{"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/vargpt-v1-1-improve-visual-autoregressive","title":"VARGPT-v1.1: Improve Visual Autoregressive Large Unified Model via Iterative Instruction Tuning and Reinforcement Learning","arxiv_id":"2504.02949","date":"2025-04-03","proceeding":null,"authors":["Xianwei Zhuang","Yuxin Xie","Yufan Deng","Dongchao Yang","Liming Liang","Jinghan Ru","Yuguo Yin","Yuexian Zou"],"abstract":"In this work, we present VARGPT-v1.1, an advanced unified visual autoregressive model that builds upon our previous framework VARGPT. The model preserves the dual paradigm of next-token prediction for visual understanding and next-scale generation for image synthesis. Specifically, VARGPT-v1.1 integrates: (1) a novel training strategy combining iterative visual instruction tuning with reinforcement learning through Direct Preference Optimization (DPO), (2) an expanded training corpus containing 8.3M visual-generative instruction pairs, (3) an upgraded language model backbone using Qwen2, (4) enhanced image generation resolution, and (5) emergent image editing capabilities without architectural modifications. These advancements enable VARGPT-v1.1 to achieve state-of-the-art performance in multimodal understanding and text-to-image instruction-following tasks, demonstrating significant improvements in both comprehension and generation metrics. Notably, through visual instruction tuning, the model acquires image editing functionality while maintaining architectural consistency with its predecessor, revealing the potential for unified visual understanding, generation, and editing. Our findings suggest that well-designed unified visual autoregressive models can effectively adopt flexible training strategies from large language models (LLMs), exhibiting promising scalability. The codebase and model weights are publicly available at https://github.com/VARGPT-family/VARGPT-v1.1.","url_abs":"https://arxiv.org/abs/2504.02949v1","url_pdf":"https://arxiv.org/pdf/2504.02949v1.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":"vargpt-v1-1-improve-visual-autoregressive","repo_url":"https://github.com/vargpt-family/vargpt-v1.1","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"vargpt-v1-1-improve-visual-autoregressive","repo_url":"https://github.com/VARGPT-family/VARGPT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"instruction-following","task_name":"Instruction Following"}],"methods":[{"method_slug":"adopt","method_name":"ADOPT"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2504.02949","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2504.02949"}},"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/vargpt-family/vargpt-v1.1","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/VARGPT-family/VARGPT","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_draft_wrong":3,"unverified":10},"by_repo_kind":{"official":{"samples":5,"ran":1,"repositories":1},"listed":{"samples":8,"ran":2,"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":"9fcdaa6e423e8aa7","entry":"Normalize","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/models/basic_vae.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/models/basic_vae.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9fcdaa6e423e8aa7"}},{"code_sha256_prefix":"3137073275f8c21a","entry":"nonlinearity","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/models/basic_vae.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/models/basic_vae.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3137073275f8c21a"}},{"code_sha256_prefix":"b99eea6376d1e212","entry":"rotate_half","repo":"vargpt-family/vargpt-v1.1","repo_kind":"official","path":"vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.py","file_url":"https://github.com/vargpt-family/vargpt-v1.1/blob/HEAD/vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b99eea6376d1e212"}},{"code_sha256_prefix":"e847ee15f089dee6","entry":"allgather","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/dist.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/dist.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":"e847ee15f089dee6"}},{"code_sha256_prefix":"ff4c548fd07140d0","entry":"allreduce","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/dist.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/dist.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":"ff4c548fd07140d0"}},{"code_sha256_prefix":"eb4d16c3bba7cff8","entry":"apply_multimodal_rotary_pos_emb","repo":"vargpt-family/vargpt-v1.1","repo_kind":"official","path":"vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.py","file_url":"https://github.com/vargpt-family/vargpt-v1.1/blob/HEAD/vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.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":"eb4d16c3bba7cff8"}},{"code_sha256_prefix":"48f12d3a9c4a15e8","entry":"apply_rotary_pos_emb_vision","repo":"vargpt-family/vargpt-v1.1","repo_kind":"official","path":"vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.py","file_url":"https://github.com/vargpt-family/vargpt-v1.1/blob/HEAD/vargpt_qwen_v1_1/modeling_vargpt_qwen2_vl.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":"48f12d3a9c4a15e8"}},{"code_sha256_prefix":"155a5802e5cdb1fb","entry":"check_file_exists","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/prepare_vargpt_llava.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/prepare_vargpt_llava.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":"155a5802e5cdb1fb"}},{"code_sha256_prefix":"793c7e2489dd8dfc","entry":"make_attn","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/models/basic_vae.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/models/basic_vae.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":"793c7e2489dd8dfc"}},{"code_sha256_prefix":"826ea3fc406a6a09","entry":"new_group","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"vargpt_llava/var_model/dist.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/vargpt_llava/var_model/dist.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":"826ea3fc406a6a09"}},{"code_sha256_prefix":"a50444314d796645","entry":"smart_resize","repo":"vargpt-family/vargpt-v1.1","repo_kind":"official","path":"vargpt_qwen_v1_1/image_processing_qwen2_vl.py","file_url":"https://github.com/vargpt-family/vargpt-v1.1/blob/HEAD/vargpt_qwen_v1_1/image_processing_qwen2_vl.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":"a50444314d796645"}},{"code_sha256_prefix":"921ef7496504eda2","entry":"vargpt_sample","repo":"VARGPT-family/VARGPT","repo_kind":"listed","path":"patching_utils/vargpt_sample.py","file_url":"https://github.com/VARGPT-family/VARGPT/blob/HEAD/patching_utils/vargpt_sample.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":"921ef7496504eda2"}},{"code_sha256_prefix":"e8ecc89e84a3116c","entry":"vargpt_sample","repo":"vargpt-family/vargpt-v1.1","repo_kind":"official","path":"patching_utils/vargpt_sample.py","file_url":"https://github.com/vargpt-family/vargpt-v1.1/blob/HEAD/patching_utils/vargpt_sample.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":"e8ecc89e84a3116c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}