---
license: apache-2.0
language:
- zh
metrics:
- accuracy
base_model:
- google-bert/bert-base-chinese
pipeline_tag: text-classification
library_name: transformers
---

最终评估结果:
======================================================================
  eval_loss............................... 0.9778
  eval_accuracy........................... 0.7857
  eval_precision.......................... 0.7908
  eval_recall............................. 0.7857
  eval_f1................................. 0.7875
  eval_1star_f1........................... 0.8641
  eval_2star_f1........................... 0.7324
  eval_3star_f1........................... 0.6420
  eval_4star_f1........................... 0.8471
  eval_5star_f1........................... 0.7903
  eval_runtime............................ 32.9546
  eval_samples_per_second................. 273.0120
  eval_steps_per_second................... 4.2790
  epoch................................... 6.0000

详细分类报告:
======================================================================
              precision    recall  f1-score   support

          1星     0.8581    0.8703    0.8641      1056
          2星     0.7533    0.7127    0.7324      1615
          3星     0.6075    0.6808    0.6420      1532
          4星     0.8688    0.8265    0.8471      3822
          5星     0.7627    0.8200    0.7903       972

    accuracy                         0.7857      8997
   macro avg     0.7701    0.7821    0.7752      8997
weighted avg     0.7908    0.7857    0.7875      8997