lucataco/magnet ❓🔢📝 → 🖼️

▶️ 2.8K runs 📅 Jan 2024 ⚙️ Cog 0.8.6 🔗 GitHub 📄 Paper ⚖️ License

music-generation sound-effect-generation

About

MAGNeT: Masked Audio Generation using a Single Non-Autoregressive Transformer

Example Output

Prompt:

"80s electronic track with melodic synthesizers, catchy beat and groovy bass"

Output

Performance Metrics

14.64s Prediction Time

110.30s Total Time

All Input Parameters

{
  "model": "facebook/magnet-small-10secs",
  "top_p": 0.9,
  "prompt": "80s electronic track with melodic synthesizers, catchy beat and groovy bass",
  "max_cfg": 10,
  "min_cfg": 1,
  "span_score": "prod-stride1",
  "variations": 3,
  "temperature": 3,
  "decoding_steps_stage_1": 20,
  "decoding_steps_stage_2": 10,
  "decoding_steps_stage_3": 10,
  "decoding_steps_stage_4": 10
}

Input Parameters

model Default: facebook/magnet-small-10secs: Model to use
top_p Type: numberDefault: 0.9Range: 0 - 1: Top p for sampling
prompt Type: stringDefault: 80s electronic track with melodic synthesizers, catchy beat and groovy bass: Input Text
max_cfg Type: numberDefault: 10: Max CFG coefficient
min_cfg Type: numberDefault: 1: Min CFG coefficient
span_score Default: prod-stride1
variations Type: integerDefault: 3Range: 1 - 4: Number of variations to generate
temperature Type: numberDefault: 3: Temperature for sampling
decoding_steps_stage_1 Type: integerDefault: 20: Number of decoding steps for stage 1
decoding_steps_stage_2 Type: integerDefault: 10: Number of decoding steps for stage 2
decoding_steps_stage_3 Type: integerDefault: 10: Number of decoding steps for stage 3
decoding_steps_stage_4 Type: integerDefault: 10: Number of decoding steps for stage 4

Output Schema

Output

Type: array • Items Type: string • Items Format: uri

Example Execution Logs

/root/.pyenv/versions/3.10.13/lib/python3.10/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
spiece.model:   0%|          | 0.00/792k [00:00<?, ?B/s]
spiece.model: 100%|██████████| 792k/792k [00:00<00:00, 31.9MB/s]
tokenizer.json:   0%|          | 0.00/1.39M [00:00<?, ?B/s]
tokenizer.json: 100%|██████████| 1.39M/1.39M [00:00<00:00, 24.7MB/s]
config.json:   0%|          | 0.00/1.21k [00:00<?, ?B/s]
config.json: 100%|██████████| 1.21k/1.21k [00:00<00:00, 10.4MB/s]
model.safetensors:   0%|          | 0.00/892M [00:00<?, ?B/s]
model.safetensors:   4%|▎         | 31.5M/892M [00:00<00:05, 171MB/s]
model.safetensors:   7%|▋         | 62.9M/892M [00:00<00:04, 193MB/s]
model.safetensors:  11%|█         | 94.4M/892M [00:00<00:03, 208MB/s]
model.safetensors:  14%|█▍        | 126M/892M [00:00<00:03, 231MB/s] 
model.safetensors:  18%|█▊        | 157M/892M [00:00<00:03, 213MB/s]
model.safetensors:  22%|██▏       | 199M/892M [00:00<00:03, 227MB/s]
model.safetensors:  26%|██▌       | 231M/892M [00:01<00:02, 232MB/s]
model.safetensors:  29%|██▉       | 262M/892M [00:01<00:02, 234MB/s]
model.safetensors:  33%|███▎      | 294M/892M [00:01<00:02, 235MB/s]
model.safetensors:  36%|███▋      | 325M/892M [00:01<00:02, 239MB/s]
model.safetensors:  40%|███▉      | 357M/892M [00:01<00:02, 233MB/s]
model.safetensors:  44%|████▎     | 388M/892M [00:01<00:02, 233MB/s]
model.safetensors:  47%|████▋     | 419M/892M [00:01<00:02, 235MB/s]
model.safetensors:  51%|█████     | 451M/892M [00:01<00:01, 236MB/s]
model.safetensors:  54%|█████▍    | 482M/892M [00:02<00:01, 227MB/s]
model.safetensors:  58%|█████▊    | 514M/892M [00:02<00:01, 229MB/s]
model.safetensors:  61%|██████    | 545M/892M [00:02<00:01, 228MB/s]
model.safetensors:  65%|██████▍   | 577M/892M [00:02<00:01, 230MB/s]
model.safetensors:  68%|██████▊   | 608M/892M [00:02<00:01, 238MB/s]
model.safetensors:  72%|███████▏  | 640M/892M [00:02<00:01, 246MB/s]
model.safetensors:  75%|███████▌  | 671M/892M [00:02<00:00, 245MB/s]
model.safetensors:  79%|███████▉  | 703M/892M [00:03<00:00, 244MB/s]
model.safetensors:  82%|████████▏ | 734M/892M [00:03<00:00, 242MB/s]
model.safetensors:  86%|████████▌ | 765M/892M [00:03<00:00, 175MB/s]
model.safetensors:  88%|████████▊ | 786M/892M [00:03<00:00, 155MB/s]
model.safetensors:  92%|█████████▏| 818M/892M [00:03<00:00, 175MB/s]
model.safetensors:  95%|█████████▌| 849M/892M [00:03<00:00, 190MB/s]
model.safetensors:  99%|█████████▉| 881M/892M [00:04<00:00, 200MB/s]
model.safetensors: 100%|██████████| 892M/892M [00:04<00:00, 217MB/s]

Version Details

Version ID: e8e2ecd4a1dabb58924aa8300b668290cafae166dd36baf65dad9875877de50e
Version Created: January 17, 2024

Run on Replicate →