Automatic Speech Recognition
Transformers
Safetensors
fun_asr_nano
text-generation
speech-recognition
asr
end-to-end
multilingual
streaming
Instructions to use FunAudioLLM/Fun-ASR-Nano-2512-hf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use FunAudioLLM/Fun-ASR-Nano-2512-hf with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="FunAudioLLM/Fun-ASR-Nano-2512-hf")# Load model directly from transformers import AutoModelForSeq2SeqLM model = AutoModelForSeq2SeqLM.from_pretrained("FunAudioLLM/Fun-ASR-Nano-2512-hf", dtype="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "activation_function": "relu", | |
| "adaptor_intermediate_size": 2048, | |
| "adaptor_num_attention_heads": 8, | |
| "adaptor_num_hidden_layers": 2, | |
| "architectures": [ | |
| "FunAsrNanoForConditionalGeneration" | |
| ], | |
| "audio_token_id": 151646, | |
| "dtype": "bfloat16", | |
| "encoder_config": { | |
| "activation_dropout": 0.1, | |
| "activation_function": "relu", | |
| "attention_dropout": 0.1, | |
| "d_model": 512, | |
| "dropout": 0.1, | |
| "encoder_attention_heads": 4, | |
| "encoder_ffn_dim": 2048, | |
| "encoder_layers": 50, | |
| "kernel_size": 11, | |
| "model_type": "fun_asr_nano_encoder", | |
| "num_mel_bins": 80, | |
| "num_stacked_frames": 7, | |
| "num_timestamp_prediction_blocks": 20 | |
| }, | |
| "initializer_range": 0.02, | |
| "model_type": "fun_asr_nano", | |
| "text_config": { | |
| "architectures": [ | |
| "Qwen3ForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 151643, | |
| "dtype": "bfloat16", | |
| "eos_token_id": 151645, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 1024, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 3072, | |
| "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" | |
| ], | |
| "max_position_embeddings": 40960, | |
| "max_window_layers": 28, | |
| "model_type": "qwen3", | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 28, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": null, | |
| "rms_norm_eps": 1e-06, | |
| "rope_parameters": { | |
| "rope_theta": 1000000, | |
| "rope_type": "default" | |
| }, | |
| "sliding_window": null, | |
| "tie_word_embeddings": true, | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 151936 | |
| }, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "5.15.0.dev0" | |
| } | |