# sig: a0ac5076.yaml

weights_path: hf://kyutai/pocket-tts/languages/english_2026-04/model.safetensors@19f95fe2df36e79fbd9f10008595cc4c977a0fcc
weights_path_without_voice_cloning: hf://kyutai/pocket-tts-without-voice-cloning/languages/english_2026-04/model.safetensors@d29db7978e464fb90cb3359ee0c69a273b9142cc
# Human evals preferred this model at 0.3 (equal WER/similarity/speech rate).
default_temperature: 0.3


# Replace characters that are absent from the training data.
replace_characters:
  '’': "'"
  '‘': "'"

flow_lm:
  insert_bos_before_voice: true
  dtype: float32
  flow:
    depth: 6
    dim: 512
  transformer:
    d_model: 1024
    hidden_scale: 4
    max_period: 10000
    num_heads: 16
    num_layers: 6
  lookup_table:
    dim: 1024
    n_bins: 4000
    tokenizer: tokenizers
    tokenizer_path: hf://kyutai/pocket-tts-without-voice-cloning/languages/english_2026-04/tokenizer.json@00eac05ed3d16bdc3f6b5d598874019c34a89214
  #weights_path: final.safetensors

mimi:
  dtype: float32
  sample_rate: 24000
  inner_dim: 32
  outer_dim: 512
  channels: 1
  frame_rate: 12.5
  seanet:
    dimension: 512
    channels: 1
    n_filters: 64
    n_residual_layers: 1
    ratios:
    - 6
    - 5
    - 4
    kernel_size: 7
    residual_kernel_size: 3
    last_kernel_size: 3
    dilation_base: 2
    pad_mode: constant
    compress: 2
  transformer:
    d_model: 512
    num_heads: 8
    num_layers: 2
    layer_scale: 0.01
    context: 250
    dim_feedforward: 2048
    input_dimension: 512
    output_dimensions:
    - 512
  quantizer:
    dimension: 32
    output_dimension: 512
  #weights_path: codec.safetensors