Papers › Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding

Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding

5 Jun 2023arXiv:2306.02858archive 2025-07-28

Hang Zhang, Xin Li, Lidong Bing

We present Video-LLaMA a multi-modal framework that empowers Large Language Models (LLMs) with the capability of understanding both visual and auditory content in the video. Video-LLaMA bootstraps cross-modal training from the frozen pre-trained visual and audio encoders and the frozen LLMs. Unlike previous works that complement LLMs to process the visual or audio signals only, Video-LLaMA enables video comprehension by tackling two challenges: (1) capturing the temporal changes in visual scenes, (2) integrating audio-visual signals. To counter the first challenge, we propose a Video Q-former to assemble a pre-trained image encoder into our video encoder and introduce a video-to-text generation task to learn video-language correspondence. For the second challenge, we leverage ImageBind, a universal embedding model aligning multiple modalities, as the pre-trained audio encoder and introduce an Audio Q-former on top of ImageBind to learn reasonable auditory query embeddings for the LLM module. To align the output of both visual and audio encoders with LLM's embedding space, we first train Video-LLaMA on massive video/image-caption pairs and then tune our model with visual-instruction datasets of moderate amount but higher quality. We found Video-LLaMA shows the ability to perceive and comprehend video content and generate meaningful responses grounded in the visual and auditory information presented in the videos.

PaperPDFCodeCode Syntology ranMethod as code

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 25 code samples harvested from 3 repositories linked to this paper: 18 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 10 where Syntology's instrument failed) · 7 unverified. Where Syntology's instrument failed: 8 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: official repository: 9 samples from 1 repository, 7 ran (1 with no instrument failure; 6 where Syntology's instrument failed); community (archive-listed): 16 samples from 2 repositories, 11 ran (7 with no instrument failure; 4 where Syntology's instrument failed). 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 1 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 3 of the 18 samples that ran have no recorded date.

damo-nlp-sg/video-llama officialmentioned in papermentioned on GitHubpytorchBSD-3-ClauseSyntology: 9 samples, 7 ran (6 instrument-failed) · 2 unverified report
damo-nlp-sg/videollama2 mentioned on GitHubpytorchApache-2.0Syntology: 12 samples, 8 ran (3 instrument-failed) · 4 unverified report
damo-nlp-sg/videollama3 mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding (arXiv:2501.13106; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 5 where Syntology's instrument failed) · 6 unverified); nothing checks that they implement this paper's method. report
xinding-sys/StreamMind mentioned on GitHubpytorchSyntology: 4 samples, 3 ran (1 instrument-failed) · 1 unverified 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

25 samples harvested: 18 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 10 where Syntology's instrument failed) · 7 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.

1ran · violated contract
8ran · our draft was wrong
2ran; the contract check could not run
7ran
7unverified

Licence: 9 of the 25 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. “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.

