| { | |
| "architectures": [ | |
| "Telechat3ForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_telechat3.Telechat3Config", | |
| "AutoModel": "modeling_telechat3.Telechat3Model", | |
| "AutoModelForCausalLM": "modeling_telechat3.Telechat3ForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 6144, | |
| "initializer_range": 0.0048, | |
| "intermediate_size": 24576, | |
| "max_position_embeddings": 32768, | |
| "mlp_bias": false, | |
| "model_type": "telechat3", | |
| "num_attention_heads": 48, | |
| "num_hidden_layers": 64, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": 3, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "beta_fast": 32.0, | |
| "beta_slow": 1.0, | |
| "factor": 4.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "telechat3-yarn", | |
| "type": "telechat3-yarn" | |
| }, | |
| "rope_theta": 1000000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.53.2", | |
| "use_cache": true, | |
| "vocab_size": 131072 | |
| } | |