README.md

metadata

library_name: transformers
license: mit
base_model: nielsr/lilt-xlm-roberta-base
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: test
    results: []

test

This model is a fine-tuned version of nielsr/lilt-xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2434
• Precision: 0.9144
• Recall: 0.9105
• F1: 0.9124
• Accuracy: 0.9725

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-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.7937	100	0.1878	0.8406	0.8761	0.8580	0.9542
No log	1.5873	200	0.1337	0.8943	0.8864	0.8903	0.9650
No log	2.3810	300	0.1259	0.9020	0.9214	0.9116	0.9716
No log	3.1746	400	0.1317	0.9100	0.9181	0.9140	0.9730
0.2107	3.9683	500	0.1159	0.9144	0.9065	0.9104	0.9710
0.2107	4.7619	600	0.1169	0.9147	0.9072	0.9109	0.9715
0.2107	5.5556	700	0.1240	0.9025	0.9144	0.9084	0.9712
0.2107	6.3492	800	0.1351	0.9160	0.9118	0.9139	0.9727
0.2107	7.1429	900	0.1469	0.9207	0.9055	0.9131	0.9722
0.0518	7.9365	1000	0.1333	0.9053	0.9158	0.9105	0.9717
0.0518	8.7302	1100	0.1367	0.9119	0.9167	0.9143	0.9724
0.0518	9.5238	1200	0.1412	0.9057	0.9134	0.9095	0.9712
0.0518	10.3175	1300	0.1666	0.9203	0.9158	0.9180	0.9740
0.0518	11.1111	1400	0.1610	0.9050	0.9062	0.9056	0.9707
0.0316	11.9048	1500	0.1677	0.9175	0.9111	0.9143	0.9720
0.0316	12.6984	1600	0.1838	0.9097	0.9052	0.9074	0.9715
0.0316	13.4921	1700	0.1622	0.9182	0.9082	0.9131	0.9725
0.0316	14.2857	1800	0.1855	0.9161	0.9092	0.9126	0.9725
0.0316	15.0794	1900	0.1739	0.9078	0.9171	0.9124	0.9725
0.0174	15.8730	2000	0.1902	0.9167	0.9167	0.9167	0.9734
0.0174	16.6667	2100	0.1729	0.9207	0.9171	0.9189	0.9739
0.0174	17.4603	2200	0.2083	0.9147	0.9171	0.9159	0.9734
0.0174	18.2540	2300	0.2233	0.9108	0.9177	0.9143	0.9724
0.0174	19.0476	2400	0.2165	0.9201	0.9134	0.9168	0.9730
0.0085	19.8413	2500	0.2138	0.9117	0.9111	0.9114	0.9721
0.0085	20.6349	2600	0.2109	0.9150	0.9108	0.9129	0.9725
0.0085	21.4286	2700	0.2118	0.9216	0.9167	0.9192	0.9742
0.0085	22.2222	2800	0.2287	0.9184	0.9184	0.9184	0.9742
0.0085	23.0159	2900	0.2350	0.9118	0.9085	0.9101	0.9719
0.0043	23.8095	3000	0.2406	0.9109	0.9158	0.9133	0.9727
0.0043	24.6032	3100	0.2480	0.9105	0.9072	0.9088	0.9715
0.0043	25.3968	3200	0.2430	0.9112	0.9055	0.9084	0.9714
0.0043	26.1905	3300	0.2396	0.9092	0.9068	0.9080	0.9712
0.0043	26.9841	3400	0.2386	0.9152	0.9164	0.9158	0.9732
0.0026	27.7778	3500	0.2417	0.9123	0.9111	0.9117	0.9720
0.0026	28.5714	3600	0.2433	0.9136	0.9085	0.9110	0.9721
0.0026	29.3651	3700	0.2434	0.9144	0.9105	0.9124	0.9725

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1