolder": "",
"type": "output",
"format": "video/h265-mp4",
"frame_rate": 24,
"workflow": "AnimateDiff_00002.png",
"fullpath": "/home/models/FireRed-Image-Edit-1.1-ComfyUI/ComfyUI-master/output/AnimateDiff_00002-audio.mp4"
}
}
}
},
{
"id": 4,
"type": "VHS_LoadVideoFFmpeg",
"pos": [
49.67605693104996,
-107.75292639896139
],
"size": [
313.4375,
511.40625
],
"flags": {},
"order": 0,
"mode": 0,
"inputs": [
{
"name": "meta_batch",
"shape": 7,
"type": "VHS_BatchManager",
"link": null
},
{
"name": "vae",
"shape": 7,
"type": "VAE",
"link": null
}
],
"outputs": [
{
"name": "IMAGE",
"type": "IMAGE",
"links": [
6
]
},
{
"name": "mask",
"type": "MASK",
"links": null
},
{
"name": "audio",
"type": "AUDIO",
"links": [
7
]
},
{
"name": "video_info",
"type": "VHS_VIDEOINFO",
"links": null
}
],
"properties": {
"cnr_id": "comfyui-videohelpersuite",
"ver": "1.7.9",
"Node name for S&R": "VHS_LoadVideoFFmpeg"
},
"widgets_values": {
"video": "MiniMax_H3_00009_.mp4",
"force_rate": 0,
"custom_width": 0,
"custom_height": 0,
"frame_load_cap": 0,
"start_time": 0,
"format": "AnimateDiff",
"videopreview": {
"hidden": false,
"paused": false,
"params": {
"force_rate": 0,
"custom_width": 0,
"custom_height": 0,
"frame_load_cap": 0,
"start_time": 0,
"filename": "MiniMax_H3_00009_.mp4",
"type": "input",
"format": "video/mp4"
}
}
}
}
],
"links": [
[
3,
2,
0,
3,
0,
"IMAGE"
],
[
4,
2,
1,
3,
1,
"AUDIO"
],
[
6,
4,
0,
2,
0,
"IMAGE"
],
[
7,
4,
2,
2,
1,
"AUDIO"
]
],
"groups": [],
"config": {},
"extra": {
"ds": {
"scale": 1.4641000000000008,
"offset": [
36.91062497247819,
179.4657356716618
]
},
"frontendVersion": "1.48.6",
"VHS_latentpreview": false,
"VHS_latentpreviewrate": 0,
"VHS_MetadataImage": true,
"VHS_KeepIntermediate": true
},
"version": 0.4
}7、Lora工作流生成视频如下：
MiniMax_H3_Lora生成视频效果
8、FlashVSR工作流生成视频如下：
FlashVSR放大视频效果
四、踩坑避坑指南（详细版）
4.1 环境与依赖相关
坑1：ComfyUI版本过低导致H3节点无法识别
-
现象：导入工作流后，MiniMax-H3相关节点显示为红色或缺失，无法加载。
-
根因 ：MiniMax-H3的原生支持是在ComfyUI v0.30.0 版本中才合并的。
-
解决方案 ：确保ComfyUI版本为 v0.30.0或更高 。若使用便携版，请下载官方包
ComfyUI_windows_portable_nvidia.7z（自带torch 2.9.1+cu130）。
坑2：依赖包缺失导致FlashVSR节点报错
-
现象 ：运行FlashVSR节点时报
ModuleNotFoundError，提示缺少block_sparse_attn等模块。 -
根因：FlashVSR依赖特定的Python wheel包，未随节点自动安装。
-
