TeleChat2-7B-32K-adv / config.json
isbondarev's picture
Upload Telechat2ForCausalLM
1987180 verified
{
"apply_residual_connection_post_layernorm": false,
"architectures": [
"Telechat2ForCausalLM"
],
"attention_dropout": 0.0,
"attention_softmax_in_fp32": true,
"auto_map": {
"AutoConfig": "configuration_telechat2.Telechat2Config",
"AutoModel": "modeling_telechat2.Telechat2ForCausalLM",
"AutoModelForCausalLM": "modeling_telechat2.Telechat2ForCausalLM"
},
"base_seqlen": 32768,
"bias_dropout_fusion": true,
"bos_token_id": 1,
"dtype": "bfloat16",
"embed_layernorm": false,
"eos_token_id": 2,
"ffn_hidden_size": 12288,
"flash_attn": true,
"hidden_dropout": 0.0,
"hidden_size": 4096,
"initializer_range": 0.02,
"layer_norm_epsilon": 1e-05,
"logn": true,
"masked_softmax_fusion": true,
"max_position_embeddings": 32768,
"model_type": "telechat",
"n_head": 32,
"n_inner": null,
"n_layer": 30,
"num_key_value_heads": 32,
"offset_alibi": 100,
"pad_token_id": 3,
"pretraining_tp": 1,
"rope_scaling": {
"factor": 1.0,
"rope_type": "dynamic"
},
"rope_theta": 1000000,
"seq_length": 32768,
"skip_bias_add": true,
"skip_bias_add_qkv": false,
"slow_but_exact": false,
"tie_word_embeddings": false,
"training_seqlen": 32768,
"transformers_version": "4.57.1",
"unk_token_id": 0,
"use_cache": true,
"vocab_size": 131072
}