Papers › Improved Baselines with Visual Instruction Tuning

Improved Baselines with Visual Instruction Tuning

5 Oct 2023CVPR 2024 1arXiv:2310.03744archive 2025-07-28

Haotian Liu, Chunyuan Li, Yuheng Li, Yong Jae Lee

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.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 9 code samples: 6 from 3 repositories linked to this paper, 3 identical to code harvested from another paper's repository. Across all 9: 6 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 3 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 3 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 6 samples from 3 repositories, 3 ran (1 with no instrument failure; 2 where Syntology's instrument failed); 3 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 5 of its repositories that Syntology harvested for other papers:

No date is recorded for any of the 6 samples that ran: their levels were written before Syntology's graph kept the date.

LLaVA-VL/LLaVA-NeXT mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
albertotestoni/ndq_visual_objects mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Grounding Language Models to Images for Multimodal Inputs and Outputs (arXiv:2301.13823; that paper's own run record: 6 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 2 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
dinhvietcuong1996/icme25-inova mentioned on GitHubpytorchSyntology: 1 sample, 0 ran · 1 unverified report
haotian-liu/LLaVA mentioned on GitHubpytorchSyntology: 4 samples, 2 ran (2 instrument-failed) · 2 unverified report
huggingface/transformers mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Attention Is All You Need (arXiv:1706.03762; that paper's own run record: 610 ran (of which 293 constructed an object rather than computing a result; 529 with no instrument failure: 45 honoured, 15 violated, 469 with no contract checked; 81 where Syntology's instrument failed) · 336 unverified), HuggingFace's Transformers: State-of-the-art Natural Language Processing (arXiv:1910.03771; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified), DeBERTa: Decoding-enhanced BERT with Disentangled Attention (arXiv:2006.03654; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 1 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 9 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
linzhiqiu/clip-flant5 mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large... (arXiv:2301.12597; that paper's own run record: 4 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 4 samples that ran constructed an object rather than computing a result); nothing checks that they implement this paper's method. report
skunkworksai/bakllava mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Visual Instruction Tuning (arXiv:2304.08485; that paper's own run record: 16 ran (of which 6 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 8 where Syntology's instrument failed) · 35 unverified); nothing checks that they implement this paper's method. report
sshh12/multi_token mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Visual Instruction Tuning (arXiv:2304.08485; that paper's own run record: 16 ran (of which 6 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 8 where Syntology's instrument failed) · 35 unverified); nothing checks that they implement this paper's method. report
x2fd/lvis-instruct4v mentioned on GitHubMITSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

9 samples harvested: 6 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 3 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 3 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

3ran · violated contract
1ran · our draft was wrong
2ran; the contract check could not run
3unverified

Licence: 8 of the 9 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. 5 samples whose link to this paper records no repository are listed under one of this paper's own repositories, because identical code was harvested from it for another paper. 3 samples are listed last, in a group without a repository. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: 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. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

LLaVA-VL/LLaVA-NeXT community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
should_only_save_mm_adapter llava/train/llava_trainer.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran · violated contract licence of this copy not recorded · 08787a49ef19e0aa · report fixture: signature args:Args/0/? · returned bool · level recorded, date not recorded
dinhvietcuong1996/icme25-inova community (archive-listed) 1 sample, 0 ran · 1 unverified
build_model_from_openai_state_dict in 42 papers llava/model/multimodal_encoder/dev_eva_clip/eva_clip/model.py unverified Apache-2.0 (permissive) · 275eb8fe6104bf3a · report
haotian-liu/LLaVA community (archive-listed) 4 samples, 2 ran (2 instrument-failed) · 2 unverified
split_to_even_chunks in 41 papers llava/train/llava_trainer.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran; the contract check could not run (RAISES) · metamorphic tier: well formed licence of this copy not recorded · 10893c4608c08075 · report fixture: signature indices:list/1/int, lengths:dict/0/?, num_chunks:int/0/int · returned list · level recorded, date not recorded
unpad_image in 72 papers
census: 4 implementations from 34 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 3, kind float, dtype float32
llava/model/llava_arch.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted licence of this copy not recorded · 55c32993da87759b · report fixture: signature tensor:Tensor/3/float, original_size:tuple/1/int · returned Tensor · level recorded, date not recorded
get_mm_adapter_state_maybe_zero_3 in 44 papers llava/train/llava_trainer.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own unverified licence of this copy not recorded · bb35e3ac741bb2c9 · report
maybe_zero_3 in 74 papers llava/train/llava_trainer.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own unverified licence of this copy not recorded · 735025744c1ab0cf · report
Repository not located 3 samples, 3 ran (1 instrument-failed) · 0 unverified
3 samples are identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
convert_to_custom_text_state_dict in 59 papers repository not located ran · violated contract licence of this copy not recorded · 7bff6f3b3cb560e7 · report fixture: signature state_dict:dict/0/? · returned dict · level recorded, date not recorded
get_cast_dtype in 67 papers repository not located ran · our draft was wrong licence of this copy not recorded · dcd422d66b0581d8 · report fixture: signature precision:str/0/str · returned dtype · level recorded, date not recorded
parse_score in 49 papers repository not located ran · violated contract fingerprinted licence of this copy not recorded · 763f6fe20bd52fbb · report fixture: signature review:str/0/str · returned list · level recorded, date not recorded

