Intel/bert-large-uncased-cola-int8-inc

INT8 BERT large uncased finetuned CoLA

Post-training static quantization

PyTorch

This is an INT8 PyTorch model quantized with huggingface/optimum-intel through the usage of Intel® Neural Compressor.

The original fp32 model comes from the fine-tuned model yoshitomo-matsubara/bert-large-uncased-cola.

Test result

	INT8	FP32
Accuracy (eval-f1)	0.6336	0.6335
Model size (MB)	388	1340

Load with optimum:

from optimum.intel import INCModelForSequenceClassification

model_id = "Intel/bert-large-uncased-cola-int8"
int8_model = INCModelForSequenceClassification.from_pretrained(model_id)