| { | |
| "_name_or_path": "./VDrontMoE-2m-5e", | |
| "architectures": [ | |
| "VDrontMoEModel" | |
| ], | |
| "attention_probs_dropout_prob": 0.0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "expert_hidden_size": 1300, | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 256, | |
| "initializer_range": 0.02, | |
| "max_position_embeddings": 2048, | |
| "model_type": "VDrontMoE", | |
| "num_attention_heads": 8, | |
| "num_experts": 5, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 2, | |
| "num_key_value_heads": 8, | |
| "num_shared_experts": 0, | |
| "pad_token_id": 0, | |
| "rms_norm_eps": 1e-06, | |
| "router_aux_loss_coef": 0.01, | |
| "router_z_loss_coef": 0.001, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.46.3", | |
| "vocab_size": 50257 | |
| } | |