xlm-roberta-base_lr5e-06_seed42_basic_original_amh-hau-eng_train

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

• Loss: 0.0289
• Spearman Corr: 0.7801

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-06
• train_batch_size: 32
• eval_batch_size: 128
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Spearman Corr
No log	1.55	200	0.0251	0.7275
0.0824	3.1	400	0.0250	0.7584
0.0309	4.65	600	0.0251	0.7713
0.0251	6.2	800	0.0254	0.7824
0.0251	7.75	1000	0.0271	0.7791
0.0222	9.3	1200	0.0258	0.7837
0.0199	10.85	1400	0.0302	0.7791
0.0182	12.4	1600	0.0244	0.7815
0.0166	13.95	1800	0.0299	0.7793
0.0166	15.5	2000	0.0256	0.7818
0.0154	17.05	2200	0.0275	0.7880
0.0145	18.6	2400	0.0283	0.7780
0.0138	20.16	2600	0.0299	0.7810
0.0138	21.71	2800	0.0298	0.7834
0.013	23.26	3000	0.0279	0.7813
0.0125	24.81	3200	0.0279	0.7802
0.0119	26.36	3400	0.0281	0.7790
0.0117	27.91	3600	0.0289	0.7801

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2