simpletuner/minimax-music3

Public

5 runs

Run simpletuner/minimax-music3 with an API

Use one of our client libraries to get started quickly. Clicking on a library will take you to the Playground tab where you can tweak different inputs, see the results, and copy the corresponding code to use in your own project.

Input schema

The fields you can use to run this model with an API. If you don't give a value for a field its default value will be used.

Field	Type	Default value	Description
train_data	string		ZIP or tar containing stem-matched audio, .txt captions, and .lyrics files
regularisation_data	string		Optional unrelated-song archive used for frozen-base prior preservation
trigger_word	string		Optional keyword used as the prompt for every training track
max_train_steps	integer	750 Min: 1 Max: 5000	Optimizer steps
checkpoint_interval	integer	250 Min: 10 Max: 1000	Save every N steps
lora_rank	None	64	LoRA rank
learning_rate	number	0.00008 Min: 0.000001 Max: 0.0005	Constant learning rate
max_audio_duration	integer	180 Min: 10 Max: 300	Maximum token sequence duration per track in seconds
seed	integer	42	Training seed

{
  "type": "object",
  "title": "Input",
  "required": [
    "train_data"
  ],
  "properties": {
    "seed": {
      "type": "integer",
      "title": "Seed",
      "default": 42,
      "minimum": 0,
      "x-order": 8,
      "description": "Training seed"
    },
    "lora_rank": {
      "enum": [
        16,
        32,
        64,
        128
      ],
      "type": "integer",
      "title": "lora_rank",
      "description": "LoRA rank",
      "default": 64,
      "x-order": 5
    },
    "train_data": {
      "type": "string",
      "title": "Train Data",
      "format": "uri",
      "x-order": 0,
      "description": "ZIP or tar containing stem-matched audio, .txt captions, and .lyrics files"
    },
    "trigger_word": {
      "type": "string",
      "title": "Trigger Word",
      "default": "",
      "x-order": 2,
      "description": "Optional keyword used as the prompt for every training track"
    },
    "learning_rate": {
      "type": "number",
      "title": "Learning Rate",
      "default": 8e-05,
      "maximum": 0.0005,
      "minimum": 1e-06,
      "x-order": 6,
      "description": "Constant learning rate"
    },
    "max_train_steps": {
      "type": "integer",
      "title": "Max Train Steps",
      "default": 750,
      "maximum": 5000,
      "minimum": 1,
      "x-order": 3,
      "description": "Optimizer steps"
    },
    "max_audio_duration": {
      "type": "integer",
      "title": "Max Audio Duration",
      "default": 180,
      "maximum": 300,
      "minimum": 10,
      "x-order": 7,
      "description": "Maximum token sequence duration per track in seconds"
    },
    "checkpoint_interval": {
      "type": "integer",
      "title": "Checkpoint Interval",
      "default": 250,
      "maximum": 1000,
      "minimum": 10,
      "x-order": 4,
      "description": "Save every N steps"
    },
    "regularisation_data": {
      "type": "string",
      "title": "Regularisation Data",
      "format": "uri",
      "x-order": 1,
      "nullable": true,
      "description": "Optional unrelated-song archive used for frozen-base prior preservation"
    }
  }
}

Output schema

The shape of the response you’ll get when you run this model with an API.

Schema

{
  "type": "string",
  "title": "Output",
  "format": "uri"
}