{"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/improved-baselines-with-visual-instruction","title":"Improved Baselines with Visual Instruction Tuning","arxiv_id":"2310.03744","date":"2023-10-05","proceeding":"CVPR 2024 1","authors":["Haotian Liu","Chunyuan Li","Yuheng Li","Yong Jae Lee"],"abstract":"Large multimodal models (LMM) have recently shown encouraging progress with visual instruction tuning. In this note, we show that the fully-connected vision-language cross-modal connector in LLaVA is surprisingly powerful and data-efficient. With simple modifications to LLaVA, namely, using CLIP-ViT-L-336px with an MLP projection and adding academic-task-oriented VQA data with simple response formatting prompts, we establish stronger baselines that achieve state-of-the-art across 11 benchmarks. Our final 13B checkpoint uses merely 1.2M publicly available data, and finishes full training in ~1 day on a single 8-A100 node. We hope this can make state-of-the-art LMM research more accessible. Code and model will be publicly available.","url_abs":"https://arxiv.org/abs/2310.03744v2","url_pdf":"https://arxiv.org/pdf/2310.03744v2.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":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/LLaVA-VL/LLaVA-NeXT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/albertotestoni/ndq_visual_objects","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/dinhvietcuong1996/icme25-inova","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/haotian-liu/LLaVA","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/huggingface/transformers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/linzhiqiu/clip-flant5","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/skunkworksai/bakllava","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/sshh12/multi_token","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"improved-baselines-with-visual-instruction","repo_url":"https://github.com/x2fd/lvis-instruct4v","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"factual-inconsistency-detection-in-chart","task_name":"Factual Inconsistency Detection in Chart Captioning"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"referring-expression-comprehension","task_name":"Referring Expression Comprehension"},{"task_slug":"referring-expression-generation","task_name":"Referring expression generation"},{"task_slug":"spatial-reasoning","task_name":"Spatial Reasoning"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"},{"task_slug":"visual-instruction-following","task_name":"visual instruction following"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/factual-inconsistency-detection-in-chart-2","task":"Factual Inconsistency Detection in Chart Captioning","dataset":"CHOCOLATE-FT","model":"LLaVA-1.5-13B","rank_in_archive_order":4,"of":5,"metrics":{"Kendall's Tau-c":"0.214"},"uses_additional_data":false},{"leaderboard":"/sota/factual-inconsistency-detection-in-chart-1","task":"Factual Inconsistency Detection in Chart Captioning","dataset":"CHOCOLATE-LLM","model":"LLaVA-1.5-13B","rank_in_archive_order":5,"of":5,"metrics":{"Kendall's Tau-c":"0.057"},"uses_additional_data":false},{"leaderboard":"/sota/factual-inconsistency-detection-in-chart-3","task":"Factual Inconsistency Detection in Chart Captioning","dataset":"CHOCOLATE-LVLM","model":"LLaVA-1.5-13B","rank_in_archive_order":4,"of":5,"metrics":{"Kendall's Tau-c":"0.002"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-coloninst-v1-seen","task":"Image Classification","dataset":"ColonINST-v1 (Seen)","model":"LLaVA-v1.5 (w/ LoRA, w/ extra data)","rank_in_archive_order":6,"of":17,"metrics":{"Accuray":"93.33"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-coloninst-v1-seen","task":"Image Classification","dataset":"ColonINST-v1 (Seen)","model":"LLaVA-v1.5 (w/ LoRA, w/o extra data)","rank_in_archive_order":9,"of":17,"metrics":{"Accuray":"92.97"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-coloninst-v1-unseen","task":"Image Classification","dataset":"ColonINST-v1 (Unseen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/ extra data)","rank_in_archive_order":2,"of":17,"metrics":{"Accuray":"80.89"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-coloninst-v1-unseen","task":"Image Classification","dataset":"ColonINST-v1 (Unseen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/o extra data)","rank_in_archive_order":6,"of":17,"metrics":{"Accuray":"79.10"},"uses_additional_data":false},{"leaderboard":"/sota/referring-expression-generation-on-coloninst","task":"Referring expression generation","dataset":"ColonINST-v1 (Seen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/ extra data)","rank_in_archive_order":2,"of":17,"metrics":{"Accuray":"99.32"},"uses_additional_data":false},{"leaderboard":"/sota/referring-expression-generation-on-coloninst","task":"Referring expression generation","dataset":"ColonINST-v1 (Seen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/o extra data)","rank_in_archive_order":5,"of":17,"metrics":{"Accuray":"98.58"},"uses_additional_data":false},{"leaderboard":"/sota/referring-expression-generation-on-coloninst-1","task":"Referring expression generation","dataset":"ColonINST-v1 (Unseen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/ extra data)","rank_in_archive_order":9,"of":17,"metrics":{"Accuray":"72.88"},"uses_additional_data":false},{"leaderboard":"/sota/referring-expression-generation-on-coloninst-1","task":"Referring expression generation","dataset":"ColonINST-v1 (Unseen)","model":"LLaVA-v1.5\n  (w/ LoRA, w/o extra data)","rank_in_archive_order":11,"of":17,"metrics":{"Accuray":"70.38"},"uses_additional_data":false},{"leaderboard":"/sota/spatial-reasoning-on-6-dof-spatialbench","task":"Spatial