{"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/pargo-bridging-vision-language-with-partial","title":"ParGo: Bridging Vision-Language with Partial and Global Views","arxiv_id":"2408.12928","date":"2024-08-23","proceeding":null,"authors":["An-Lan Wang","Bin Shan","Wei Shi","Kun-Yu Lin","Xiang Fei","Guozhi Tang","Lei Liao","Jingqun Tang","Can Huang","Wei-Shi Zheng"],"abstract":"This work presents ParGo, a novel Partial-Global projector designed to connect the vision and language modalities for Multimodal Large Language Models (MLLMs). Unlike previous works that rely on global attention-based projectors, our ParGo bridges the representation gap between the separately pre-trained vision encoders and the LLMs by integrating global and partial views, which alleviates the overemphasis on prominent regions. To facilitate the effective training of ParGo, we collect a large-scale detail-captioned image-text dataset named ParGoCap-1M-PT, consisting of 1 million images paired with high-quality captions. Extensive experiments on several MLLM benchmarks demonstrate the effectiveness of our ParGo, highlighting its superiority in aligning vision and language modalities. Compared to conventional Q-Former projector, our ParGo achieves an improvement of 259.96 in MME benchmark. Furthermore, our experiments reveal that ParGo significantly outperforms other projectors, particularly in tasks that emphasize detail perception ability.","url_abs":"https://arxiv.org/abs/2408.12928v2","url_pdf":"https://arxiv.org/pdf/2408.12928v2.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":"pargo-bridging-vision-language-with-partial","repo_url":"https://github.com/bytedance/pargo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"BSD-3-Clause"}}],"tasks":[{"task_slug":"mme","task_name":"MME"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2408.12928","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2408.12928"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/bytedance/pargo","reach":{"status":"ok","spdx":"BSD-3-Clause"}}],"summary":{"ran":7,"ran_violates":1,"unverified":3},"by_repo_kind":{"official":{"samples":11,"ran":8,"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":"09380b456c8170e0","entry":"bertTokenizer","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/language/tokenizer.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/language/tokenizer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"09380b456c8170e0"}},{"code_sha256_prefix":"8d2a5d47063ea17a","entry":"compute_itm","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/fusion/blip2.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/fusion/blip2.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"8d2a5d47063ea17a"}},{"code_sha256_prefix":"38f46ef4e03fee1c","entry":"create_position_ids_from_input_ids","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/language/xlm_roberta.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/language/xlm_roberta.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"38f46ef4e03fee1c"}},{"code_sha256_prefix":"4cb732f513d69dfd","entry":"disabled_train","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/fusion/minigpt.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/fusion/minigpt.py","link_basis":"harvester_set","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"4cb732f513d69dfd"}},{"code_sha256_prefix":"6f022e25aeb2e11a","entry":"encode_img","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/fusion/minigpt.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/fusion/minigpt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"6f022e25aeb2e11a"}},{"code_sha256_prefix":"014a05d25d682eb3","entry":"maybe_autocast","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/fusion/minigpt.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/fusion/minigpt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"014a05d25d682eb3"}},{"code_sha256_prefix":"92e88aa752ef7f19","entry":"split_into_patches","repo":"bytedance/pargo","repo_kind":"official","path":"eval/init_models.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/eval/init_models.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"92e88aa752ef7f19"}},{"code_sha256_prefix":"5790f409b12a5693","entry":"t5TokenizerFast","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/language/tokenizer.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/language/tokenizer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"5790f409b12a5693"}},{"code_sha256_prefix":"0ec9fc2025c16f65","entry":"all_gather_with_grad","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/utils/disttools.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/utils/disttools.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"0ec9fc2025c16f65"}},{"code_sha256_prefix":"b5443bddb79c822e","entry":"autoTokenizerBlip","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/backbone/language/tokenizer.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/backbone/language/tokenizer.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"b5443bddb79c822e"}},{"code_sha256_prefix":"73cecca9f3575f09","entry":"concat_all_gather","repo":"bytedance/pargo","repo_kind":"official","path":"pargo/utils/disttools.py","file_url":"https://github.com/bytedance/pargo/blob/HEAD/pargo/utils/disttools.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"73cecca9f3575f09"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}