{"componentChunkName":"component---src-templates-acg-portal-new-template-tsx","path":"/9mu0o7cif","result":{"data":{"markdownRemark":{"html":"

简介

\n

图图相似度算子：两张图各过一遍 CLIP 图像塔，取余弦相似度，逐行产出分数，过滤交给 pipeline。CLIP 封装复用 daft/aihc/functions/multimodal/_clip_dual_tower.py。

\n

功能描述

\n\n

算子参数

\n

输入

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
输入含义
images_a第一路图像列，内容形式由 image_src_type 决定
images_b第二路图像列，与 images_a 按行配对，形式相同
\n

输出

\n\n\n\n\n\n\n\n\n\n\n\n\n\n
输出含义
simfloat64，两张图的 CLIP 余弦相似度，取值 [-1, 1]；任一侧为空或该微批失败时为 None
\n

参数

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
参数名称类型默认值描述
image_src_typestr\"image_url\"两列图像共用的编码形式：image_url / image_base64 / image_binary
model_pathstr\"/opt/aihc/model\"权重根目录
model_namestr\"openai/clip-vit-base-patch32\"相对 model_path 的 CLIP 权重子目录
dtypestr\"float32\"权重精度：float16 / float32 / bfloat16，其他值抛 ValueError
batch_sizeint16推理微批大小（算子内部分批，与 aihc_udf(batch_size=...) 的行批是两回事）
rankint0多卡场景 worker 序号，用来选 cuda:{rank % 卡数}
\n

注意事项

\n\n

调用示例

\n\n
\n
\n
\n Python\n \n
\n
\n 
1from __future__ import annotations\n2\n3import os\n4\n5import daft\n6from daft import col\n7\n8from daft.aihc.common.udf import aihc_udf\n9from daft.aihc.functions.image.image_pair_similarity import ImagePairSimilarity\n10\n11if __name__ == \"__main__\":\n12    if os.getenv(\"DAFT_RUNNER\", \"native\") == \"ray\":\n13        import ray\n14        ray.init(dashboard_host=\"0.0.0.0\", ignore_reinit_error=True)\n15        daft.set_runner_ray()\n16    daft.set_execution_config(actor_udf_ready_timeout=6000, min_cpu_per_task=0)\n17\n18samples = {\n19        \"a\": [\"bos://your-bucket/sample.jpg\"],\n20        \"b\": [\"bos://your-bucket/sample.jpg\"],\n21    }\n22    ds = daft.from_pydict(samples)\n23    ds = ds.with_column(\n24        \"sim\",\n25        aihc_udf(\n26            ImagePairSimilarity,\n27            construct_args={\n28                \"image_src_type\": \"image_url\",\n29                \"model_path\": \"/path/to/models\",\n30                \"model_name\": \"openai/clip-vit-base-patch32\",\n31                \"dtype\": \"float32\",\n32                \"batch_size\": 16,\n33            },\n34            num_cpus=1,\n35            num_gpus=1,\n36            concurrency=1,\n37            batch_size=16,\n38        )(col(\"a\"), col(\"b\")),\n39    )\n40    ds.show()
\n
\n
\n
\n ","fields":{"slug":"9mu0o7cif","title":"图像对相似度（CLIP）","date":"2026-09-14","extractedHeadings":[]},"headings":[{"value":"简介","depth":2},{"value":"功能描述","depth":3},{"value":"算子参数","depth":2},{"value":"输入","depth":3},{"value":"输出","depth":3},{"value":"参数","depth":3},{"value":"注意事项","depth":2},{"value":"调用示例","depth":2}]}},"pageContext":{"isCreatedByStatefulCreatePages":false,"slug":"9mu0o7cif","prev":{"id":"Vmu0o6bo5","name":"图像字节大小","path":"Vmu0o6bo5","filePath":"操作指南/AI数据处理/算子列表/图片/图像字节大小.md","seo":null,"parentIds":["ilib2qygp","Ymo88m8hi","Imob3m6so","6mob45l3j"],"parents":[{"id":"ilib2qygp","documentId":"bfa43a8b-968a-41a1-8c9d-906507eeaed9","name":"操作指南","repoName":"AIHC","filePath":"操作指南","disabled":false,"path":"ilib2qygp","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":null,"solutionTag":null,"solutionColor":null},{"id":"Ymo88m8hi","documentId":"c8cb5e38-f8c5-40f4-a424-b0c7895f0c0a","name":"AI数据处理","repoName":"AIHC","filePath":"操作指南/AI数据处理","disabled":false,"path":"Ymo88m8hi","lastMergeTime":"2026-04-21 14:23:10","isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"Imob3m6so","documentId":"fe548e34-6659-4ff5-86f6-eee2c43aec90","name":"算子列表","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表","disabled":false,"path":"Imob3m6so","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"6mob45l3j","documentId":"f121147a-d538-4da6-b212-1afecb2ecd42","name":"图片","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表/图片","disabled":false,"path":"6mob45l3j","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null}]},"next":{"id":"nmu0o8ds3","name":"图像扩散生成","path":"nmu0o8ds3","filePath":"操作指南/AI数据处理/算子列表/图片/图像扩散生成.md","seo":null,"parentIds":["ilib2qygp","Ymo88m8hi","Imob3m6so","6mob45l3j"],"parents":[{"id":"ilib2qygp","documentId":"bfa43a8b-968a-41a1-8c9d-906507eeaed9","name":"操作指南","repoName":"AIHC","filePath":"操作指南","disabled":false,"path":"ilib2qygp","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":null,"solutionTag":null,"solutionColor":null},{"id":"Ymo88m8hi","documentId":"c8cb5e38-f8c5-40f4-a424-b0c7895f0c0a","name":"AI数据处理","repoName":"AIHC","filePath":"操作指南/AI数据处理","disabled":false,"path":"Ymo88m8hi","lastMergeTime":"2026-04-21 14:23:10","isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"Imob3m6so","documentId":"fe548e34-6659-4ff5-86f6-eee2c43aec90","name":"算子列表","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表","disabled":false,"path":"Imob3m6so","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"6mob45l3j","documentId":"f121147a-d538-4da6-b212-1afecb2ecd42","name":"图片","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表/图片","disabled":false,"path":"6mob45l3j","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null}]},"parents":[{"id":"ilib2qygp","documentId":"bfa43a8b-968a-41a1-8c9d-906507eeaed9","name":"操作指南","repoName":"AIHC","filePath":"操作指南","disabled":false,"path":"ilib2qygp","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":null,"solutionTag":null,"solutionColor":null},{"id":"Ymo88m8hi","documentId":"c8cb5e38-f8c5-40f4-a424-b0c7895f0c0a","name":"AI数据处理","repoName":"AIHC","filePath":"操作指南/AI数据处理","disabled":false,"path":"Ymo88m8hi","lastMergeTime":"2026-04-21 14:23:10","isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"Imob3m6so","documentId":"fe548e34-6659-4ff5-86f6-eee2c43aec90","name":"算子列表","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表","disabled":false,"path":"Imob3m6so","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null},{"id":"6mob45l3j","documentId":"f121147a-d538-4da6-b212-1afecb2ecd42","name":"图片","repoName":"AIHC","filePath":"操作指南/AI数据处理/算子列表/图片","disabled":false,"path":"6mob45l3j","lastMergeTime":null,"isApiDoc":null,"httpMethod":null,"seo":null,"sourceOrgName":null,"sourceRepoName":null,"sourceDocumentId":"","solutionTag":null,"solutionColor":null}],"specificSeo":null}}}