Upload config.json
#3
by CoreloneH - opened
- config.json +34 -0
config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "Lance",
|
| 3 |
+
"model_size": "3B active parameters",
|
| 4 |
+
"license": "apache-2.0",
|
| 5 |
+
"organization": "bytedance-research",
|
| 6 |
+
"description": "Lance is a lightweight native unified multimodal model trained from scratch for image and video understanding, generation, and editing in a single framework.",
|
| 7 |
+
"supported_tasks": [
|
| 8 |
+
"text-to-image generation",
|
| 9 |
+
"text-to-video generation",
|
| 10 |
+
"image editing",
|
| 11 |
+
"video editing",
|
| 12 |
+
"image understanding",
|
| 13 |
+
"video understanding"
|
| 14 |
+
],
|
| 15 |
+
"task_names": [
|
| 16 |
+
"t2i",
|
| 17 |
+
"t2v",
|
| 18 |
+
"image_edit",
|
| 19 |
+
"video_edit",
|
| 20 |
+
"x2t_image",
|
| 21 |
+
"x2t_video"
|
| 22 |
+
],
|
| 23 |
+
"recommended_environment": {
|
| 24 |
+
"python": ">=3.10",
|
| 25 |
+
"cuda": ">=12.4",
|
| 26 |
+
"gpu_memory": ">=40GB VRAM for inference"
|
| 27 |
+
},
|
| 28 |
+
"checkpoint_directories": [
|
| 29 |
+
"Lance_3B",
|
| 30 |
+
"Lance_3B_Video",
|
| 31 |
+
"Qwen2.5-VL-ViT"
|
| 32 |
+
],
|
| 33 |
+
"homepage": "https://huggingface.co/bytedance-research/Lance"
|
| 34 |
+
}
|