返回模型

说明文档

模型	F1 分数	精确率	召回率	阈值	Cluster_2	Cluster_Free
ccip-caformer_b36-24	0.940925	0.938254	0.943612	0.213231	0.89508	0.957017
ccip-caformer-24-randaug-pruned	0.917211	0.933481	0.901499	0.178475	0.890366	0.922375
ccip-v2-caformer_s36-10	0.906422	0.932779	0.881513	0.207757	0.874592	0.89241
ccip-caformer-6-randaug-pruned_fp32	0.878403	0.893648	0.863669	0.195122	0.810176	0.897904
ccip-caformer-5_fp32	0.864363	0.90155	0.830121	0.183973	0.792051	0.862289
ccip-caformer-4_fp32	0.844967	0.870553	0.820842	0.18367	0.795565	0.868133
ccip-caformer_query-12	0.823928	0.871122	0.781585	0.141308	0.787237	0.809426
ccip-caformer-23_randaug_fp32	0.81625	0.854134	0.781585	0.136797	0.745697	0.8068
ccip-caformer-2-randaug-pruned_fp32	0.78561	0.800148	0.771592	0.171053	0.686617	0.728195
ccip-caformer-2_fp32	0.755125	0.790172	0.723055	0.141275	0.64977	0.718516

F1 分数、精确率和召回率的计算将"两张图片中的角色相同"视为正例。阈值是通过在 F1 分数曲线上找到最大值来确定的。
Cluster_2 表示在 DBSCAN 聚类算法中将 min_samples 设置为 2，通过调整 eps 值获得的近似最优聚类结果，并使用 random_adjust_score 评估所得聚类与真实分布之间的相似度。
Cluster_Free 表示在 OPTICS 聚类算法中通过调整 max_eps 和 min_samples 值获得的近似最优解，并使用 random_adjust_score 评估所得聚类与真实分布之间的相似度。

deepghs/ccip_onnx

作者 deepghs

zero-shot-image-classification

↓ 0 ♥ 8

创建时间: 2023-06-25 05:11:05+00:00

更新时间: 2024-09-09 08:34:16+00:00

在 Hugging Face 上查看

文件 (122)

