| { | |
| "last_node_id": 5, | |
| "last_link_id": 4, | |
| "nodes": [ | |
| { | |
| "id": 1, | |
| "type": "LoadAudio", | |
| "title": "Audio to describe", | |
| "pos": [ | |
| 30, | |
| 80 | |
| ], | |
| "size": [ | |
| 420, | |
| 300 | |
| ], | |
| "flags": {}, | |
| "order": 0, | |
| "mode": 0, | |
| "inputs": [], | |
| "outputs": [ | |
| { | |
| "name": "AUDIO", | |
| "type": "AUDIO", | |
| "links": [ | |
| 1 | |
| ], | |
| "slot_index": 0 | |
| } | |
| ], | |
| "properties": { | |
| "Node name for S&R": "LoadAudio" | |
| }, | |
| "widgets_values": [ | |
| "moss_audio_demo.wav" | |
| ] | |
| }, | |
| { | |
| "id": 2, | |
| "type": "CLIPLoader", | |
| "title": "MOSS-Audio 8B Thinking \u00b7 BF16", | |
| "pos": [ | |
| 30, | |
| 550 | |
| ], | |
| "size": [ | |
| 420, | |
| 150 | |
| ], | |
| "flags": {}, | |
| "order": 1, | |
| "mode": 0, | |
| "inputs": [], | |
| "outputs": [ | |
| { | |
| "name": "CLIP", | |
| "type": "CLIP", | |
| "links": [ | |
| 2 | |
| ], | |
| "slot_index": 0 | |
| } | |
| ], | |
| "properties": { | |
| "Node name for S&R": "CLIPLoader" | |
| }, | |
| "widgets_values": [ | |
| "moss_audio_8b_instruct_bf16.safetensors", | |
| "moss_audio", | |
| "default" | |
| ] | |
| }, | |
| { | |
| "id": 3, | |
| "type": "TextGenerate", | |
| "title": "Describe audio", | |
| "pos": [ | |
| 590, | |
| 80 | |
| ], | |
| "size": [ | |
| 520, | |
| 720 | |
| ], | |
| "flags": {}, | |
| "order": 2, | |
| "mode": 0, | |
| "inputs": [ | |
| { | |
| "name": "clip", | |
| "type": "CLIP", | |
| "link": 2 | |
| }, | |
| { | |
| "name": "image", | |
| "type": "IMAGE", | |
| "link": null | |
| }, | |
| { | |
| "name": "video", | |
| "type": "IMAGE", | |
| "link": null | |
| }, | |
| { | |
| "name": "audio", | |
| "type": "AUDIO", | |
| "link": 1 | |
| }, | |
| { | |
| "name": "system_prompt", | |
| "type": "STRING", | |
| "link": null | |
| } | |
| ], | |
| "outputs": [ | |
| { | |
| "name": "generated_text", | |
| "type": "STRING", | |
| "links": [ | |
| 3 | |
| ], | |
| "slot_index": 0 | |
| }, | |
| { | |
| "name": "thinking", | |
| "type": "STRING", | |
| "links": [ | |
| 4 | |
| ], | |
| "slot_index": 1 | |
| } | |
| ], | |
| "properties": { | |
| "Node name for S&R": "TextGenerate" | |
| }, | |
| "widgets_values": [ | |
| "Write a detailed audio caption in 3\u20135 sentences. Describe foreground and background sounds and their order, distinguishing speech, music and environmental sounds. Include only audible evidence and state uncertainty about ambiguous sounds. Do not give exact timestamps.", | |
| 2048, | |
| "on", | |
| 1.0, | |
| 50, | |
| 1.0, | |
| 0.0, | |
| 1.0, | |
| 42, | |
| "fixed", | |
| 0.0, | |
| false, | |
| true, | |
| "auto" | |
| ] | |
| }, | |
| { | |
| "id": 4, | |
| "type": "PreviewAny", | |
| "title": "Audio description", | |
| "pos": [ | |
| 1250, | |
| 80 | |
| ], | |
| "size": [ | |
| 500, | |
| 340 | |
| ], | |
| "flags": {}, | |
| "order": 3, | |
| "mode": 0, | |
| "inputs": [ | |
| { | |
| "name": "source", | |
| "type": "*", | |
| "link": 3 | |
| } | |
| ], | |
| "outputs": [ | |
| { | |
| "name": "STRING", | |
| "type": "STRING", | |
| "links": [], | |
| "slot_index": 0 | |
| } | |
| ], | |
| "properties": { | |
| "Node name for S&R": "PreviewAny" | |
| }, | |
| "widgets_values": [] | |
| }, | |
| { | |
| "id": 5, | |
| "type": "PreviewAny", | |
| "title": "Thinking", | |
| "pos": [ | |
| 1250, | |
| 460 | |
| ], | |
| "size": [ | |
| 500, | |
| 340 | |
| ], | |
| "flags": {}, | |
| "order": 4, | |
| "mode": 0, | |
| "inputs": [ | |
| { | |
| "name": "source", | |
| "type": "*", | |
| "link": 4 | |
| } | |
| ], | |
| "outputs": [ | |
| { | |
| "name": "STRING", | |
| "type": "STRING", | |
| "links": [], | |
| "slot_index": 0 | |
| } | |
| ], | |
| "properties": { | |
| "Node name for S&R": "PreviewAny" | |
| }, | |
| "widgets_values": [] | |
| } | |
| ], | |
| "links": [ | |
| [ | |
| 1, | |
| 1, | |
| 0, | |
| 3, | |
| 3, | |
| "AUDIO" | |
| ], | |
| [ | |
| 2, | |
| 2, | |
| 0, | |
| 3, | |
| 0, | |
| "CLIP" | |
| ], | |
| [ | |
| 3, | |
| 3, | |
| 0, | |
| 4, | |
| 0, | |
| "STRING" | |
| ], | |
| [ | |
| 4, | |
| 3, | |
| 1, | |
| 5, | |
| 0, | |
| "STRING" | |
| ] | |
| ], | |
| "groups": [ | |
| { | |
| "id": 1, | |
| "title": "1 \u00b7 Inputs", | |
| "bounding": [ | |
| 0, | |
| 0, | |
| 480, | |
| 410 | |
| ], | |
| "color": "#357c91", | |
| "font_size": 24, | |
| "flags": {} | |
| }, | |
| { | |
| "id": 2, | |
| "title": "2 \u00b7 Model Loading", | |
| "bounding": [ | |
| 0, | |
| 470, | |
| 480, | |
| 260 | |
| ], | |
| "color": "#63598f", | |
| "font_size": 24, | |
| "flags": {} | |
| }, | |
| { | |
| "id": 3, | |
| "title": "3 \u00b7 Configuration and Audio Description", | |
| "bounding": [ | |
| 560, | |
| 0, | |
| 580, | |
| 830 | |
| ], | |
| "color": "#8c7438", | |
| "font_size": 24, | |
| "flags": {} | |
| }, | |
| { | |
| "id": 4, | |
| "title": "4 \u00b7 Text Outputs", | |
| "bounding": [ | |
| 1220, | |
| 0, | |
| 560, | |
| 830 | |
| ], | |
| "color": "#427852", | |
| "font_size": 24, | |
| "flags": {} | |
| } | |
| ], | |
| "config": {}, | |
| "extra": { | |
| "ds": { | |
| "scale": 0.7, | |
| "offset": [ | |
| 50, | |
| 50 | |
| ] | |
| } | |
| }, | |
| "version": 0.4 | |
| } | |