models/experiments/config.dac.yaml

sample_rate: 44_000
audio_backend: "dac"

models:
- name: "ar+nar-dac"
  size:
    audio_tokens: 1024
    text_tokens: 256
    dim: 1024
    heads: 16
    layers: 16
  resp_levels: 8
  prom_levels: 8
  tasks: 8
  langs: 2
  tones: 1
  arch_type: llama
  training: True
  version: 5
  attention: auto
  dropout: 0.1

  loss_factors:
    text: 0.01
    prom: 0.5
    resp: 1.0

  capabilities: ["ar", "nar"]

  experimental:
    hf: False
    interleave: False
    audio_embedding_sums: False
    rvq_level_range: []

hyperparameters:
  autotune: False
  autotune_params:
    start_profile_step: 1
    end_profile_step: 50
    num_tuning_micro_batch_sizes: 8

  batch_size: 16
  gradient_accumulation_steps: 2
  gradient_clipping: 1.0
  warmup_steps: 10

  optimizer: Prodigy
  learning_rate: 1.0
  torch_optimizer: True
  
  scheduler: "" # ScheduleFree
  torch_scheduler: True

evaluation:
  batch_size: 4
  frequency: 1000
  size: 4
  
  steps: 500
  ar_temperature: 1.0
  nar_temperature: 0.0
  load_disabled_engines: True

trainer:
  #no_logger: True
  ddp: False
  check_for_oom: False
  iterations: 1_000_000
  
  save_tag: step
  save_on_oom: True
  save_on_quit: True
  save_frequency: 500
  export_on_save: True

  keep_last_checkpoints: 4

  aggressive_optimizations: False
  load_disabled_engines: False
  gradient_checkpointing: True

  #load_state_dict: True
  strict_loading: False
  #load_tag: "9500"
  #load_states: False
  #restart_step_count: True
  
  gc_mode: None # "global_step"

  weight_dtype: bfloat16
  amp: False

  backend: deepspeed
  deepspeed:
    inferencing: False
    zero_optimization_level: 0
    use_compression_training: False

    amp: False

  load_webui: False

inference:
  backend: local
  normalize: False

  weight_dtype: bfloat16
  amp: False

optimizations:
  injects: False
  replace: True

  linear: False
  embedding: False
  optimizers: True

  bitsandbytes: False
  dadaptation: False
  bitnet: False
  fp8: False

dataset:
  speaker_name_getter: "lambda p: f'{p.parts[-3]}_{p.parts[-2]}'"
  speaker_group_getter: "lambda p: f'{p.parts[-3]}'"
  speaker_languages:
    ja: []

  use_hdf5: True
  use_metadata: True
  hdf5_flag: r
  validate: True

  workers: 6
  cache: True

  duration_range: [2.0, 60.0]

  random_utterance: 1.0
  max_prompts: 1
  prompt_duration_range: [3.0, 3.0]
  
  max_resps: 1
  p_resp_append: 0.25

  sample_type: path # path | speaker | group
  sample_order: duration
  sample_max_duration_batch: 0

  tasks_list: [ "tts" ] # , [ "tts", "tts-c", "ns", "sr", "tse", "cse", "nse", "tts"]

  training: []
  validation: []
  noise: []