config.json

{
  "card": 2048,
  "n_q": 32,
  "dep_q": 32,
  "delays": [
    0,
    0,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2,
    2
  ],
  "dim": 2048,
  "text_card": 8000,
  "existing_text_padding_id": 3,
  "num_heads": 16,
  "num_layers": 16,
  "hidden_scale": 4.125,
  "causal": true,
  "layer_scale": null,
  "context": 500,
  "max_period": 10000,
  "gating": "silu",
  "norm": "rms_norm_f32",
  "positional_embedding": "rope",
  "depformer_dim": 1024,
  "depformer_num_heads": 16,
  "depformer_num_layers": 4,
  "depformer_dim_feedforward": 3072,
  "depformer_multi_linear": true,
  "depformer_pos_emb": "none",
  "depformer_weights_per_step": true,
  "depformer_low_rank_embeddings": 128,
  "demux_second_stream": true,
  "text_card_out": null,
  "conditioners": {
    "speaker_wavs": {
      "type": "tensor",
      "tensor": {
        "dim": 512
      }
    },
    "cfg": {
      "type": "lut",
      "lut": {
        "n_bins": 7,
        "dim": 16,
        "tokenizer": "noop",
        "possible_values": [
          "1.0",
          "1.5",
          "2.0",
          "2.5",
          "3.0",
          "3.5",
          "4.0"
        ]
      }
    },
    "control": {
      "type": "lut",
      "lut": {
        "dim": 2048,
        "n_bins": 1,
        "tokenizer": "noop",
        "possible_values": [
          "ok"
        ]
      }
    }
  },
  "fuser": {
    "cross_attention_pos_emb": true,
    "cross_attention_pos_emb_scale": 1,
    "sum": [
      "control",
      "cfg"
    ],
    "prepend": [],
    "cross": [
      "speaker_wavs"
    ]
  },
  "cross_attention": true,
  "tts_config": {
    "audio_delay": 1.28,
    "second_stream_ahead": 2
  },
  "model_id": {
    "sig": "1e68beda",
    "epoch": 240
  },
  "depformer_weights_per_step_schedule": [
    0,
    1,
    2,
    3,
    4,
    5,
    6,
    7,
    8,
    8,
    8,
    8,
    8,
    8,
    8,
    8,
    9,
    9,
    9,
    9,
    9,
    9,
    9,
    9,
    10,
    10,
    10,
    10,
    10,
    10,
    10,
    10
  ],
  "model_type": "tts",
  "lm_gen_config": {
    "temp": 0.6,
    "text_temp": 0.6
  },
  "tokenizer_name": "tokenizer_spm_8k_en_fr_audio.model",
  "mimi_name": "tokenizer-e351c8d8-checkpoint125.safetensors",
  "moshi_name": "dsm_tts_1e68beda@240.safetensors"
}