{ "architectures": [ "MoonshineModel" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 98, "decoder_hidden_act": "silu", "decoder_num_attention_heads": 2, "decoder_num_hidden_layers": 2, "decoder_num_key_value_heads": 2, "decoder_start_token_id": 85, "dtype": "float32", "encoder_hidden_act": "gelu", "encoder_num_attention_heads": 2, "encoder_num_hidden_layers": 2, "encoder_num_key_value_heads": 2, "eos_token_id": 98, "hidden_size": 8, "initializer_range": 0.02, "intermediate_size": 32, "is_encoder_decoder": true, "max_position_embeddings": 512, "model_type": "moonshine", "pad_head_dim_to_multiple_of": null, "pad_token_id": null, "partial_rotary_factor": 0.9, "rope_parameters": { "partial_rotary_factor": 0.9, "rope_theta": 10000.0, "rope_type": "default" }, "tie_word_embeddings": true, "transformers_version": "5.16.0.dev0", "use_cache": true, "vocab_size": 32768 }