Modelpublic

thrunlab/t5-base_cola_moe_ex9_sp0_05_ar0_0_mare_mlp

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face

0likes8downloads

Model Card

t5-basecolamoeex9sp005ar00mare_mlp

This model is a fine-tuned version of t5-base on the glue dataset. It achieves the following results on the evaluation set:

—Loss: 3.9386
—Accuracy: 0.0

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: 5e-05
—trainbatchsize: 32
—evalbatchsize: 64
—seed: 1
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 20
—num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2328	0.09	25	1.1651	0.7383
0.764	0.19	50	0.7678	0.7287
0.6109	0.28	75	0.6739	0.7718
0.5633	0.37	100	0.5954	0.7661
0.5133	0.47	125	0.5870	0.7814
0.5224	0.56	150	0.5766	0.7785
0.4876	0.65	175	0.5574	0.7881
0.5157	0.75	200	0.5760	0.7881
0.4745	0.84	225	0.5327	0.7824
0.4612	0.93	250	0.5576	0.7900
0.4491	1.03	275	0.5174	0.7881
0.358	1.12	300	0.6065	0.7900
0.3363	1.21	325	0.6949	0.7919
0.4065	1.31	350	0.5112	0.7987
0.4044	1.4	375	0.5681	0.8063
0.3888	1.49	400	0.5422	0.7996
0.4992	1.59	425	0.5294	0.7900
0.4231	1.68	450	0.5221	0.8044
0.4912	1.77	475	0.4984	0.8130
0.4951	1.87	500	0.5109	0.8015
0.3117	1.96	525	0.5640	0.8044
0.3822	2.05	550	0.5524	0.8130
0.3886	2.15	575	0.6092	0.8121
0.305	2.24	600	0.5380	0.8111
0.4815	2.33	625	0.5478	0.8111
0.3298	2.43	650	0.5298	0.8150
0.3533	2.52	675	0.5043	0.8140
0.3706	2.61	700	0.5810	0.8178

Framework versions

—Transformers 4.33.2
—Pytorch 2.0.1+cu117
—Datasets 2.14.5
—Tokenizers 0.11.6