使用Super-gradients训练YOLO-NAS首次验证后遇编码错误求助

环境版本 #

• torch: 1.13.1
• super-gradients: 3.1.2

训练参数 #

from super_gradients.training.losses import PPYoloELoss
from super_gradients.training.metrics import DetectionMetrics_050
from super_gradients.training.metrics import DetectionMetrics_050_095

from super_gradients.training.models.detection_models.pp_yolo_e import PPYoloEPostPredictionCallback

train_params = {
    'silent_mode': False,
    "average_best_models":True,
    "warmup_mode": "linear_epoch_step",
    "warmup_initial_lr": 1e-6,
    "lr_warmup_epochs": 3,
    "initial_lr": 5e-4,
    "lr_mode": "cosine",
    "cosine_final_lr_ratio": 0.1,
    "optimizer": "Adam",
    "optimizer_params": {"weight_decay": 0.0001},
    "zero_weight_decay_on_bias_and_bn": True,
    "ema": True,
    "ema_params": {"decay": 0.9, "decay_type": "threshold"},
    "max_epochs": EPOCHS,
    "mixed_precision": True,
    "loss": PPYoloELoss(
        use_static_assigner=False,
        num_classes=len(dataset_params['classes']),
        reg_max=16
    ),
    "valid_metrics_list": [
        DetectionMetrics_050(
            score_thres=0.1,
            top_k_predictions=300,
            num_cls=len(dataset_params['classes']),
            normalize_targets=True,
            post_prediction_callback=PPYoloEPostPredictionCallback(
                score_threshold=0.01,
                nms_top_k=1000,
                max_predictions=300,
                nms_threshold=0.7
            )
        ),
        DetectionMetrics_050_095(
            score_thres=0.1,
            top_k_predictions=300,
            num_cls=len(dataset_params['classes']),
            normalize_targets=True,
            post_prediction_callback=PPYoloEPostPredictionCallback(
                score_threshold=0.01,
                nms_top_k=1000,
                max_predictions=300,
                nms_threshold=0.7
            )
        )
    ],
    "metric_to_watch": 'mAP@0.50:0.95'
}

训练流程 #

models_to_train = ['yolo_nas_s']
 
CHECKPOINT_DIR = 'checkpoints'
 
trainer = Trainer(
    experiment_name='yolo_nas_s', 
    ckpt_root_dir=CHECKPOINT_DIR
)

model = models.get(
    'yolo_nas_s', 
    num_classes=len(dataset_params['classes']), 
    pretrained_weights="coco"
)

trainer.train(
    model, 
    training_params=train_params, 
    train_loader=train_data, 
    valid_loader=val_data
)

错误信息 #

"'charmap' codec can't encode characters in position 20-22: character maps to "

这个编码错误多出现于Windows环境，核心原因是程序默认使用系统编码（如GBK）读取含非ASCII字符的文件（标注文件、类别名等）时出现兼容问题，可尝试以下方案：

1. 强制设置Python默认编码为UTF-8
   在训练脚本开头添加代码，覆盖默认输出编码：

import sys
import io
sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8')
sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding='utf-8')

2. 显式指定数据集文件编码
   如果自定义数据集的标注文件（XML/JSON/TXT等）含中文或特殊字符，读取时强制指定utf-8编码：

with open(label_file_path, 'r', encoding='utf-8') as f:
    content = f.read()

若使用SuperGradients内置数据集类，检查是否支持通过参数指定编码，或修改类内文件读取逻辑。

3. 修改控制台编码
   在Windows命令行执行以下命令切换编码为UTF-8，再运行训练脚本：

chcp 65001

4. 清理路径与名称中的特殊字符
   检查数据集的文件路径、文件名、类别名称是否包含中文、生僻字或特殊符号，若有则改为英文/数字命名。

内容的提问来源于stack exchange，提问作者nost99