timesheet_estimator

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

• Loss: 0.5383
• Mse: 0.5383
• Rmse: 0.7337
• Mae: 0.5091
• R2: 0.4827
• Smape: 89.7730

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Mae	Mse	R2	Rmse	Smape
No log	0.46	300	0.7945	0.6404	0.7945	0.2018	0.8914	120.6558
No log	0.91	600	0.6674	0.5948	0.6674	0.3295	0.8169	109.9213
No log	1.37	900	0.6795	0.5998	0.6795	0.3174	0.8243	108.4883
0.7094	1.82	1200	0.6327	0.5916	0.6327	0.3643	0.7954	111.5772
0.7094	2.28	1500	0.6216	0.5712	0.6216	0.3755	0.7884	99.9257
0.7094	2.73	1800	0.5770	0.5397	0.5770	0.4203	0.7596	100.1235
0.5129	3.19	2100	0.5791	0.5391	0.5791	0.4182	0.7610	99.9525
0.5129	3.64	2400	0.5796	0.5421	0.5796	0.4177	0.7613	99.3905
0.5129	4.1	2700	0.5720	0.5354	0.5720	0.4254	0.7563	98.9299
0.4448	4.55	3000	0.5801	0.5381	0.5801	0.4173	0.7616	96.1430
0.4448	5.01	3300	0.5437	0.5185	0.5437	0.4775	0.7373	94.1203
0.4448	5.46	3600	0.5111	0.4949	0.5111	0.5088	0.7149	92.1147
0.4448	5.92	3900	0.5234	0.5106	0.5234	0.4970	0.7235	95.4636
0.4877	6.37	4200	0.5478	0.5249	0.5478	0.4735	0.7402	94.5022
0.4877	6.83	4500	0.5172	0.5172	0.7192	0.4998	0.5029	93.0563
0.4877	7.28	4800	0.5318	0.5318	0.7293	0.5083	0.4889	90.5273
0.3889	7.74	5100	0.5845	0.5845	0.7645	0.5377	0.4383	93.8608
0.3889	8.19	5400	0.5315	0.5315	0.7291	0.5014	0.4892	90.2302
0.3889	8.65	5700	0.5356	0.5356	0.7319	0.5010	0.4852	88.9946
0.324	9.1	6000	0.5345	0.5345	0.7311	0.5028	0.4864	89.7148
0.324	9.56	6300	0.5383	0.5383	0.7337	0.5091	0.4827	89.7730

Framework versions

• Transformers 4.27.0.dev0
• Pytorch 1.13.1
• Datasets 2.9.0
• Tokenizers 0.13.2