| { |
| "add_input_local_transformer": true, |
| "add_speech_sosp_eosp": false, |
| "architectures": [ |
| "MiMoV2ASRForCausalLM" |
| ], |
| "attention_bias": true, |
| "attention_dropout": 0.0, |
| "audio_channels": 8, |
| "delay_pattern": "0-1-2-3-4-5-6-7", |
| "dtype": "bfloat16", |
| "empty_loss_weight": 0.01, |
| "group_size": 4, |
| "head_dim": 128, |
| "hidden_act": "silu", |
| "hidden_size": 4096, |
| "initializer_range": 0.02, |
| "input_full_attention": true, |
| "input_local_dim": 1024, |
| "input_local_layers": 6, |
| "intermediate_size": 11008, |
| "layer_types": [ |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention" |
| ], |
| "local_attn_dropout": 0.1, |
| "local_attn_heads": 64, |
| "local_dim": 1024, |
| "local_ffn_dim": 4096, |
| "local_hidden_dropout": 0.1, |
| "local_layers": 16, |
| "local_rotary_base": 640000, |
| "max_position_embeddings": 8192, |
| "max_window_layers": 28, |
| "mlp_layers": 1, |
| "model_type": "qwen2", |
| "n_rvq": 20, |
| "no_speech_loss": false, |
| "no_text_loss": false, |
| "num_attention_heads": 32, |
| "num_hidden_layers": 36, |
| "num_key_value_heads": 8, |
| "rms_norm_eps": 1e-06, |
| "rope_scaling": null, |
| "rope_theta": 640000, |
| "sliding_window": null, |
| "speech_vocab_size": "1025-1025-129-129-129-129-129-129", |
| "speech_zeroemb_idx": "1024-1024-128-128-128-128-128-128", |
| "tie_word_embeddings": false, |
| "transformers_version": "4.57.1", |
| "use_cache": true, |
| "use_sliding_window": false, |
| "vocab_size": 151680, |
| "audio_config": { |
| "tokenizer_version": "v1", |
| "speech_vocab_size": "1025-1025-129-129-129-129-129-129", |
| "speech_zeroemb_idx": "1024-1024-128-128-128-128-128-128", |
| "group_size": 4, |
| "audio_channels": 8, |
| "input_local_layers": 6, |
| "input_local_dim": 1024, |
| "input_full_attention": true, |
| "input_local_attn_heads": 64, |
| "input_local_head_dim": 16, |
| "input_local_intermediate_size": 4096, |
| "input_local_hidden_dropout": 0.1, |
| "out_hidden_size": 4096, |
| "rope_theta": 640000, |
| "partial_rotary_factor": 1.0, |
| "projection_layers": 1, |
| "add_post_norm": true, |
| "audio_segment_size": 6000 |
| }, |
| "quantization": { |
| "group_size": 64, |
| "bits": 4, |
| "mode": "affine" |
| }, |
| "quantization_config": { |
| "group_size": 64, |
| "bits": 4, |
| "mode": "affine" |
| } |
| } |
|
|
