Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
TheBloke
/
phi-2-GPTQ
like
29
Text Generation
Transformers
Safetensors
English
phi-msft
nlp
code
custom_code
4-bit precision
gptq
License:
microsoft-research-license (other)
Model card
Files
Files and versions
Community
6
Train
Use this model
gptq-4bit-32g-actorder_True
phi-2-GPTQ
/
quantize_config.json
TheBloke
Fix quantize_config.json
fdc0568
6 months ago
raw
history
blame
133 Bytes
{
"bits"
:
4
,
"group_size"
:
32
,
"damp_percent"
:
0.1
,
"desc_act"
:
true
,
"sym"
:
true
,
"true_sequential"
:
true
}