解决方案：手动安装依赖：
bashpip install block_sparse_attn-*.whl # 匹配py3.12 + torch2.8+ pip install opencv-python numpy einops坑3：FlashVSR模型自动下载失败
-
现象：首次运行FlashVSR节点时，模型下载卡住或报错。
-
根因：Hugging Face国内访问不稳定。
-
解决方案：
-
方案一：从 hf-mirror.com 或 ModelScope 镜像站手动下载
1038lab/FlashVSR仓库中的模型文件。 -
方案二：删除
ComfyUI/models/FlashVSR/文件夹，让ComfyUI重新尝试下载。
-
4.2 模型与LoRA相关
坑4：选错了主模型版本导致LoRA无法生效
-
现象：加载Turbo LoRA后，生成效果无明显变化，或报错提示层结构不匹配。
-
根因 ：Turbo LoRA要求主模型必须保留完整的AdaLN层结构 ，而名称中带
pruned（剪枝）的版本可能移除了LoRA所依赖的结构。 -
解决方案 ：选择 完整（非pruned） 的主模型：
bashminimax_h3_fl2va_bf16.safetensors（BF16精度） minimax_h3_fl2va_int8_convrot.safetensors（INT8 ConvRot版本） 不要使用 *_pruned_* 版本。坑5：Turbo LoRA键名缺少
diffusion_model.前缀 -
现象：LoRA加载成功但生成效果异常，或模型完全不受LoRA影响。
-
根因 ：作者发布的原始LoRA文件键名缺少ComfyUI所需的
diffusion_model.前缀。 -
解决方案：使用官方提供的转换脚本处理LoRA文件：
bashpython custom_nodes/ComfyUI-MiniMaxH3DualClockSampler/convert_h3_lora_for_comfyui.py \ /path/to/minimax_h3_turbo_v4_step600_ema.safetensors \ models/loras/minimax_h3_turbo_v4_step600_ema.safetensors转换只修改键名前缀，不改变tensor数值、dtype、shape或rank。
坑6：Turbo LoRA + 有音频输入时的Tensor形状不匹配
-
现象 ：使用
ref2va等涉及音频输入的工作流时，报RuntimeError: shape mismatch。 -
根因：当音频存在时，Turbo LoRA的patch矩阵会携带3个batch/modulation segments，导致按维度0进行逐元素相加时失败。
-
解决方案 ：关注
ComfyUI-MiniMax-H3-Turbo仓库的最新修复，或暂时避免在有音频输入的场景下使用Turbo LoRA。
4.3 生成与采样相关
坑7：SageAttention导致H3输出纯噪声
-
现象 ：启用
--use-sage-attention后，生成的视频和音频均为纯噪声，无任何有效内容。 -
根因 ：H3 DiT的attention调用未传递
low_precision_attention=False参数，导致SageAttention的INT8 QK路径错误处理了H3的attention。 -
解决方案 ：该问题已在ComfyUI主仓库的PR #15263中修复。更新ComfyUI至包含该修复的最新版本即可。作为临时方案，可禁用
--use-sage-attention。
坑8：长提示词导致H3崩溃（C++ assertion abort）
-
现象 ：使用较长提示词时，
MiniMaxH3MemoryEfficientSageAttentionPatch节点触发C++断言崩溃。 -
根因 ：当
kv_len % 128 ∈ [65, 127]时，SageAttention的SM90分支会触发断言失败。 -
解决方案 ：该问题已在PR #729中修复。更新
ComfyUI-KJNodes至最新版本。
坑9：4步Turbo生成出现严重音频爆音
-
现象：使用Turbo LoRA以4步生成时，音频出现严重爆音、削波、噪声化。
-
根因：标准Euler采样器在4步低步数下无法妥善处理音视频联合生成的时序一致性。
-
解决方案 ：使用 双时钟Euler采样器 （
MiniMaxH3DualClockEulerSampler）替换标准采样器。该节点专门为解决Turbo LoRA的音频问题而设计。安装方式：bashcd ComfyUI/custom_nodes git clone https://github.com/shuaixn/ComfyUI-MiniMaxH3DualClockSampler.git注意：不要将双时钟节点与
MiniMax H3 Sigma Shift节点同时使用。
坑10：VAE解码时报设备不匹配错误
-
现象 ：采样完成后，VAE解码阶段报错：
