Modelpublic

rajevan123/STS-Lora-Fine-Tuning-Capstone-roberta-base-filtered-150-with-higher-r-mid

sourceHugging Facemitupdated 3y agoView on Hugging Face

0likes7downloads

Model Card

STS-Lora-Fine-Tuning-Capstone-roberta-base-filtered-150-with-higher-r-mid

This model is a fine-tuned version of FacebookAI/roberta-base on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.7443
—Accuracy: 0.6873

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: 3e-05
—trainbatchsize: 16
—evalbatchsize: 16
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	449	1.0408	0.4625
0.9593	2.0	898	1.0281	0.4981
0.9206	3.0	1347	0.9842	0.5225
0.8912	4.0	1796	0.9107	0.5693
0.8539	5.0	2245	0.8257	0.6273
0.8107	6.0	2694	0.8062	0.6685
0.779	7.0	3143	0.7672	0.6648
0.7797	8.0	3592	0.7709	0.6704
0.7649	9.0	4041	0.7509	0.6873
0.7649	10.0	4490	0.7376	0.6816
0.7527	11.0	4939	0.7360	0.6835
0.7526	12.0	5388	0.7493	0.6816
0.7476	13.0	5837	0.7421	0.6723
0.741	14.0	6286	0.7331	0.6929
0.7284	15.0	6735	0.7404	0.6854
0.7321	16.0	7184	0.7372	0.6798
0.7269	17.0	7633	0.7344	0.6816
0.7237	18.0	8082	0.7428	0.6723
0.7261	19.0	8531	0.7368	0.6854
0.7261	20.0	8980	0.7591	0.6704
0.715	21.0	9429	0.7434	0.6835
0.7088	22.0	9878	0.7504	0.6854
0.7228	23.0	10327	0.7500	0.6835
0.7127	24.0	10776	0.7583	0.6835
0.706	25.0	11225	0.7353	0.6948
0.7104	26.0	11674	0.7423	0.6891
0.7068	27.0	12123	0.7426	0.6910
0.7046	28.0	12572	0.7494	0.6873
0.7036	29.0	13021	0.7460	0.6910
0.7036	30.0	13470	0.7443	0.6873

Framework versions

—PEFT 0.10.0
—Transformers 4.38.2
—Pytorch 2.2.1+cu121
—Datasets 2.18.0
—Tokenizers 0.15.2