hf-tiny-v2/tiny-random-GraniteMoeSWAModel
013
1{2 "architectures": [3 "GraniteMoeSWAModel"4 ],5 "attention_bias": false,6 "attention_dropout": 0.0,7 "attention_multiplier": 1.0,8 "bos_token_id": 100257,9 "dtype": "float32",10 "embedding_multiplier": 1.0,11 "eos_token_id": 100257,12 "hidden_act": "gelu",13 "hidden_size": 32,14 "initializer_range": 0.02,15 "intermediate_size": 32,16 "layer_rope_theta": [17 10000.0,18 10000.019 ],20 "layer_types": [21 "full_attention",22 "sliding_attention"23 ],24 "logits_scaling": 1.0,25 "max_position_embeddings": 512,26 "model_type": "granitemoe_swa",27 "num_attention_heads": 2,28 "num_experts_per_tok": 2,29 "num_hidden_layers": 2,30 "num_key_value_heads": 2,31 "num_local_experts": 8,32 "output_router_logits": false,33 "pad_token_id": 100256,34 "residual_multiplier": 1.0,35 "rms_norm_eps": 1e-06,36 "rope_parameters": {37 "rope_theta": 10000.0,38 "rope_type": "default"39 },40 "router_aux_loss_coef": 0.001,41 "shared_intermediate_size": 0,42 "sliding_window": 128,43 "tie_word_embeddings": false,44 "transformers_version": "5.16.0.dev0",45 "use_cache": true,46 "vocab_size": 10035247}48 