Text Generation
Transformers
bloom
Eval Results
bloomz-176B-GPTQ / config.json
TheBloke's picture
Upload GPTQ split into three parts to avoid HF upload limit
b904a1c
raw
history blame
666 Bytes
{
"_name_or_path": "/home/ubuntu/models/bloomz/",
"apply_residual_connection_post_layernorm": false,
"architectures": [
"BloomForCausalLM"
],
"attention_dropout": 0.0,
"attention_softmax_in_fp32": true,
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_dropout": 0.0,
"hidden_size": 14336,
"initializer_range": 0.02,
"layer_norm_epsilon": 1e-05,
"masked_softmax_fusion": true,
"model_type": "bloom",
"n_head": 112,
"n_layer": 70,
"pad_token_id": 3,
"pretraining_tp": 4,
"seq_length": 2048,
"slow_but_exact": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.28.1",
"use_cache": true,
"vocab_size": 250880
}