config.json

{
  "card": 2048,
  "n_q": 32,
  "dep_q": 0,
  "delays": [
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0,
    0
  ],
  "dim": 2048,
  "text_card": 4000,
  "existing_text_padding_id": 3,
  "num_heads": 32,
  "num_layers": 48,
  "hidden_scale": 4.125,
  "causal": true,
  "layer_scale": null,
  "context": 375,
  "max_period": 100000.0,
  "gating": "silu",
  "norm": "rms_norm_f32",
  "positional_embedding": "rope",
  "depformer_dim": 1024,
  "depformer_num_heads": 16,
  "depformer_num_layers": 6,
  "depformer_dim_feedforward": null,
  "depformer_multi_linear": true,
  "depformer_pos_emb": "none",
  "depformer_weights_per_step": true,
  "conditioners": {},
  "cross_attention": false,
  "model_id": {
    "sig": "dabcc802",
    "epoch": 50
  },
  "lm_gen_config": {
    "temp": 0.0,
    "temp_text": 0.0,
    "top_k": 250,
    "top_k_text": 50
  },
  "stt_config": {
    "audio_delay_seconds": 2.5,
    "audio_silence_prefix_seconds": 1.0
  },
  "model_type": "stt",
  "mimi_name": "[email protected]",
  "tokenizer_name": "tokenizer_en_audio_4000.model",
  "architectures": [
    "KyutaiSpeechToTextForConditionalGeneration"
  ],
  "attention_dropout": 0.0,
  "audio_bos_token_id": 2048,
  "audio_pad_token_id": 69569,
  "bos_token_id": 48000,
  "codebook_vocab_size": 2049,
  "codec_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "audio_channels": 1,
    "codebook_dim": 256,
    "codebook_size": 2048,
    "compress": 2,
    "dilation_growth_rate": 2,
    "frame_rate": 12.5,
    "head_dim": 64,
    "hidden_act": "gelu",
    "hidden_size": 512,
    "initializer_range": 0.02,
    "intermediate_size": 2048,
    "kernel_size": 7,
    "last_kernel_size": 3,
    "layer_scale_initial_scale": 0.01,
    "max_position_embeddings": 8000,
    "model_type": "mimi",
    "norm_eps": 1e-05,
    "num_attention_heads": 8,
    "num_filters": 64,
    "num_hidden_layers": 8,
    "num_key_value_heads": 8,
    "num_quantizers": 32,
    "num_residual_layers": 1,
    "num_semantic_quantizers": 1,
    "pad_mode": "constant",
    "residual_kernel_size": 3,
    "rope_theta": 10000.0,
    "sampling_rate": 24000,
    "sliding_window": 250,
    "trim_right_ratio": 1.0,
    "upsample_groups": 512,
    "upsampling_ratios": [
      8,
      6,
      5,
      4
    ],
    "use_cache": false,
    "use_causal_conv": true,
    "use_conv_shortcut": false,
    "use_streaming": false,
    "vector_quantization_hidden_dimension": 256
  },
  "ffn_dim": 11264,
  "frame_size": 1920,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "max_position_embeddings": 750,
  "num_attention_heads": 32,
  "num_codebooks": 32,
  "num_hidden_layers": 48,
  "num_key_value_heads": 32,
  "pad_token_id": 3,
  "rms_norm_eps": 1e-08,
  "rope_theta": 100000.0,
  "sliding_window": 375,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.53.0.dev0",
  "use_cache": true,
  "vocab_size": 4001,
  "transformers_weights": "transformers.safetensors.index.json"
}