{ "last_node_id": 15, "last_link_id": 15, "nodes": [ { "id": 1, "type": "UNETLoader", "title": "AuK-Flash transformer", "pos": [ 30, 80 ], "size": [ 360, 130 ], "flags": {}, "order": 0, "mode": 0, "inputs": [], "outputs": [ { "name": "MODEL", "type": "MODEL", "links": [ 2 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "UNETLoader" }, "widgets_values": [ "auk_flash.safetensors", "default" ] }, { "id": 2, "type": "CLIPLoader", "title": "AuK Qwen conditioning", "pos": [ 30, 250 ], "size": [ 360, 150 ], "flags": {}, "order": 1, "mode": 0, "inputs": [], "outputs": [ { "name": "CLIP", "type": "CLIP", "links": [ 1 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "CLIPLoader" }, "widgets_values": [ "auk_qwen25_omni_3b.safetensors", "auk", "default" ] }, { "id": 3, "type": "VAELoader", "title": "AuK audio VAE", "pos": [ 30, 440 ], "size": [ 360, 100 ], "flags": {}, "order": 2, "mode": 0, "inputs": [], "outputs": [ { "name": "VAE", "type": "VAE", "links": [ 10, 13 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "VAELoader" }, "widgets_values": [ "auk_audio_vae.safetensors" ] }, { "id": 4, "type": "PrimitiveStringMultiline", "title": "Instruction", "pos": [ 530, 80 ], "size": [ 360, 200 ], "flags": {}, "order": 3, "mode": 0, "inputs": [], "outputs": [ { "name": "STRING", "type": "STRING", "links": [ 14 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "PrimitiveStringMultiline" }, "widgets_values": [ "Say the following with the same voice: \"Hello there. This is a short audio test.\"" ] }, { "id": 15, "type": "LoadAudio", "title": "Choose reference audio", "pos": [ 530, 320 ], "size": [ 360, 240 ], "flags": {}, "order": 8, "mode": 0, "inputs": [], "outputs": [ { "name": "AUDIO", "type": "AUDIO", "links": [ 12 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "LoadAudio" }, "widgets_values": [ "reference.wav" ] }, { "id": 5, "type": "PrimitiveFloat", "title": "Target duration (seconds)", "pos": [ 1030, 80 ], "size": [ 360, 100 ], "flags": {}, "order": 4, "mode": 0, "inputs": [], "outputs": [ { "name": "FLOAT", "type": "FLOAT", "links": [ 15 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "PrimitiveFloat" }, "widgets_values": [ 5.0 ] }, { "id": 9, "type": "RandomNoise", "title": "Initial noise seed", "pos": [ 1030, 220 ], "size": [ 360, 120 ], "flags": {}, "order": 5, "mode": 0, "inputs": [], "outputs": [ { "name": "NOISE", "type": "NOISE", "links": [ 4 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "RandomNoise" }, "widgets_values": [ 42, "fixed" ] }, { "id": 10, "type": "KSamplerSelect", "title": "Four-step Euler", "pos": [ 1030, 380 ], "size": [ 360, 100 ], "flags": {}, "order": 6, "mode": 0, "inputs": [], "outputs": [ { "name": "SAMPLER", "type": "SAMPLER", "links": [ 6 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "KSamplerSelect" }, "widgets_values": [ "euler" ] }, { "id": 11, "type": "ManualSigmas", "title": "Exact AuK-Flash schedule", "pos": [ 1030, 520 ], "size": [ 360, 120 ], "flags": {}, "order": 7, "mode": 0, "inputs": [], "outputs": [ { "name": "SIGMAS", "type": "SIGMAS", "links": [ 7 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "ManualSigmas" }, "widgets_values": [ "1.0, 0.9238795042037964, 0.7071067690849304, 0.3826833963394165, 0.0" ] }, { "id": 6, "type": "AuKConditioning", "title": "Encode instruction and reference", "pos": [ 1530, 80 ], "size": [ 360, 210 ], "flags": {}, "order": 9, "mode": 0, "inputs": [ { "name": "clip", "type": "CLIP", "link": 1 }, { "name": "instruction", "type": "STRING", "link": 14, "widget": { "name": "instruction" } }, { "name": "reference_audio", "type": "AUDIO", "link": 12 }, { "name": "vae", "type": "VAE", "link": 13 } ], "outputs": [ { "name": "CONDITIONING", "type": "CONDITIONING", "links": [ 3 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "AuKConditioning" }, "widgets_values": [ "", 0 ] }, { "id": 7, "type": "EmptyAuKLatentAudio", "title": "24 kHz \u00b7 50 latent frames/sec", "pos": [ 1530, 330 ], "size": [ 360, 130 ], "flags": {}, "order": 10, "mode": 0, "inputs": [ { "name": "seconds", "type": "FLOAT", "link": 15, "widget": { "name": "seconds" } } ], "outputs": [ { "name": "LATENT", "type": "LATENT", "links": [ 8 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "EmptyAuKLatentAudio" }, "widgets_values": [ 5.0, 1 ] }, { "id": 8, "type": "BasicGuider", "title": "Conditional only \u00b7 no extra CFG", "pos": [ 2030, 80 ], "size": [ 360, 120 ], "flags": {}, "order": 11, "mode": 0, "inputs": [ { "name": "model", "type": "MODEL", "link": 2 }, { "name": "conditioning", "type": "CONDITIONING", "link": 3 } ], "outputs": [ { "name": "GUIDER", "type": "GUIDER", "links": [ 5 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "BasicGuider" }, "widgets_values": [] }, { "id": 12, "type": "SamplerCustomAdvanced", "title": "Generate audio latents", "pos": [ 2030, 240 ], "size": [ 360, 220 ], "flags": {}, "order": 12, "mode": 0, "inputs": [ { "name": "noise", "type": "NOISE", "link": 4 }, { "name": "guider", "type": "GUIDER", "link": 5 }, { "name": "sampler", "type": "SAMPLER", "link": 6 }, { "name": "sigmas", "type": "SIGMAS", "link": 7 }, { "name": "latent_image", "type": "LATENT", "link": 8 } ], "outputs": [ { "name": "output", "type": "LATENT", "links": [ 9 ], "slot_index": 0 }, { "name": "denoised_output", "type": "LATENT", "links": [], "slot_index": 1 } ], "properties": { "Node name for S&R": "SamplerCustomAdvanced" }, "widgets_values": [] }, { "id": 13, "type": "VAEDecodeAudio", "title": "Decode \u00b7 preserve waveform amplitude", "pos": [ 2530, 80 ], "size": [ 360, 130 ], "flags": {}, "order": 13, "mode": 0, "inputs": [ { "name": "samples", "type": "LATENT", "link": 9 }, { "name": "vae", "type": "VAE", "link": 10 } ], "outputs": [ { "name": "AUDIO", "type": "AUDIO", "links": [ 11 ], "slot_index": 0 } ], "properties": { "Node name for S&R": "VAEDecodeAudio" }, "widgets_values": [ false ] }, { "id": 14, "type": "SaveAudioAdvanced", "title": "Save AuK audio", "pos": [ 2530, 250 ], "size": [ 360, 180 ], "flags": {}, "order": 14, "mode": 0, "inputs": [ { "name": "audio", "type": "AUDIO", "link": 11 } ], "outputs": [ { "name": "audio", "type": "AUDIO", "links": [], "slot_index": 0 } ], "properties": { "Node name for S&R": "SaveAudioAdvanced" }, "widgets_values": [ "audio/auk_reference", "flac" ] } ], "links": [ [ 1, 2, 0, 6, 0, "CLIP" ], [ 2, 1, 0, 8, 0, "MODEL" ], [ 3, 6, 0, 8, 1, "CONDITIONING" ], [ 4, 9, 0, 12, 0, "NOISE" ], [ 5, 8, 0, 12, 1, "GUIDER" ], [ 6, 10, 0, 12, 2, "SAMPLER" ], [ 7, 11, 0, 12, 3, "SIGMAS" ], [ 8, 7, 0, 12, 4, "LATENT" ], [ 9, 12, 0, 13, 0, "LATENT" ], [ 10, 3, 0, 13, 1, "VAE" ], [ 11, 13, 0, 14, 0, "AUDIO" ], [ 12, 15, 0, 6, 2, "AUDIO" ], [ 13, 3, 0, 6, 3, "VAE" ], [ 14, 4, 0, 6, 1, "STRING" ], [ 15, 5, 0, 7, 0, "FLOAT" ] ], "groups": [ { "title": "1 \u00b7 Model Loading", "bounding": [ 0, 0, 420, 570 ], "color": "#35576b", "font_size": 24, "flags": {} }, { "title": "2 \u00b7 Inputs", "bounding": [ 500, 0, 420, 590 ], "color": "#4f7350", "font_size": 24, "flags": {} }, { "title": "3 \u00b7 Configuration", "bounding": [ 1000, 0, 420, 670 ], "color": "#82673d", "font_size": 24, "flags": {} }, { "title": "4 \u00b7 Conditioning", "bounding": [ 1500, 0, 420, 490 ], "color": "#685080", "font_size": 24, "flags": {} }, { "title": "5 \u00b7 Sampling", "bounding": [ 2000, 0, 420, 490 ], "color": "#496f79", "font_size": 24, "flags": {} }, { "title": "6 \u00b7 Output", "bounding": [ 2500, 0, 420, 460 ], "color": "#754b59", "font_size": 24, "flags": {} } ], "config": {}, "extra": { "ds": { "scale": 0.5, "offset": [ 50, 50 ] } }, "version": 0.4 }