job-listing-filtering-model

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

• Loss: 0.1992

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss
0.4639	1.55	50	0.4343
0.407	3.12	100	0.3589
0.3459	4.68	150	0.3110
0.2871	6.25	200	0.2604
0.1966	7.8	250	0.2004
0.0994	9.37	300	0.1766
0.0961	10.92	350	0.2007
0.0954	12.49	400	0.1716
0.0498	14.06	450	0.1642
0.0419	15.62	500	0.1811
0.0232	17.18	550	0.1872
0.0146	18.74	600	0.1789
0.0356	20.31	650	0.1984
0.0325	21.86	700	0.1845
0.0381	23.43	750	0.1994
0.0063	24.98	800	0.1992

Framework versions

• Transformers 4.17.0
• Pytorch 1.11.0+cu113
• Datasets 2.0.0
• Tokenizers 0.11.6