AEmotionStudio commited on
Commit
b68ff8b
·
verified ·
1 Parent(s): f45f86a

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +81 -0
README.md ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: stability-ai-community
4
+ license_link: LICENSE
5
+ base_model:
6
+ - stabilityai/stable-audio-open-1.0
7
+ tags:
8
+ - music
9
+ - audio
10
+ - text-to-audio
11
+ - music-generation
12
+ - loop
13
+ - sample
14
+ - stable-audio
15
+ - safetensors
16
+ - comfyui
17
+ pipeline_tag: text-to-audio
18
+ ---
19
+
20
+ # Foundation-1 — Mirror
21
+
22
+ **BPM/Key-Aware Music Sample Generator**
23
+
24
+ [Original Model](https://huggingface.co/RoyalCities/Foundation-1) by [RoyalCities](https://huggingface.co/RoyalCities) · Fine-tuned on [stable-audio-open-1.0](https://huggingface.co/stabilityai/stable-audio-open-1.0)
25
+
26
+ > This is an **ungated mirror** of the Foundation-1 model weights for use with [ComfyUI-FFMPEGA](https://github.com/AEmotionStudio/ComfyUI-FFMPEGA). All credits go to the original authors.
27
+
28
+ ## What's in This Repo
29
+
30
+ | File | Description | Size |
31
+ |------|-------------|------|
32
+ | `Foundation_1.safetensors` | FP16 model checkpoint | ~3 GB |
33
+ | `model_config.json` | Model architecture config | ~1 KB |
34
+
35
+ ## What Foundation-1 Does
36
+
37
+ Foundation-1 generates production-ready musical loops with fine-grained control over:
38
+
39
+ - **Musical structure**: BPM, bars, time signatures, key/mode
40
+ - **Instrument identity**: 30+ instrument families (synth, strings, brass, guitar, etc.)
41
+ - **Timbral control**: 100+ timbre tags (warm, bright, gritty, glassy, etc.)
42
+ - **FX prompting**: Reverb, delay, chorus, distortion, and more
43
+ - **Loop fidelity**: Seamless, tempo-synced loops designed for layering
44
+
45
+ ## Usage with ComfyUI-FFMPEGA
46
+
47
+ These weights are auto-downloaded by the **FFMPEGA Agent** node:
48
+
49
+ 1. Set `llm_model` to `none`
50
+ 2. Set `no_llm_mode` to `generate_sample`
51
+ 3. Enter a prompt like: *"Synth, Pad, Warm, Wide, Lush, 120 BPM, 4 Bars, C major"*
52
+
53
+ The model (~3 GB) downloads on first use to `ComfyUI/models/foundation1/`.
54
+
55
+ ## Hardware Requirements
56
+
57
+ - **VRAM**: ~7 GB during generation
58
+ - **Generation speed**: ~7–8 seconds per sample (RTX 3090)
59
+
60
+ ## Prompt Structure
61
+
62
+ Foundation-1 uses a layered prompt system:
63
+
64
+ ```
65
+ [Instrument Family], [Sub-Family], [Timbre Tags], [FX Tags], [BPM], [Bars], [Key]
66
+ ```
67
+
68
+ Example: `"Synth, Lead, Bright, Sharp, Saw, Detune, Delay Ping Pong, 140 BPM, 8 Bars, A minor"`
69
+
70
+ ## License
71
+
72
+ These model weights are released under the [Stability AI Community License](./LICENSE).
73
+
74
+ - ✅ Free for non-commercial use
75
+ - ✅ Free for commercial use by entities with annual revenue < $1M USD
76
+ - ❌ Entities with annual revenue ≥ $1M need an Enterprise license from Stability AI
77
+
78
+ ## Acknowledgements
79
+
80
+ - [Foundation-1](https://huggingface.co/RoyalCities/Foundation-1) by RoyalCities
81
+ - [Stable Audio Open](https://huggingface.co/stabilityai/stable-audio-open-1.0) by Stability AI