onnx-internal-testing/tiny-random-GraniteMoeHybridForCausalLM

1{2  "architectures": [3    "GraniteMoeHybridForCausalLM"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "attention_multiplier": 0.0078125,8  "bos_token_id": 100257,9  "dtype": "bfloat16",10  "embedding_multiplier": 12,11  "eos_token_id": 100257,12  "hidden_act": "silu",13  "hidden_size": 32,14  "initializer_range": 0.1,15  "intermediate_size": 128,16  "layer_types": [17    "mamba",18    "attention"19  ],20  "logits_scaling": 16,21  "mamba_chunk_size": 256,22  "mamba_conv_bias": true,23  "mamba_d_conv": 4,24  "mamba_d_head": 64,25  "mamba_d_state": 128,26  "mamba_expand": 256,27  "mamba_n_groups": 1,28  "mamba_n_heads": 128,29  "mamba_proj_bias": false,30  "max_position_embeddings": 131072,31  "model_type": "granitemoehybrid",32  "normalization_function": "rmsnorm",33  "num_attention_heads": 2,34  "num_experts_per_tok": 10,35  "num_hidden_layers": 2,36  "num_key_value_heads": 2,37  "num_local_experts": 72,38  "output_router_logits": false,39  "pad_token_id": 100256,40  "position_embedding_type": "nope",41  "residual_multiplier": 0.22,42  "rms_norm_eps": 1e-05,43  "rope_parameters": {44    "rope_theta": 10000,45    "rope_type": "default"46  },47  "router_aux_loss_coef": 0.0,48  "shared_intermediate_size": 128,49  "tie_word_embeddings": true,50  "time_step_limit": [51    0.0,52    {53      "__float__": "Infinity"54    }55  ],56  "time_step_max": 0.1,57  "time_step_min": 0.001,58  "transformers_version": "5.3.0.dev0",59  "use_cache": true,60  "vocab_size": 100352,61  "transformers.js_config": {62    "use_external_data_format": {63      "model.onnx": 1,64      "model_fp16.onnx": 165    },66    "kv_cache_dtype": {67      "q4f16": "float16",68      "fp16": "float16"69    }70  }71}