README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
base_model: albert/albert-base-v2
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: modelsent_test
    results: []

modelsent_test

This model is a fine-tuned version of albert/albert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2510
• Accuracy: 0.9261
• F1: 0.9261
• Precision: 0.9261
• Recall: 0.9261
• Accuracy Label Negative: 0.9255
• Accuracy Label Positive: 0.9266

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Accuracy Label Negative	Accuracy Label Positive
0.5848	0.2442	100	0.5668	0.7783	0.7774	0.7869	0.7783	0.8548	0.7065
0.2761	0.4884	200	0.2858	0.8913	0.8912	0.8944	0.8913	0.9318	0.8533
0.2099	0.7326	300	0.2412	0.9114	0.9114	0.9116	0.9114	0.8965	0.9254
0.2717	0.9768	400	0.2532	0.9133	0.9133	0.9141	0.9133	0.9318	0.8959
0.2076	1.2210	500	0.2588	0.9084	0.9083	0.9111	0.9084	0.9457	0.8734
0.1745	1.4652	600	0.2217	0.9133	0.9132	0.9133	0.9133	0.9028	0.9231
0.21	1.7094	700	0.2161	0.9157	0.9157	0.9157	0.9157	0.9078	0.9231
0.1349	1.9536	800	0.2092	0.9243	0.9242	0.9245	0.9243	0.9078	0.9396
0.1795	2.1978	900	0.2492	0.9175	0.9175	0.9189	0.9175	0.9432	0.8935
0.107	2.4420	1000	0.2743	0.9120	0.9120	0.9163	0.9120	0.9596	0.8675
0.08	2.6862	1100	0.2606	0.9188	0.9188	0.9200	0.9188	0.9432	0.8959
0.1275	2.9304	1200	0.2550	0.9255	0.9255	0.9255	0.9255	0.9167	0.9337

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1