| { | |
| "apply_residual_connection_post_layernorm": false, | |
| "architectures": [ | |
| "TelechatForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_telechat.TelechatConfig", | |
| "AutoModelForCausalLM": "modeling_telechat.TelechatForCausalLM" | |
| }, | |
| "attention_dropout": 0.0, | |
| "attention_softmax_in_fp32": true, | |
| "bias_dropout_fusion": true, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "hidden_dropout": 0.0, | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "masked_softmax_fusion": true, | |
| "model_type": "telechat", | |
| "n_head": 64, | |
| "n_inner": null, | |
| "num_key_value_heads": 8, | |
| "n_layer": 96, | |
| "pad_token_id": 3, | |
| "pretraining_tp": 2, | |
| "skip_bias_add": false, | |
| "skip_bias_add_qkv": false, | |
| "slow_but_exact": false, | |
| "unk_token_id": 0, | |
| "use_cache": true, | |
| "vocab_size": 131072, | |
| "ffn_hidden_size": 40960, | |
| "flash_attn":true, | |
| "tie_word_embeddings":false, | |
| "training_seqlen":8192, | |
| "base_seqlen":8192, | |
| "seq_length": 8192 | |
| } | |