qinweijia/finbert-tone-chinese-finetuned-sentiment

来源：🤗有风险🏴‍☠️ 救援者：isilicontext-classificationapache-2.0102M 参数410 MBsafetensors✓ 2 个校验和705 天前更新

finbert-tone-chinese-finetuned-sentiment

This model is a fine-tuned version of yiyanghkust/finbert-tone-chinese on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.2524
Accuracy: 0.6399
Matthews Correlation: 0.4352

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: 32
eval_batch_size: 32
seed: 42
optimizer: Use 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Matthews Correlation
1.0134	1.0	9044	1.0035	0.5918	0.3035
0.8984	2.0	18088	0.9505	0.6123	0.3471
0.77	3.0	27132	0.9348	0.6265	0.3735
0.6549	4.0	36176	0.9536	0.6281	0.3976
0.5588	5.0	45220	1.0170	0.6374	0.4332
0.472	6.0	54264	1.0852	0.6260	0.4124
0.412	7.0	63308	1.1077	0.6342	0.4284
0.3624	8.0	72352	1.2524	0.6399	0.4352
0.3182	9.0	81396	1.3556	0.6287	0.4215
0.3092	10.0	90440	1.4456	0.6264	0.4177

Framework versions

Transformers 4.46.3
Pytorch 2.4.1+cu121
Datasets 3.1.0
Tokenizers 0.20.3

权重

原版

F32409 MB
2 个文件
model.safetensors 409 MBtraining_args.bin 5.4 kB
旧版本磁力链接