Views
No views yet
tempo (AudioLDM2)1from src.steering import SteerableAudioLDMModel, AudioLDMCAASteeringController
2
3model = SteerableAudioLDMModel(device="cuda")
4ctrl = AudioLDMCAASteeringController.from_pretrained("lukasz-staniszewski/audioldm2-caa-tempo", alpha=1.0)
5
6with model.steer(ctrl):
7 out = model.generate(
8 prompt="instrumental music",
9 num_inference_steps=30, audio_length_in_s=10.0,
10 guidance_scale=3.5, seed=0,
11 )1{
2 "method": "standard_caa_audioldm",
3 "model": "cvssp/audioldm2-large",
4 "concept": "tempo",
5 "num_inference_steps": 100,
6 "audio_length_in_s": 10.0,
7 "guidance_scale": 4.5,
8 "seed": 10,
9 "device": "cuda",
10 "dtype": "float16",
11 "save_all_cfg_passes": true,
12 "layers_preset": "all",
13 "layers_to_steer": [
14 ".unet.down_blocks.1.attentions.1.transformer_blocks.0.attn2",
15 ".unet.down_blocks.1.attentions.1.transformer_blocks.1.attn2",
16 ".unet.down_blocks.1.attentions.2.transformer_blocks.0.attn2",
17 ".unet.down_blocks.1.attentions.2.transformer_blocks.1.attn2",
18 ".unet.down_blocks.1.attentions.5.transformer_blocks.0.attn2",
19 ".unet.down_blocks.1.attentions.5.transformer_blocks.1.attn2",
20 ".unet.down_blocks.1.attentions.6.transformer_blocks.0.attn2",
21 ".unet.down_blocks.1.attentions.6.transformer_blocks.1.attn2",
22 ".unet.down_blocks.2.attentions.1.transformer_blocks.0.attn2",
23 ".unet.down_blocks.2.attentions.1.transformer_blocks.1.attn2",
24 ".unet.down_blocks.2.attentions.2.transformer_blocks.0.attn2",
25 ".unet.down_blocks.2.attentions.2.transformer_blocks.1.attn2",
26 ".unet.down_blocks.2.attentions.5.transformer_blocks.0.attn2",
27 ".unet.down_blocks.2.attentions.5.transformer_blocks.1.attn2",
28 ".unet.down_blocks.2.attentions.6.transformer_blocks.0.attn2",
29 ".unet.down_blocks.2.attentions.6.transformer_blocks.1.attn2",
30 ".unet.down_blocks.3.attentions.1.transformer_blocks.0.attn2",
31 ".unet.down_blocks.3.attentions.1.transformer_blocks.1.attn2",
32 ".unet.down_blocks.3.attentions.2.transformer_blocks.0.attn2",
33 ".unet.down_blocks.3.attentions.2.transformer_blocks.1.attn2",
34 ".unet.down_blocks.3.attentions.5.transformer_blocks.0.attn2",
35 ".unet.down_blocks.3.attentions.5.transformer_blocks.1.attn2",
36 ".unet.down_blocks.3.attentions.6.transformer_blocks.0.attn2",
37 ".unet.down_blocks.3.attentions.6.transformer_blocks.1.attn2",
38 ".unet.mid_block.attentions.1.transformer_blocks.0.attn2",
39 ".unet.mid_block.attentions.1.transformer_blocks.1.attn2",
40 ".unet.mid_block.attentions.2.transformer_blocks.0.attn2",
41 ".unet.mid_block.attentions.2.transformer_blocks.1.attn2",
42 ".unet.up_blocks.0.attentions.1.transformer_blocks.0.attn2",
43 ".unet.up_blocks.0.attentions.1.transformer_blocks.1.attn2",
44 ".unet.up_blocks.0.attentions.2.transformer_blocks.0.attn2",
45 ".unet.up_blocks.0.attentions.2.transformer_blocks.1.attn2",
46 ".unet.up_blocks.0.attentions.5.transformer_blocks.0.attn2",
47 ".unet.up_blocks.0.attentions.5.transformer_blocks.1.attn2",
48 ".unet.up_blocks.0.attentions.6.transformer_blocks.0.attn2",
49 ".unet.up_blocks.0.attentions.6.transformer_blocks.1.attn2",
50 ".unet.up_blocks.0.attentions.9.transformer_blocks.0.attn2",
51 ".unet.up_blocks.0.attentions.9.transformer_blocks.1.attn2",
52 ".unet.up_blocks.0.attentions.10.transformer_blocks.0.attn2",
53 ".unet.up_blocks.0.attentions.10.transformer_blocks.1.attn2",
54 ".unet.up_blocks.1.attentions.1.transformer_blocks.0.attn2",
55 ".unet.up_blocks.1.attentions.1.transformer_blocks.1.attn2",
56 ".unet.up_blocks.1.attentions.2.transformer_blocks.0.attn2",
57 ".unet.up_blocks.1.attentions.2.transformer_blocks.1.attn2",
58 ".unet.up_blocks.1.attentions.5.transformer_blocks.0.attn2",
59 ".unet.up_blocks.1.attentions.5.transformer_blocks.1.attn2",
60 ".unet.up_blocks.1.attentions.6.transformer_blocks.0.attn2",
61 ".unet.up_blocks.1.attentions.6.transformer_blocks.1.attn2",
62 ".unet.up_blocks.1.attentions.9.transformer_blocks.0.attn2",
63 ".unet.up_blocks.1.attentions.9.transformer_blocks.1.attn2",
64 ".unet.up_blocks.1.attentions.10.transformer_blocks.0.attn2",
65 ".unet.up_blocks.1.attentions.10.transformer_blocks.1.attn2",
66 ".unet.up_blocks.2.attentions.1.transformer_blocks.0.attn2",
67 ".unet.up_blocks.2.attentions.1.transformer_blocks.1.attn2",
68 ".unet.up_blocks.2.attentions.2.transformer_blocks.0.attn2",
69 ".unet.up_blocks.2.attentions.2.transformer_blocks.1.attn2",
70 ".unet.up_blocks.2.attentions.5.transformer_blocks.0.attn2",
71 ".unet.up_blocks.2.attentions.5.transformer_blocks.1.attn2",
72 ".unet.up_blocks.2.attentions.6.transformer_blocks.0.attn2",
73 ".unet.up_blocks.2.attentions.6.transformer_blocks.1.attn2",
74 ".unet.up_blocks.2.attentions.9.transformer_blocks.0.attn2",
75 ".unet.up_blocks.2.attentions.9.transformer_blocks.1.attn2",
76 ".unet.up_blocks.2.attentions.10.transformer_blocks.0.attn2",
77 ".unet.up_blocks.2.attentions.10.transformer_blocks.1.attn2"
78 ],
79 "normalize_sv": true
80}