Reasoning","dataset":"6-DoF SpatialBench","model":"LLaVA-1.5","rank_in_archive_order":7,"of":7,"metrics":{"Orientation-abs":"25.8","Orientation-rel":"28.3","Position-abs":"24.5","Position-rel":"30.9","Total":"27.2"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-benchlmm","task":"Visual Question Answering","dataset":"BenchLMM","model":"LLaVA-1.5-13B","rank_in_archive_order":3,"of":10,"metrics":{"GPT-3.5 score":"55.53"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet","task":"Visual Question Answering","dataset":"MM-Vet","model":"LLaVA-1.5-13B","rank_in_archive_order":145,"of":231,"metrics":{"GPT-4 score":"36.3±0.2","Params":"13B"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet","task":"Visual Question Answering","dataset":"MM-Vet","model":"LLaVA-1.5-7B","rank_in_archive_order":197,"of":231,"metrics":{"GPT-4 score":"31.1±0.2","Params":"7B"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet-v2","task":"Visual Question Answering","dataset":"MM-Vet v2","model":"LLaVA-v1.5-13B","rank_in_archive_order":21,"of":24,"metrics":{"GPT-4 score":"33.2±0.1","Params":"13B"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet-v2","task":"Visual Question Answering","dataset":"MM-Vet v2","model":"LLaVA-v1.5-7B","rank_in_archive_order":22,"of":24,"metrics":{"GPT-4 score":"28.3±0.2","Params":"7B"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-vip-bench","task":"Visual Question Answering","dataset":"ViP-Bench","model":"LLaVA-1.5-13B (Coordinates)","rank_in_archive_order":5,"of":13,"metrics":{"GPT-4 score (bbox)":"47.1"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-vip-bench","task":"Visual Question Answering","dataset":"ViP-Bench","model":"LLaVA-1.5-13B (Visual Prompt)","rank_in_archive_order":8,"of":13,"metrics":{"GPT-4 score (bbox)":"41.8","GPT-4 score (human)":"42.9"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-vqa-on-5","task":"Visual Question Answering (VQA)","dataset":"AutoHallusion","model":"LLaVA-1.5","rank_in_archive_order":4,"of":5,"metrics":{"Overall Accuracy":"44.5"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-vqa-on-core-mm","task":"Visual Question Answering (VQA)","dataset":"InfiMM-Eval","model":"LLaVA-1.5","rank_in_archive_order":5,"of":14,"metrics":{"Abductive":"47.91","Analogical":"24.31","Deductive":"30.94","Overall score":"32.62","Params":"13B"},"uses_additional_data":false},{"leaderboard":"/sota/visual-instruction-following-on-llava-bench","task":"visual instruction following","dataset":"LLaVA-Bench","model":"LLaVA-v1.5-13B","rank_in_archive_order":4,"of":8,"metrics":{"avg score":"70.7"},"uses_additional_data":false},{"leaderboard":"/sota/visual-instruction-following-on-llava-bench","task":"visual instruction following","dataset":"LLaVA-Bench","model":"LLaVA-v1.5-7B","rank_in_archive_order":5,"of":8,"metrics":{"avg score":"63.4"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2310.03744","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.03744"}},"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/huggingface/transformers","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/haotian-liu/LLaVA","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/albertotestoni/ndq_visual_objects","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/x2fd/lvis-instruct4v","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sshh12/multi_token","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dinhvietcuong1996/icme25-inova","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/linzhiqiu/clip-flant5","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/skunkworksai/bakllava","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/LLaVA-VL/LLaVA-NeXT","reach":null}],"summary":{"ran_violates":3,"ran_draft_wrong":1,"ran_fixture":2,"unverified":3},"by_repo_kind":{"listed":{"samples":1,"ran":0,"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":8,"samples":[{"code_sha256_prefix":"7bff6f3b3cb560e7","entry":"convert_to_custom_text_state_dict","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"7bff6f3b3cb560e7"}},{"code_sha256_prefix":"dcd422d66b0581d8","entry":"get_cast_dtype","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"dcd422d66b0581d8"}},{"code_sha256_prefix":"763f6fe20bd52fbb","entry":"parse_score","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"763f6fe20bd52fbb"}},{"code_sha256_prefix":"08787a49ef19e0aa","entry":"should_only_save_mm_adapter","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"08787a49ef19e0aa"}},{"code_sha256_prefix":"10893c4608c08075","entry":"split_to_even_chunks","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"10893c4608c08075"}},{"code_sha256_prefix":"55c32993da87759b","entry":"unpad_image","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"55c32993da87759b"}},{"code_sha256_prefix":"275eb8fe6104bf3a","entry":"build_model_from_openai_state_dict","repo":"dinhvietcuong1996/icme25-inova","repo_kind":"listed","path":"llava/model/multimodal_encoder/dev_eva_clip/eva_clip/model.py","file_url":"https://github.com/dinhvietcuong1996/icme25-inova/blob/HEAD/llava/model/multimodal_encoder/dev_eva_clip/eva_clip/model.py","link_basis":"first_harvest_node","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":"275eb8fe6104bf3a"}},{"code_sha256_prefix":"bb35e3ac741bb2c9","entry":"get_mm_adapter_state_maybe_zero_3","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"bb35e3ac741bb2c9"}},{"code_sha256_prefix":"735025744c1ab0cf","entry":"maybe_zero_3","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"735025744c1ab0cf"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}