| { | |
| "architectures": [ | |
| "RishCoreForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 50256, | |
| "dtype": "float32", | |
| "eos_token_id": 50256, | |
| "hidden_act": "silu", | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 384, | |
| "max_position_embeddings": 1024, | |
| "model_type": "rish_ai", | |
| "norm_topk_prob": false, | |
| "num_attention_heads": 12, | |
| "num_experts": 16, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 12, | |
| "num_key_value_heads": 12, | |
| "output_router_logits": false, | |
| "pad_token_id": 50256, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 500000, | |
| "router_aux_loss_coef": 0.01, | |
| "tie_word_embeddings": false, | |
| "transformers_version": "4.57.3", | |
| "use_cache": true, | |
| "vocab_size": 50257 | |
| } | |