Add DiffusionPipeline support and auto mode retry logic

- Add DiffusionPipeline to PIPELINE_CLASS_MAP for generic model loading
- Add fallback to DiffusionPipeline for unknown pipeline classes
- Add return_all parameter to select_best_model() for getting all candidates
- Store alternative models in auto mode for retry support
- Implement retry logic when model loading fails in auto mode
- Retry up to 3 times with alternative models before failing
- Add debug output for model loading troubleshooting
- Improve error messages with troubleshooting hints

videogen_models.json

+{
+  "models": {
+    "wan_1.3b_i2v": {
+      "id": "Wan-AI/Wan2.1-I2V-1.3B-Diffusers",
+      "vram": "~8–10 GB",
+      "class": "WanPipeline",
+      "desc": "Small Wan I2V variant – NSFW friendly, no filter",
+      "extra": {"use_custom_vae": true},
+      "supports_i2v": true,
+      "tags": ["i2v", "video", "wan", "nsfw-friendly"]
+    },
+    "svd_xt_1.1": {
+      "id": "stabilityai/stable-video-diffusion-img2vid-xt-1-1",
+      "vram": "~14–18 GB",
+      "class": "StableVideoDiffusionPipeline",
+      "desc": "Stable Video Diffusion XT 1.1 – high-quality I2V, NSFW capable",
+      "supports_i2v": true,
+      "tags": ["i2v", "video", "svd", "nsfw-friendly"]
+    },
+    "i2vgen_xl": {
+      "id": "ali-vilab/i2vgen-xl",
+      "vram": "~18–24 GB",
+      "class": "I2VGenXLPipeline",
+      "desc": "Strong image-to-video XL – works with NSFW content",
+      "supports_i2v": true,
+      "tags": ["i2v", "video", "nsfw-friendly"]
+    },
+    "ltx_video": {
+      "id": "Lightricks/LTX-Video",
+      "vram": "~12–16 GB",
+      "class": "LTXVideoPipeline",
+      "desc": "Fast & efficient local video gen – NSFW friendly",
+      "supports_i2v": true,
+      "tags": ["i2v", "video", "ltx", "nsfw-friendly"]
+    },
+    "wan_14b_i2v": {
+      "id": "Wan-AI/Wan2.1-I2V-14B-Diffusers",
+      "vram": "~20–24 GB",
+      "class": "WanPipeline",
+      "desc": "Strong Wan I2V variant",
+      "extra": {"use_custom_vae": true},
+      "supports_i2v": true,
+      "tags": ["i2v", "video", "wan", "nsfw-friendly"]
+    },
+    "pony_uncensored_v6": {
+      "id": "AstraliteHeart/pony-diffusion-v6-xl-uncensored",
+      "vram": "~10–14 GB",
+      "class": "StableDiffusionXLPipeline",
+      "desc": "Uncensored Pony V6 XL – extremely NSFW-capable",
+      "supports_i2v": false,
+      "tags": ["image", "pony", "nsfw", "uncensored", "sdxl"]
+    },
+    "pony_realism_v2.2": {
+      "id": "AstraliteHeart/pony-realism-v2.2",
+      "vram": "~11–15 GB",
+      "class": "StableDiffusionXLPipeline",
+      "desc": "Realistic Pony merge – good for NSFW photoreal",
+      "supports_i2v": false,
+      "tags": ["image", "pony", "nsfw", "realism", "sdxl"]
+    },
+    "flux_nsfw_uncensored": {
+      "id": "lllyasviel/flux-nsfw-uncensored",
+      "vram": "~20–25 GB",
+      "class": "FluxPipeline",
+      "desc": "Uncensored NSFW Flux variant",
+      "supports_i2v": false,
+      "tags": ["image", "flux", "nsfw", "uncensored"]
+    },
+    "realistic_vision_v6": {
+      "id": "SG161222/Realistic_Vision_V6.0_B1_noVAE",
+      "vram": "~12–16 GB",
+      "class": "StableDiffusionXLPipeline",
+      "desc": "Realistic Vision V6 – strong NSFW realism",
+      "supports_i2v": false,
+      "tags": ["image", "realism", "nsfw-friendly", "sdxl"]
+    }
+  },
+  "search_queries": [
+    "i2v video generation nsfw",
+    "stable video diffusion",
+    "image to video",
+    "animatediff nsfw",
+    "svd fine-tune",
+    "wan i2v",
+    "ltx video",
+    "cogvideox"
+  ]
+}
\ No newline at end of file