| { | |
| "_name_or_path": "Qwen/Qwen2.5-Math-7B", | |
| "architectures": [ | |
| "LatentModelForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoModel": "latent_modeling.LatentModelForCausalLM", | |
| "AutoModelForCausalLM": "latent_modeling.LatentModelForCausalLM" | |
| }, | |
| "bos_token_id": 151643, | |
| "end_think_token_id": 151666, | |
| "eos_token_id": 151643, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 3584, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 18944, | |
| "latent_token_id": 151667, | |
| "max_position_embeddings": 4096, | |
| "max_window_layers": 28, | |
| "model_type": "qwen2", | |
| "num_attention_heads": 28, | |
| "num_hidden_layers": 28, | |
| "num_key_value_heads": 4, | |
| "pad_token_id": 151643, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 10000, | |
| "sliding_window": 4096, | |
| "start_think_token_id": 151665, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": false, | |
| "use_mrope": false, | |
| "use_sliding_window": false, | |
| "vocab_size": 151668 | |
| } | |