| { | |
| "architectures": [ | |
| "SpeechTextAny2Any" | |
| ], | |
| "bos_token_id": 1, | |
| "encoder_config": { | |
| "attn_dropout": 0.0, | |
| "attn_flash": true, | |
| "conv_dropout": 0.0, | |
| "conv_expansion_factor": 2, | |
| "conv_kernel_size": 5, | |
| "depth": 8, | |
| "dim_head": 32, | |
| "ff_dropout": 0.0, | |
| "ff_mult": 4, | |
| "heads": 16 | |
| }, | |
| "eos_token_id": 2, | |
| "hidden_size": 512, | |
| "label_smoothing": 0.0, | |
| "model_type": "speech_text_any2any", | |
| "num_speech_tokens": 1024, | |
| "num_text_tokens": 256, | |
| "pad_token_id": 0, | |
| "special_tokens": { | |
| "[MASK]": 3, | |
| "[PAD]": 0, | |
| "[SEP]": 4, | |
| "[SPEECH]": 2, | |
| "[TEXT]": 1 | |
| }, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.42.0.dev0" | |
| } | |