metadata

base_model: FacebookAI/xlm-roberta-base
library_name: transformers
license: mit
metrics:
  - precision
  - recall
  - f1
  - accuracy
tags:
  - generated_from_trainer
model-index:
  - name: scenario-kd-pre-ner-full-xlmr_data-univner_full44
    results: []

scenario-kd-pre-ner-full-xlmr_data-univner_full44

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: 0.4609
Precision: 0.8151
Recall: 0.8199
F1: 0.8175
Accuracy: 0.9812

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: 3e-05
train_batch_size: 8
eval_batch_size: 32
seed: 44
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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
1.4889	0.2911	500	0.8586	0.6595	0.7179	0.6875	0.9708
0.756	0.5822	1000	0.7200	0.7071	0.7746	0.7393	0.9739
0.6626	0.8732	1500	0.6671	0.7417	0.7530	0.7473	0.9760
0.5626	1.1643	2000	0.6205	0.7676	0.7785	0.7730	0.9772
0.514	1.4554	2500	0.6545	0.8094	0.7420	0.7743	0.9774
0.4905	1.7465	3000	0.5754	0.7770	0.7869	0.7819	0.9782
0.461	2.0375	3500	0.5559	0.7697	0.8101	0.7894	0.9790
0.4097	2.3286	4000	0.5613	0.7862	0.7836	0.7849	0.9785
0.3973	2.6197	4500	0.5514	0.7850	0.8003	0.7926	0.9795
0.3878	2.9108	5000	0.5299	0.7913	0.8039	0.7975	0.9791
0.3579	3.2019	5500	0.5424	0.8023	0.7852	0.7936	0.9790
0.3434	3.4929	6000	0.5077	0.7881	0.8085	0.7982	0.9795
0.3362	3.7840	6500	0.5244	0.8012	0.7943	0.7977	0.9793
0.3243	4.0751	7000	0.5158	0.8068	0.8108	0.8088	0.9801
0.3134	4.3662	7500	0.5081	0.8001	0.8137	0.8069	0.9799
0.3027	4.6573	8000	0.4989	0.8003	0.8169	0.8085	0.9803
0.2977	4.9483	8500	0.4926	0.8013	0.8121	0.8067	0.9804
0.2822	5.2394	9000	0.4905	0.8052	0.8081	0.8067	0.9801
0.2773	5.5305	9500	0.4864	0.8012	0.8049	0.8031	0.9798
0.2803	5.8216	10000	0.4883	0.7963	0.8090	0.8026	0.9798
0.2717	6.1126	10500	0.4941	0.8169	0.7909	0.8037	0.9798
0.258	6.4037	11000	0.4842	0.8008	0.8078	0.8043	0.9802
0.2572	6.6948	11500	0.4760	0.8129	0.8097	0.8113	0.9805
0.2553	6.9859	12000	0.4742	0.8119	0.8116	0.8117	0.9809
0.2462	7.2770	12500	0.4791	0.8116	0.8054	0.8085	0.9806
0.2447	7.5680	13000	0.4750	0.8017	0.8171	0.8093	0.9804
0.2463	7.8591	13500	0.4657	0.8179	0.8113	0.8146	0.9811
0.2381	8.1502	14000	0.4677	0.8025	0.8153	0.8088	0.9805
0.2357	8.4413	14500	0.4658	0.8135	0.8184	0.8159	0.9810
0.2333	8.7324	15000	0.4638	0.8144	0.8116	0.8130	0.9807
0.234	9.0234	15500	0.4605	0.8126	0.8165	0.8145	0.9810
0.2297	9.3145	16000	0.4670	0.8116	0.8080	0.8098	0.9808
0.2258	9.6056	16500	0.4651	0.8095	0.8142	0.8118	0.9808
0.2272	9.8967	17000	0.4609	0.8151	0.8199	0.8175	0.9812

Framework versions

Transformers 4.44.2
Pytorch 2.1.1+cu121
Datasets 2.14.5
Tokenizers 0.19.1