Modelpublic

optimum/MiniLMv2-L12-H384-distilled-finetuned-clinc

sourceHugging Faceupdated 4y agoView on Hugging Face

0likes19downloads

Model Card

MiniLMv2-L12-H384-distilled-from-RoBERTa-Large-distilled-clinc

This model is a fine-tuned version of nreimers/MiniLMv2-L12-H384-distilled-from-RoBERTa-Large on the clinc_oos dataset. It achieves the following results on the evaluation set:

—Loss: 0.3479
—Accuracy: 0.94

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 0.0001
—trainbatchsize: 256
—evalbatchsize: 256
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	60	0.8171	0.2490
No log	2.0	120	0.7039	0.6568
No log	3.0	180	0.6067	0.7932
0.7269	4.0	240	0.5270	0.8674
0.7269	5.0	300	0.4659	0.9010
0.7269	6.0	360	0.4201	0.9194
0.7269	7.0	420	0.3867	0.9352
0.4426	8.0	480	0.3649	0.9352
0.4426	9.0	540	0.3520	0.9403
0.4426	10.0	600	0.3479	0.94

Framework versions

—Transformers 4.11.3
—Pytorch 1.11.0
—Datasets 1.16.1
—Tokenizers 0.10.3