Modelpublic

gokuls/model_v1_complete_training_wt_init_48_small_emb_comp

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes8downloads

Model Card

modelv1completetrainingwtinit48smallemb_comp

This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset. It achieves the following results on the evaluation set:

—Loss: 5.3637
—Accuracy: 0.2281

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: 1e-05
—trainbatchsize: 48
—evalbatchsize: 48
—seed: 10
—distributed_type: multi-GPU
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 10000
—num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
6.5703	0.25	30000	6.5612	0.1162
6.3542	0.49	60000	6.3484	0.1335
6.2455	0.74	90000	6.2454	0.1406
6.1931	0.98	120000	6.1884	0.1438
6.1452	1.23	150000	6.1441	0.1460
6.1194	1.47	180000	6.1122	0.1481
6.0939	1.72	210000	6.0831	0.1488
6.0463	1.97	240000	6.0348	0.1504
6.0078	2.21	270000	6.0020	0.1516
5.9732	2.46	300000	5.9705	0.1531
5.9408	2.7	330000	5.9340	0.1542
5.9028	2.95	360000	5.8904	0.1567
5.8675	3.2	390000	5.8527	0.1597
5.8365	3.44	420000	5.8116	0.1655
5.7944	3.69	450000	5.7580	0.1753
5.7198	3.93	480000	5.6695	0.1896
5.6264	4.18	510000	5.5688	0.2028
5.5337	4.42	540000	5.4742	0.2137
5.4748	4.67	570000	5.4033	0.2228
5.4301	4.92	600000	5.3637	0.2281

Framework versions

—Transformers 4.30.2
—Pytorch 1.14.0a0+410ce96
—Datasets 2.13.1
—Tokenizers 0.13.3