Text Generation
Transformers
PyTorch
skywork
custom_code
LoneStriker's picture
ExLLaMA V2 quant of Skywork-13B-Airo-Claude-Pippa-Puffin-4.0bpw-h6-exl2
a131d93
{
"_name_or_path": "/models/models/Skywork-13B-base",
"architectures": [
"SkyworkForCausalLM"
],
"auto_map": {
"AutoConfig": "configuration_skywork.SkyworkConfig",
"AutoModelForCausalLM": "modeling_skywork.SkyworkForCausalLM"
},
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 4608,
"initializer_range": 0.01,
"intermediate_size": 12288,
"max_position_embeddings": 131072,
"model_type": "skywork",
"num_attention_heads": 36,
"num_hidden_layers": 52,
"num_key_value_heads": 36,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 10000.0,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.34.1",
"use_cache": false,
"vocab_size": 65519
}