Fun-ASR-Nano-2512-hf / config.json
langgz's picture
Add Fun-ASR-Nano in HuggingFace Transformers format (safetensors)
1b829ba verified
{
"model_type": "fun_asr_nano",
"architectures": [
"FunAsrNanoForConditionalGeneration"
],
"audio_encoder_config": {
"model_type": "fun_asr_nano_encoder",
"input_size": 560,
"output_size": 512,
"attention_heads": 4,
"linear_units": 2048,
"num_blocks": 50,
"tp_blocks": 20,
"dropout_rate": 0.1,
"attention_dropout_rate": 0.0,
"kernel_size": 11,
"sanm_shift": 0,
"initializer_range": 0.02
},
"adaptor_config": {
"model_type": "fun_asr_nano_adaptor",
"downsample_rate": 1,
"encoder_dim": 512,
"llm_dim": 1024,
"ffn_dim": 2048,
"num_layers": 2,
"attention_heads": 8,
"dropout_rate": 0.0,
"use_low_frame_rate": true
},
"text_config": {
"architectures": [
"Qwen3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 151643,
"eos_token_id": 151645,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 3072,
"max_position_embeddings": 40960,
"max_window_layers": 28,
"model_type": "qwen3",
"num_attention_heads": 16,
"num_hidden_layers": 28,
"num_key_value_heads": 8,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000,
"sliding_window": null,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.51.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151936
},
"ctc_config": {
"model_type": "fun_asr_nano_ctc",
"vocab_size": 60515,
"encoder_dim": 512,
"decoder_dim": 512,
"ffn_dim": 2048,
"num_layers": 5,
"downsample_rate": 1,
"blank_id": 60514,
"dropout_rate": 0.0
},
"audio_token_index": 151646,
"initializer_range": 0.02,
"torch_dtype": "bfloat16"
}