Modelpublic

logiczmaksimka/training_results

sourceHugging Facemitupdated 3y agoView on Hugging Face

0likes8downloads

Model Card

training_results

This model is a fine-tuned version of ai-forever/ruElectra-medium on the None dataset. It achieves the following results on the evaluation set:

—Loss: 2.6856
—Accuracy: 0.7135
—Recall: 0.6688
—Precision: 0.7321
—F1: 0.6855

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: 16
—evalbatchsize: 16
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Recall	Precision	F1
No log	1.0	200	1.0503	0.6462	0.5404	0.5573	0.5309
No log	2.0	400	0.9312	0.6842	0.6358	0.6068	0.5981
0.9761	3.0	600	0.9141	0.7193	0.6410	0.6629	0.6447
0.9761	4.0	800	1.1036	0.7193	0.6453	0.6843	0.6516
0.3389	5.0	1000	1.3396	0.7135	0.6512	0.7203	0.6576
0.3389	6.0	1200	1.4660	0.7251	0.6688	0.7587	0.6759
0.3389	7.0	1400	1.4835	0.7135	0.6656	0.6910	0.6640
0.1627	8.0	1600	1.8635	0.7135	0.6535	0.7441	0.6673
0.1627	9.0	1800	1.5689	0.7368	0.7140	0.7412	0.7192
0.0893	10.0	2000	1.9628	0.7047	0.6885	0.7050	0.6842
0.0893	11.0	2200	1.9155	0.7339	0.6814	0.7328	0.6995
0.0893	12.0	2400	2.0020	0.7398	0.7086	0.7351	0.7064
0.0781	13.0	2600	2.0432	0.7193	0.7005	0.7265	0.6876
0.0781	14.0	2800	1.8877	0.7544	0.7385	0.7634	0.7415
0.0435	15.0	3000	2.2208	0.7281	0.6876	0.7271	0.6871
0.0435	16.0	3200	1.9514	0.7485	0.7071	0.7438	0.7169
0.0435	17.0	3400	2.0358	0.7368	0.7551	0.7406	0.7402
0.0405	18.0	3600	2.2364	0.7310	0.6250	0.6655	0.6307
0.0405	19.0	3800	2.3225	0.7164	0.6779	0.7234	0.6868
0.0511	20.0	4000	2.1369	0.7310	0.6826	0.7670	0.7089
0.0511	21.0	4200	2.2229	0.7427	0.6981	0.7783	0.7145
0.0511	22.0	4400	2.2711	0.7222	0.6650	0.7214	0.6671
0.0382	23.0	4600	2.4241	0.7222	0.6556	0.7826	0.6834
0.0382	24.0	4800	2.0575	0.7368	0.6767	0.7238	0.6804
0.0413	25.0	5000	2.5485	0.7076	0.6681	0.6842	0.6682
0.0413	26.0	5200	2.2235	0.7222	0.6474	0.6889	0.6536
0.0413	27.0	5400	2.5252	0.7105	0.6835	0.7028	0.6793
0.035	28.0	5600	2.5843	0.7164	0.6438	0.7341	0.6654
0.035	29.0	5800	2.6856	0.7135	0.6688	0.7321	0.6855

Framework versions

—Transformers 4.34.0
—Pytorch 2.1.0+cu121
—Datasets 2.14.5
—Tokenizers 0.14.1