RuntimeError: Expected all tensors to be on the same device, but got tensors is on cpu, different from other tensors on cuda:0。 -
根因 ：H3的Video VAE和Audio VAE在forward中直接访问了多个
nn.Parameters（如register_tokens、scale1/scale2、Snake1d.alpha等），绕过了ComfyUI的weight-cast机制。当VAE在offload/manual-cast模式下运行时，这些参数留在CPU而激活张量在GPU上，导致设备不匹配。 -
解决方案：该问题已在PR #15268中修复。更新ComfyUI至包含该修复的版本。若临时规避，可尝试在显存充裕时运行（避免VAE进入offload模式）。
4.4 视频放大相关（FlashVSR）
坑11：FlashVSR输出视频只有声音没有图像
-
现象：运行FlashVSR工作流后，输出的视频文件只有音频轨道，画面黑屏或无画面。
-
根因 ：视频加载节点（如
Load Video (Upload)）可能存在解码异常，只读取了音频轨道。 -
解决方案 ：切换视频加载节点，使用
VHS_LoadVideoFFmpeg节点（来自VideoHelperSuite），其兼容性更好。
坑12：FlashVSR长视频处理时显存溢出（OOM）
-
现象：处理超过约100帧的视频时，触发CUDA OOM崩溃。
-
根因：FlashVSR默认一次性加载所有帧到显存进行处理。
-
解决方案：
-
启用
enable_tiling=true+tile_size=384（12GB低显存模式） -
设置
frame_chunk_size为较小值（如20-50），分批次处理 -
启用
unload_dit选项，在解码前卸载DiT以降低显存峰值 -
使用
resize_factor=0.5先缩小输入再处理
-
坑13：FlashVSR输出画面出现黑边或画面撕裂
-
现象：放大后的视频边缘出现黑边，或画面出现视觉伪影、撕裂。
-
根因：旧版本FlashVSR存在Padding→Process→Crop逻辑缺陷，以及Tensor排列顺序（B,C,F,H,W vs B,H,W,C）处理错误。
-
解决方案 ：更新ComfyUI-FlashVSR至 v1.1.0或更高版本，该版本已修复黑边和画面撕裂问题。
4.5 性能与资源相关
坑14：768P生成速度极慢，"一步到位"不划算
-
现象：直接生成768P（1344×768）视频耗时极长，试错成本高。
-
根因：768P的像素量是480p的约2.5倍（41.5万→103.2万），加上显存卸载开销，实际耗时放大到3-5倍。
-
解决方案 ：采用 "480p生产、768p精选" 的分阶段策略：
-
构图预览：832×480，5s，8-12步 → 快速筛选提示词
-
质量测试：1024×576，5s，16-20步 → 判断动作/人物一致性
-
最终成片 ：1344×768，5s，20步 → 仅对入选镜头重跑
480p试5次≈15分钟，768p试5次≈1小时+。
-
坑15：Turbo LoRA的步数不是越少越好
-
现象：4步生成虽然最快，但画面出现塑料感皮肤、过度锐化的颗粒感等已知伪影。
-
根因：Turbo LoRA的当前checkpoint在4步极端压缩下，画质与细节的平衡尚未达到完美。
-
解决方案 ：建议使用 6-8步 作为速度与画质的平衡点。若追求极致速度可尝试4步，但需接受画质妥协。
坑16：模型总下载量约40GB，需预留足够空间
-
现象：下载过程中磁盘空间不足导致下载失败。
-
根因：完整模型组合约40GB，包括：FL2VA pruned INT8 DiT（19.5GB）+ Qwen3-VL NVFP4编码器（14.6GB）+ 双VAE（5.4GB）。
-
解决方案 ：确保
ComfyUI/models/所在磁盘至少有 50GB 以上可用空间。建议使用 ModelScope 国内镜像直连下载，支持断点续传。
4.6 海光DCU特殊适配相关
坑17：海光DCU上PyTorch未正确识别DCU
-
现象：ComfyUI启动后显示未检测到CUDA设备，或只能使用CPU模式。
-
根因：海光DCU使用ROCm而非NVIDIA CUDA，需要安装对应版本的ROCm PyTorch。
-
解决方案：
-
确认已安装正确的ROCm驱动和DTK（DCU Toolkit）
-
安装ROCm版本的PyTorch（如
torch-2.9.1+rocm） -
启动ComfyUI时添加
--disable-cuda-malloc参数
-
五、总结与展望
本文成功验证了一条在消费级算力（64GB显存）上，利用开源H3模型生成高清视频的可行路径：LoRA加速解决了"生成慢"的问题，FlashVSR解决了"分辨率低"的问题。这套组合方案使得在有限硬件上高效产出2K级AI视频成为可能。
展望 ：未来，随着官方H3-Regenerate-2K等原生高清技术的普及，以及社区生态的完善，AI视频生成将更加高效、高清，进一步降低高质量视频创作的门槛。