Files
niknah da9776ac5b Whisper example.
Use the built in preview as text node instead.
Cut sample audio down to 12secs
2026-05-11 22:02:47 +10:00

366 lines
6.2 KiB
JSON

{
"id": "790a8801-3396-429b-8047-a6a8dcddb760",
"revision": 0,
"last_node_id": 18,
"last_link_id": 12,
"nodes": [
{
"id": 13,
"type": "Apply Whisper",
"pos": [
610.0144856770834,
238.638427734375
],
"size": [
315,
146
],
"flags": {},
"order": 2,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 10
}
],
"outputs": [
{
"name": "text",
"type": "STRING",
"slot_index": 0,
"links": [
5,
8
]
},
{
"name": "segments_alignment",
"type": "whisper_alignment",
"links": null
},
{
"name": "words_alignment",
"type": "whisper_alignment",
"links": null
}
],
"properties": {
"cnr_id": "ComfyUI-Whisper",
"ver": "74aa5a217b1c4da906d7ee8bf5c1efd555ff257e",
"Node name for S&R": "Apply Whisper"
},
"widgets_values": [
"base",
"auto",
""
]
},
{
"id": 15,
"type": "PreviewAny",
"pos": [
754.5203993055555,
479.6483696831598
],
"size": [
240.70978867966846,
200.72181460785532
],
"flags": {},
"order": 5,
"mode": 0,
"inputs": [
{
"name": "source",
"type": "*",
"link": 8
}
],
"outputs": [
{
"name": "STRING",
"type": "STRING",
"links": null
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.20.1",
"Node name for S&R": "PreviewAny"
},
"widgets_values": [
null,
null,
null
]
},
{
"id": 18,
"type": "PreviewAudio",
"pos": [
401.9883611367486,
592.378366074352
],
"size": [
270,
88
],
"flags": {},
"order": 3,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 12
}
],
"outputs": [],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.20.1",
"Node name for S&R": "PreviewAudio"
},
"widgets_values": []
},
{
"id": 16,
"type": "TrimAudioDuration",
"pos": [
180.3947493270065,
398.90515191672586
],
"size": [
270,
82
],
"flags": {},
"order": 1,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 9
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
10,
11,
12
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.20.1",
"Node name for S&R": "TrimAudioDuration"
},
"widgets_values": [
0,
12
]
},
{
"id": 2,
"type": "PreviewAudio",
"pos": [
1066.7879326714412,
492.73924018012156
],
"size": [
315,
88
],
"flags": {},
"order": 6,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 4
}
],
"outputs": [],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.20.1",
"Node name for S&R": "PreviewAudio"
},
"widgets_values": []
},
{
"id": 12,
"type": "F5TTSAudioInputs",
"pos": [
1025,
155
],
"size": [
400,
256
],
"flags": {},
"order": 4,
"mode": 0,
"inputs": [
{
"name": "sample_audio",
"type": "AUDIO",
"link": 11
},
{
"name": "sample_text",
"type": "STRING",
"widget": {
"name": "sample_text"
},
"link": 5
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"slot_index": 0,
"links": [
4
]
}
],
"properties": {
"cnr_id": "comfyui-f5-tts",
"ver": "b23a1fe1cbe5a72cb7b8281619aa84227895adc3",
"Node name for S&R": "F5TTSAudioInputs"
},
"widgets_values": [
"Text of sample_audio",
"This is what I want to say",
1950,
"randomize",
"F5v1",
"auto",
1,
"F5TTS_Base"
]
},
{
"id": 4,
"type": "LoadAudio",
"pos": [
100.14046888369758,
162.98056233825855
],
"size": [
315,
136
],
"flags": {},
"order": 0,
"mode": 0,
"inputs": [],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"slot_index": 0,
"links": [
9
]
}
],
"properties": {
"cnr_id": "comfy-core",
"ver": "0.20.1",
"Node name for S&R": "LoadAudio"
},
"widgets_values": [
"F5TTS_test_en_1_ref_short.wav",
null,
null
]
}
],
"links": [
[
4,
12,
0,
2,
0,
"AUDIO"
],
[
5,
13,
0,
12,
1,
"STRING"
],
[
8,
13,
0,
15,
0,
"STRING"
],
[
9,
4,
0,
16,
0,
"AUDIO"
],
[
10,
16,
0,
13,
0,
"AUDIO"
],
[
11,
16,
0,
12,
0,
"AUDIO"
],
[
12,
16,
0,
18,
0,
"AUDIO"
]
],
"groups": [],
"config": {},
"extra": {
"ds": {
"scale": 1.157740863787376,
"offset": [
-30.787250223228813,
-20.539357326764627
]
},
"frontendVersion": "1.43.17",
"VHS_latentpreview": false,
"VHS_latentpreviewrate": 0,
"VHS_MetadataImage": true,
"VHS_KeepIntermediate": true
},
"version": 0.4
}