Tasks

Factual Inconsistency Detection in Chart CaptioningImage ClassificationReferring Expression ComprehensionReferring expression generationSpatial ReasoningVisual Question AnsweringVisual Question Answering (VQA)visual instruction following

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Factual Inconsistency Detection in Chart Captioning CHOCOLATE-FT LLaVA-1.5-13B Kendall's Tau-c 0.214 #4 of 5 Archive leaderboard report
Factual Inconsistency Detection in Chart Captioning CHOCOLATE-LLM LLaVA-1.5-13B Kendall's Tau-c 0.057 #5 of 5 Archive leaderboard report
Factual Inconsistency Detection in Chart Captioning CHOCOLATE-LVLM LLaVA-1.5-13B Kendall's Tau-c 0.002 #4 of 5 Archive leaderboard report
Image Classification ColonINST-v1 (Seen) LLaVA-v1.5 (w/ LoRA, w/ extra data) Accuray 93.33 #6 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Seen) LLaVA-v1.5 (w/ LoRA, w/o extra data) Accuray 92.97 #9 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) LLaVA-v1.5 (w/ LoRA, w/ extra data) Accuray 80.89 #2 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) LLaVA-v1.5 (w/ LoRA, w/o extra data) Accuray 79.10 #6 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) LLaVA-v1.5 (w/ LoRA, w/ extra data) Accuray 99.32 #2 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) LLaVA-v1.5 (w/ LoRA, w/o extra data) Accuray 98.58 #5 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) LLaVA-v1.5 (w/ LoRA, w/ extra data) Accuray 72.88 #9 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) LLaVA-v1.5 (w/ LoRA, w/o extra data) Accuray 70.38 #11 of 17 Archive leaderboard report
Spatial Reasoning 6-DoF SpatialBench LLaVA-1.5 Orientation-abs 25.8 #7 of 7 Archive leaderboard report
Spatial Reasoning 6-DoF SpatialBench LLaVA-1.5 Orientation-rel 28.3 #7 of 7 Archive leaderboard report
Spatial Reasoning 6-DoF SpatialBench LLaVA-1.5 Position-abs 24.5 #7 of 7 Archive leaderboard report
Spatial Reasoning 6-DoF SpatialBench LLaVA-1.5 Position-rel 30.9 #7 of 7 Archive leaderboard report
Spatial Reasoning 6-DoF SpatialBench LLaVA-1.5 Total 27.2 #7 of 7 Archive leaderboard report
Visual Question Answering BenchLMM LLaVA-1.5-13B GPT-3.5 score 55.53 #3 of 10 Archive leaderboard report
Visual Question Answering MM-Vet LLaVA-1.5-13B GPT-4 score 36.3±0.2 #145 of 231 Archive leaderboard report
Visual Question Answering MM-Vet LLaVA-1.5-13B Params 13B #145 of 231 Archive leaderboard report
Visual Question Answering MM-Vet LLaVA-1.5-7B GPT-4 score 31.1±0.2 #197 of 231 Archive leaderboard report
Visual Question Answering MM-Vet LLaVA-1.5-7B Params 7B #197 of 231 Archive leaderboard report
Visual Question Answering MM-Vet v2 LLaVA-v1.5-13B GPT-4 score 33.2±0.1 #21 of 24 Archive leaderboard report
Visual Question Answering MM-Vet v2 LLaVA-v1.5-13B Params 13B #21 of 24 Archive leaderboard report
Visual Question Answering MM-Vet v2 LLaVA-v1.5-7B GPT-4 score 28.3±0.2 #22 of 24 Archive leaderboard report
Visual Question Answering MM-Vet v2 LLaVA-v1.5-7B Params 7B #22 of 24 Archive leaderboard report
Visual Question Answering ViP-Bench LLaVA-1.5-13B (Coordinates) GPT-4 score (bbox) 47.1 #5 of 13 Archive leaderboard report
Visual Question Answering ViP-Bench LLaVA-1.5-13B (Visual Prompt) GPT-4 score (bbox) 41.8 #8 of 13 Archive leaderboard report
Visual Question Answering ViP-Bench LLaVA-1.5-13B (Visual Prompt) GPT-4 score (human) 42.9 #8 of 13 Archive leaderboard report
Visual Question Answering (VQA) AutoHallusion LLaVA-1.5 Overall Accuracy 44.5 #4 of 5 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval LLaVA-1.5 Abductive 47.91 #5 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval LLaVA-1.5 Analogical 24.31 #5 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval LLaVA-1.5 Deductive 30.94 #5 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval LLaVA-1.5 Overall score 32.62 #5 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval LLaVA-1.5 Params 13B #5 of 14 Archive leaderboard report
visual instruction following LLaVA-Bench LLaVA-v1.5-13B avg score 70.7 #4 of 8 Archive leaderboard report
visual instruction following LLaVA-Bench LLaVA-v1.5-7B avg score 63.4 #5 of 8 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections