Update README.md
Browse files
README.md
CHANGED
|
@@ -4,10 +4,6 @@
|
|
| 4 |
🚀 模型性能
|
| 5 |
该模型在 IMDB 数据集的评估集上取得了以下性能：
|
| 6 |
|
| 7 |
-
指标 (Metric)
|
| 8 |
-
|
| 9 |
-
数值 (Value)
|
| 10 |
-
|
| 11 |
评估损失 (Loss)
|
| 12 |
|
| 13 |
0.3455
|
|
@@ -53,78 +49,3 @@ print(f"情感分析结果: {result_neg}")
|
|
| 53 |
|
| 54 |
训练过程
|
| 55 |
模型微调是基于 Hugging Face transformers 库的 Trainer API 完成的。
|
| 56 |
-
|
| 57 |
-
超参数 (Hyperparameters)
|
| 58 |
-
超参数
|
| 59 |
-
|
| 60 |
-
值
|
| 61 |
-
|
| 62 |
-
learning_rate
|
| 63 |
-
|
| 64 |
-
2e-05
|
| 65 |
-
|
| 66 |
-
train_batch_size
|
| 67 |
-
|
| 68 |
-
16
|
| 69 |
-
|
| 70 |
-
eval_batch_size
|
| 71 |
-
|
| 72 |
-
16
|
| 73 |
-
|
| 74 |
-
seed
|
| 75 |
-
|
| 76 |
-
42
|
| 77 |
-
|
| 78 |
-
optimizer
|
| 79 |
-
|
| 80 |
-
AdamW (betas=(0.9,0.999), epsilon=1e-08)
|
| 81 |
-
|
| 82 |
-
lr_scheduler_type
|
| 83 |
-
|
| 84 |
-
linear
|
| 85 |
-
|
| 86 |
-
num_epochs
|
| 87 |
-
|
| 88 |
-
2
|
| 89 |
-
|
| 90 |
-
训练结果日志
|
| 91 |
-
Training Loss
|
| 92 |
-
|
| 93 |
-
Epoch
|
| 94 |
-
|
| 95 |
-
Step
|
| 96 |
-
|
| 97 |
-
Validation Loss
|
| 98 |
-
|
| 99 |
-
Accuracy
|
| 100 |
-
|
| 101 |
-
No log
|
| 102 |
-
|
| 103 |
-
1.0
|
| 104 |
-
|
| 105 |
-
63
|
| 106 |
-
|
| 107 |
-
0.4222
|
| 108 |
-
|
| 109 |
-
0.844
|
| 110 |
-
|
| 111 |
-
No log
|
| 112 |
-
|
| 113 |
-
2.0
|
| 114 |
-
|
| 115 |
-
126
|
| 116 |
-
|
| 117 |
-
0.3455
|
| 118 |
-
|
| 119 |
-
0.85
|
| 120 |
-
|
| 121 |
-
框架版本
|
| 122 |
-
Transformers: 4.56.0
|
| 123 |
-
|
| 124 |
-
Pytorch: 2.8.0+cu126
|
| 125 |
-
|
| 126 |
-
Datasets: 4.0.0
|
| 127 |
-
|
| 128 |
-
Tokenizers: 0.22.0
|
| 129 |
-
|
| 130 |
-
该模型卡由 Gemini 协助生成。
|
|
|
|
| 4 |
🚀 模型性能
|
| 5 |
该模型在 IMDB 数据集的评估集上取得了以下性能：
|
| 6 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
评估损失 (Loss)
|
| 8 |
|
| 9 |
0.3455
|
|
|
|
| 49 |
|
| 50 |
训练过程
|
| 51 |
模型微调是基于 Hugging Face transformers 库的 Trainer API 完成的。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|