config.json

{
  "_name_or_path": "Llama-3.1-8B-Omni",
  "architectures": [
    "OmniSpeech2SLlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 128000,
  "ctc_add_pos": null,
  "ctc_decoder_config": "(2,4096,32,11008)",
  "ctc_lookahead_attn": false,
  "ctc_loss_weight": 1.0,
  "ctc_upsample_factor": 25,
  "eos_token_id": [
    128001,
    128008,
    128009
  ],
  "freeze_speech_projector": false,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "max_position_embeddings": 131072,
  "mlp_bias": false,
  "model_type": "omni_speech2s_llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "pretraining_tp": 1,
  "qformer_query_len": 64,
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "factor": 8.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "speech_encoder": "models/speech_encoder/large-v3.pt",
  "speech_encoder_ds_rate": 5,
  "speech_encoder_hidden_size": 1280,
  "speech_encoder_type": "whisper",
  "speech_generator_type": "ctc",
  "speech_normalize": false,
  "speech_projector_lr": null,
  "speech_projector_type": "linear",
  "tie_word_embeddings": false,
  "tokenizer_model_max_length": 2048,
  "tokenizer_padding_side": "right",
  "torch_dtype": "float16",
  "transformers_version": "4.43.4",
  "tune_speech_projector": false,
  "unit_vocab_size": 1000,
  "use_cache": true,
  "vocab_size": 128256
}