README.md

metadata

library_name: transformers
license: apache-2.0
base_model: TurboPascal/ChineseModernBert
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: ner_based_ChineseModernBert
    results: []

ner_based_ChineseModernBert

This model is a fine-tuned version of TurboPascal/ChineseModernBert on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0221
• Precision: 0.9440
• Recall: 0.9449
• F1: 0.9445
• Accuracy: 0.9963

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• 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: cosine
• lr_scheduler_warmup_steps: 20
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	327	0.0271	0.8688	0.8590	0.8639	0.9916
0.0912	2.0	654	0.0196	0.8851	0.9370	0.9103	0.9942
0.0912	3.0	981	0.0177	0.9227	0.9283	0.9255	0.9953
0.012	4.0	1308	0.0178	0.9329	0.9387	0.9358	0.9957
0.0047	5.0	1635	0.0198	0.9340	0.9314	0.9327	0.9957
0.0047	6.0	1962	0.0191	0.9334	0.9435	0.9384	0.9960
0.0024	7.0	2289	0.0224	0.9426	0.9404	0.9415	0.9961
0.0014	8.0	2616	0.0223	0.9420	0.9466	0.9443	0.9963
0.0014	9.0	2943	0.0220	0.9344	0.9410	0.9377	0.9960
0.0009	10.0	3270	0.0221	0.9440	0.9449	0.9445	0.9963

Framework versions

• Transformers 4.54.0
• Pytorch 2.7.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4