.gitattributes

README.md

ccip-caformer-2-randaug-pruned_fp32/cluster.json

ccip-caformer-2-randaug-pruned_fp32/metrics.json

ccip-caformer-2-randaug-pruned_fp32/model.ckpt

ccip-caformer-2-randaug-pruned_fp32/model_feat.onnx ONNX

ccip-caformer-2-randaug-pruned_fp32/model_metrics.onnx ONNX

ccip-caformer-2-randaug-pruned_fp32/plt_confusion_matrix_pred.png

ccip-caformer-2-randaug-pruned_fp32/plt_confusion_matrix_true.png

ccip-caformer-2-randaug-pruned_fp32/plt_f1.png

ccip-caformer-2-randaug-pruned_fp32/plt_pr.png

ccip-caformer-2-randaug-pruned_fp32/plt_precision.png

ccip-caformer-2-randaug-pruned_fp32/plt_recall.png

ccip-caformer-2-randaug-pruned_fp32/plt_roc.png

ccip-caformer-23_randaug_fp32/cluster.json

ccip-caformer-23_randaug_fp32/metrics.json

ccip-caformer-23_randaug_fp32/model.ckpt

ccip-caformer-23_randaug_fp32/model_feat.onnx ONNX

ccip-caformer-23_randaug_fp32/model_metrics.onnx ONNX

ccip-caformer-23_randaug_fp32/plt_confusion_matrix_pred.png

ccip-caformer-23_randaug_fp32/plt_confusion_matrix_true.png

ccip-caformer-23_randaug_fp32/plt_f1.png

ccip-caformer-23_randaug_fp32/plt_pr.png

ccip-caformer-23_randaug_fp32/plt_precision.png

ccip-caformer-23_randaug_fp32/plt_recall.png

ccip-caformer-23_randaug_fp32/plt_roc.png

ccip-caformer-24-randaug-pruned/cluster.json

ccip-caformer-24-randaug-pruned/metrics.json

ccip-caformer-24-randaug-pruned/model.ckpt

ccip-caformer-24-randaug-pruned/model_feat.onnx ONNX

ccip-caformer-24-randaug-pruned/model_metrics.onnx ONNX

ccip-caformer-24-randaug-pruned/plt_confusion_matrix_pred.png

ccip-caformer-24-randaug-pruned/plt_confusion_matrix_true.png

ccip-caformer-24-randaug-pruned/plt_f1.png

ccip-caformer-24-randaug-pruned/plt_pr.png

ccip-caformer-24-randaug-pruned/plt_precision.png

ccip-caformer-24-randaug-pruned/plt_recall.png

ccip-caformer-24-randaug-pruned/plt_roc.png

ccip-caformer-2_fp32/cluster.json

ccip-caformer-2_fp32/metrics.json

ccip-caformer-2_fp32/model.ckpt

ccip-caformer-2_fp32/model_feat.onnx ONNX

ccip-caformer-2_fp32/model_metrics.onnx ONNX

ccip-caformer-2_fp32/plt_confusion_matrix_pred.png

ccip-caformer-2_fp32/plt_confusion_matrix_true.png

ccip-caformer-2_fp32/plt_f1.png

ccip-caformer-2_fp32/plt_pr.png

ccip-caformer-2_fp32/plt_precision.png

ccip-caformer-2_fp32/plt_recall.png

ccip-caformer-2_fp32/plt_roc.png

ccip-caformer-4_fp32/cluster.json

ccip-caformer-4_fp32/metrics.json

ccip-caformer-4_fp32/model.ckpt

ccip-caformer-4_fp32/model_feat.onnx ONNX

ccip-caformer-4_fp32/model_metrics.onnx ONNX

ccip-caformer-4_fp32/plt_confusion_matrix_pred.png

ccip-caformer-4_fp32/plt_confusion_matrix_true.png

ccip-caformer-4_fp32/plt_f1.png

ccip-caformer-4_fp32/plt_pr.png

ccip-caformer-4_fp32/plt_precision.png

ccip-caformer-4_fp32/plt_recall.png

ccip-caformer-4_fp32/plt_roc.png

ccip-caformer-5_fp32/cluster.json

ccip-caformer-5_fp32/metrics.json

ccip-caformer-5_fp32/model.ckpt

ccip-caformer-5_fp32/model_feat.onnx ONNX

ccip-caformer-5_fp32/model_metrics.onnx ONNX

ccip-caformer-5_fp32/plt_confusion_matrix_pred.png

ccip-caformer-5_fp32/plt_confusion_matrix_true.png

ccip-caformer-5_fp32/plt_f1.png

ccip-caformer-5_fp32/plt_pr.png

ccip-caformer-5_fp32/plt_precision.png

ccip-caformer-5_fp32/plt_recall.png

ccip-caformer-5_fp32/plt_roc.png

ccip-caformer-6-randaug-pruned_fp32/cluster.json

ccip-caformer-6-randaug-pruned_fp32/metrics.json

ccip-caformer-6-randaug-pruned_fp32/model.ckpt

ccip-caformer-6-randaug-pruned_fp32/model_feat.onnx ONNX

ccip-caformer-6-randaug-pruned_fp32/model_metrics.onnx ONNX

ccip-caformer-6-randaug-pruned_fp32/plt_confusion_matrix_pred.png

ccip-caformer-6-randaug-pruned_fp32/plt_confusion_matrix_true.png

ccip-caformer-6-randaug-pruned_fp32/plt_f1.png

ccip-caformer-6-randaug-pruned_fp32/plt_pr.png

ccip-caformer-6-randaug-pruned_fp32/plt_precision.png

ccip-caformer-6-randaug-pruned_fp32/plt_recall.png

ccip-caformer-6-randaug-pruned_fp32/plt_roc.png

ccip-caformer_b36-24/cluster.json

ccip-caformer_b36-24/metrics.json

ccip-caformer_b36-24/model.ckpt

ccip-caformer_b36-24/model_feat.onnx ONNX

ccip-caformer_b36-24/model_metrics.onnx ONNX

ccip-caformer_b36-24/plt_confusion_matrix_pred.png

ccip-caformer_b36-24/plt_confusion_matrix_true.png

ccip-caformer_b36-24/plt_f1.png

ccip-caformer_b36-24/plt_pr.png

ccip-caformer_b36-24/plt_precision.png

ccip-caformer_b36-24/plt_recall.png

ccip-caformer_b36-24/plt_roc.png

ccip-caformer_query-12/cluster.json

ccip-caformer_query-12/metrics.json

ccip-caformer_query-12/model.ckpt

ccip-caformer_query-12/model_feat.onnx ONNX

ccip-caformer_query-12/model_metrics.onnx ONNX

ccip-caformer_query-12/plt_confusion_matrix_pred.png

ccip-caformer_query-12/plt_confusion_matrix_true.png

ccip-caformer_query-12/plt_f1.png

ccip-caformer_query-12/plt_pr.png

ccip-caformer_query-12/plt_precision.png

ccip-caformer_query-12/plt_recall.png

ccip-caformer_query-12/plt_roc.png

ccip-v2-caformer_s36-10/cluster.json

ccip-v2-caformer_s36-10/metrics.json

ccip-v2-caformer_s36-10/model.ckpt

ccip-v2-caformer_s36-10/model_feat.onnx ONNX

ccip-v2-caformer_s36-10/model_metrics.onnx ONNX

ccip-v2-caformer_s36-10/plt_confusion_matrix_pred.png

ccip-v2-caformer_s36-10/plt_confusion_matrix_true.png

ccip-v2-caformer_s36-10/plt_f1.png

ccip-v2-caformer_s36-10/plt_pr.png

ccip-v2-caformer_s36-10/plt_precision.png

ccip-v2-caformer_s36-10/plt_recall.png

ccip-v2-caformer_s36-10/plt_roc.png