{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/altclip-altering-the-language-encoder-in-clip","title":"AltCLIP: Altering the Language Encoder in CLIP for Extended Language Capabilities","arxiv_id":"2211.06679","date":"2022-11-12","proceeding":null,"authors":["Zhongzhi Chen","Guang Liu","Bo-Wen Zhang","Fulong Ye","Qinghong Yang","Ledell Wu"],"abstract":"In this work, we present a conceptually simple and effective method to train a strong bilingual/multilingual multimodal representation model. Starting from the pre-trained multimodal representation model CLIP released by OpenAI, we altered its text encoder with a pre-trained multilingual text encoder XLM-R, and aligned both languages and image representations by a two-stage training schema consisting of teacher learning and contrastive learning. We validate our method through evaluations of a wide range of tasks. We set new state-of-the-art performances on a bunch of tasks including ImageNet-CN, Flicker30k-CN, COCO-CN and XTD. Further, we obtain very close performances with CLIP on almost all tasks, suggesting that one can simply alter the text encoder in CLIP for extended capabilities such as multilingual understanding. Our models and code are available at https://github.com/FlagAI-Open/FlagAI.","url_abs":"https://arxiv.org/abs/2211.06679v2","url_pdf":"https://arxiv.org/pdf/2211.06679v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"altclip-altering-the-language-encoder-in-clip","repo_url":"https://github.com/flagai-open/flagai","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"altclip-altering-the-language-encoder-in-clip","repo_url":"https://github.com/pwc-1/Paper-8/tree/main/altclip","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"contrastive-learning","task_name":"Contrastive Learning"},{"task_slug":"cross-modal-retrieval","task_name":"Cross-Modal Retrieval"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"image-to-text-retrieval","task_name":"Image-to-Text Retrieval"},{"task_slug":"text-to-image-generation","task_name":"Text-to-Image Generation"},{"task_slug":"xlm-r","task_name":"XLM-R"},{"task_slug":"zero-shot-cross-modal-retrieval","task_name":"Zero-Shot Cross-Modal Retrieval"},{"task_slug":"zero-shot-image-classification","task_name":"Zero-Shot Image Classification"},{"task_slug":"zero-shot-transfer-image-classification","task_name":"Zero-Shot Transfer Image Classification"},{"task_slug":"zero-shot-transfer-image-classification-cn","task_name":"Zero-Shot Transfer Image Classification (CN)"},{"task_slug":"zero-shot-image-retrieval","task_name":"Zero-shot Image Retrieval"},{"task_slug":"zero-shot-text-retrieval","task_name":"Zero-shot Text Retrieval"}],"methods":[{"method_slug":"altclip","method_name":"AltCLIP"},{"method_slug":"altdiffusion","method_name":"AltDiffusion"},{"method_slug":"clip","method_name":"CLIP"},{"method_slug":"diffusion","method_name":"Diffusion"},{"method_slug":"xlm-r","method_name":"XLM-R"}],"datasets_introduced":[],"methods_introduced":[{"slug":"altclip","name":"AltCLIP","full_name":"AltCLIP"},{"slug":"altdiffusion","name":"AltDiffusion","full_name":"AltDiffusion"}],"results":[{"leaderboard":"/sota/zero-shot-cross-modal-retrieval-on-flickr30k","task":"Zero-Shot Cross-Modal Retrieval","dataset":"Flickr30k","model":"AltCLIP","rank_in_archive_order":17,"of":22,"metrics":{"Image-to-text R@1":"86","Image-to-text R@10":"99.1","Image-to-text R@5":"98","Text-to-image R@1":"72.5","Text-to-image R@10":"95.4","Text-to-image R@5":"91.6"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-cn","task":"Zero-Shot Transfer Image Classification","dataset":"CN-ImageNet","model":"AltCLIP","rank_in_archive_order":2,"of":2,"metrics":{"Accuracy (Private)":"59.6"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-cn-4","task":"Zero-Shot Transfer Image Classification","dataset":"CN-ImageNet V2","model":"AltCLIP","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy (Private)":"50.9"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-cn-2","task":"Zero-Shot Transfer Image Classification","dataset":"CN-ImageNet-A","model":"AltCLIP","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy (Private)":"58.5"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-cn-3","task":"Zero-Shot Transfer Image Classification","dataset":"CN-ImageNet-R","model":"AltCLIP","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy (Private)":"79.9"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-cn-1","task":"Zero-Shot Transfer Image Classification","dataset":"CN-ImageNet-Sketch","model":"AltCLIP","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy (Private)":"46.5"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-1","task":"Zero-Shot Transfer Image Classification","dataset":"ImageNet","model":"AltCLIP","rank_in_archive_order":18,"of":23,"metrics":{"Accuracy (Private)":"74.5"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-3","task":"Zero-Shot Transfer Image Classification","dataset":"ImageNet V2","model":"AltCLIP","rank_in_archive_order":12,"of":13,"metrics":{"Accuracy (Private)":"68.1"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-5","task":"Zero-Shot Transfer Image Classification","dataset":"ImageNet-A","model":"AltCLIP","rank_in_archive_order":12,"of":13,"metrics":{"Accuracy (Private)":"69.5"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-4","task":"Zero-Shot Transfer Image Classification","dataset":"ImageNet-R","model":"AltCLIP","rank_in_archive_order":11,"of":12,"metrics":{"Accuracy":"87.2"},"uses_additional_data":false},{"leaderboard":"/sota/zero-shot-transfer-image-classification-on-8","task":"Zero-Shot Transfer Image Classification","dataset":"ImageNet-Sketch","model":"AltCLIP","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (Private)":"58.7"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2211.06679","atlas_url":"https://app.syntology.ai/?focus=2211.06679","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2211.06679"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/flagai-open/flagai","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/pwc-1/Paper-8/tree/main/altclip","reach":null}],"summary":{"ran":10,"unverified":1},"by_repo_kind":{"official":{"samples":11,"ran":10,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"e923a6d5aaa37270","entry":"accuracy_metric","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/metrics.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/metrics.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e923a6d5aaa37270"}},{"code_sha256_prefix":"9a58da55dcd68a4b","entry":"change_json_to_cls","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/model/base_model.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/model/base_model.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9a58da55dcd68a4b"}},{"code_sha256_prefix":"329c97f25378e517","entry":"cmd_load_hyperparam","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/launch.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/launch.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"329c97f25378e517"}},{"code_sha256_prefix":"fd6777c43a4e9d66","entry":"create_custom_forward","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/model/aquila_model.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/model/aquila_model.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"fd6777c43a4e9d66"}},{"code_sha256_prefix":"2774d56117e0ff9f","entry":"get_args_list","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/env_trainer.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/env_trainer.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"2774d56117e0ff9f"}},{"code_sha256_prefix":"8977724aa4056e15","entry":"get_args_list","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/env_trainer_v1.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/env_trainer_v1.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8977724aa4056e15"}},{"code_sha256_prefix":"5011d61e17e5dcd5","entry":"save_best","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/env_args.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/env_args.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"5011d61e17e5dcd5"}},{"code_sha256_prefix":"98a2b2f174e040a3","entry":"sigmoid","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/metrics.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/metrics.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"98a2b2f174e040a3"}},{"code_sha256_prefix":"92a5f0d0912c3a91","entry":"str2bool","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/env_args.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/env_args.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"92a5f0d0912c3a91"}},{"code_sha256_prefix":"6b5d363ab4dbf012","entry":"to_python_float","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/fp16/loss_scaler.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/fp16/loss_scaler.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6b5d363ab4dbf012"}},{"code_sha256_prefix":"4e8fb0839eac7d8c","entry":"should_log_le","repo":"flagai-open/flagai","repo_kind":"official","path":"flagai/logger.py","file_url":"https://github.com/flagai-open/flagai/blob/HEAD/flagai/logger.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"4e8fb0839eac7d8c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}