damo-nlp-sg/video-llama official repository 9 samples, 7 ran (6 instrument-failed) · 2 unverified
apply_rotary_pos_emb in 393 papers
census: 14 implementations from 15 papers, 8 distinct outputs across 7 buckets, one shared input per bucket
video_llama/models/modeling_llama.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant BSD-3-Clause recorded; this copy not marked cleared · pointer only · 9b4dff79d5e6102c · report fixture: signature q:Tensor/4/float, k:Tensor/4/float, cos:Tensor/4/float, sin:Tensor/4/float, position_ids:Tensor/2/int · returned tuple · level recorded, date not recorded
disabled_train in 172 papers video_llama/models/blip2.py ran · violated contract BSD-3-Clause recorded; this copy not marked cleared · pointer only · 4cb732f513d69dfd · report fixture: signature self:callable/0/?, mode:bool/0/bool · returned SimpleModel · level recorded, date not recorded
getAttMap in 30 papers video_llama/common/gradcam.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted BSD-3-Clause recorded; this copy not marked cleared · pointer only · 7d214655954e2bc6 · report fixture: signature img:ndarray/3/float, attMap:ndarray/2/float, blur:bool/0/bool, overlap:bool/0/bool · returned ndarray · level recorded 2026-09-22
main_process in 45 papers video_llama/common/dist_utils.py ran · our draft was wrong BSD-3-Clause recorded; this copy not marked cleared · pointer only · 98589643273920ba · report fixture: signature func:callable/0/? · returned function · level recorded 2026-09-24
node_to_dict in 40 papers video_llama/common/config.py ran · our draft was wrong BSD-3-Clause recorded; this copy not marked cleared · pointer only · 791c72070c0b1cfb · report fixture: signature node:DictConfig/0/? · returned dict · level recorded 2026-09-24
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
video_llama/models/modeling_llama.py ran · our draft was wrong fingerprinted BSD-3-Clause recorded; this copy not marked cleared · pointer only · b99eea6376d1e212 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
tile in 109 papers
census: 5 implementations from 7 papers, 4 distinct outputs across 3 buckets, one shared input per bucket
video_llama/models/base_model.py ran · our draft was wrong fingerprinted BSD-3-Clause recorded; this copy not marked cleared · pointer only · cb33571427334815 · report fixture: signature x:Tensor/4/float, dim:int/0/int, n_tile:int/0/int · returned Tensor · level recorded 2026-09-24
all_gather_with_grad in 93 papers video_llama/models/base_model.py unverified BSD-3-Clause recorded; this copy not marked cleared · pointer only · 0ec9fc2025c16f65 · report
download_cached_file in 40 papers video_llama/common/dist_utils.py unverified BSD-3-Clause recorded; this copy not marked cleared · pointer only · 643539116e049080 · report
damo-nlp-sg/videollama2 community (archive-listed) 12 samples, 8 ran (3 instrument-failed) · 4 unverified
STCConnector videollama2/model/videollama2_arch.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 143af1a0e1416655 · report fixture: returned STCConnector · level recorded 2026-09-22
STCConnectorV35 videollama2/model/videollama2_arch.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · d25dbb06b89a36bd · report fixture: returned STCConnectorV35 · level recorded 2026-09-22
STPConnector videollama2/model/videollama2_arch.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · a1260df691548981 · report fixture: returned STPConnector · level recorded 2026-09-22
SpatialConv videollama2/model/videollama2_arch.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 75d05c81526b6ca0 · report fixture: returned SpatialConv · level recorded 2026-09-22
SpatialPool videollama2/model/videollama2_arch.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 27e6bfb755d3e81b · report fixture: returned SpatialPool · level recorded 2026-09-22
build_vision_projector in 23 papers videollama2/model/videollama2_arch.py ran · our draft was wrong Apache-2.0 (permissive) · 7aa936fbe4f3d6fe · report fixture: signature config:MockConfig/0/?, delay_load:bool/0/bool · returned Linear · level recorded 2026-09-22
load_mm_projector videollama2/model/videollama2_arch.py ran · our draft was wrong Apache-2.0 (permissive) · d8e25fe3aa75d2ce · report fixture: signature model_path:str/0/str, cache_dir:NoneType/0/?, token:NoneType/0/? · returned dict · level recorded 2026-09-22
parse_snapshot_folder videollama2/model/videollama2_arch.py ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 04129d1d261c322a · report fixture: signature repo_id:str/0/str, cache_dir:str/0/str, repo_type:str/0/str · returned str · level recorded 2026-09-22
CLIPVisionTower in 11 papers videollama2/model/videollama2_arch.py unverified Apache-2.0 (permissive) · 351648024debfdef · report
SiglipVisionTower videollama2/model/videollama2_arch.py unverified Apache-2.0 (permissive) · 93c7869a74572a70 · report
Videollama2MetaModel videollama2/model/videollama2_arch.py unverified Apache-2.0 (permissive) · e0abf348c09035c4 · report
build_vision_tower in 35 papers videollama2/model/videollama2_arch.py unverified Apache-2.0 (permissive) · f61dd94ea06b9f92 · report
xinding-sys/StreamMind community (archive-listed) 4 samples, 3 ran (1 instrument-failed) · 1 unverified
ClsNet streammind/model/multimodal_projector/builder.py ran Apache-2.0 (permissive) · 0d7afd7e8c6f79a7 · report fixture: returned ClsNet · level recorded 2026-09-22
STCConnector streammind/model/multimodal_projector/builder.py ran Apache-2.0 (permissive) · 034c07582f342e3f · report fixture: returned STCConnector · level recorded 2026-09-22
build_mlp in 74 papers streammind/model/multimodal_projector/builder.py ran · our draft was wrong Apache-2.0 (permissive) · d62ac690ebadd7fa · report fixture: signature depth:int/0/int, hidden_size:int/0/int, output_hidden_size:int/0/int · returned Sequential · level recorded 2026-09-22
MistralForCausalLM_cls streammind/model/multimodal_projector/builder.py unverified Apache-2.0 (permissive) · 44b8362633634eda · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

2 assemblies on this paper: 2 harvested, 0 generated.

harvested Assembly for Video Q-former (graph method: Video Q-former). 4 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified. The assembly has not been run end-to-end.

  1. STCConnector entry · level 1 · ran · 034c07582f342e3f
  2. ClsNet helper · level 1 · ran · 0d7afd7e8c6f79a7
  3. build_mlp helper · level 1 · ran · our draft was wrong · d62ac690ebadd7fa
  4. MistralForCausalLM_cls helper · level 0 · unverified · 44b8362633634eda

harvested Assembly for Video-LLaMA (graph method: Video-LLaMA). 14 bricks; individually on a fixture: 10 ran (of which 6 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 4 where Syntology's instrument failed) · 4 unverified. The assembly has not been run end-to-end.

  1. Videollama2MetaModel entry · level 0 · unverified · e0abf348c09035c4
  2. build_vision_projector helper · level 1 · ran · our draft was wrong · 7aa936fbe4f3d6fe
  3. build_vision_tower helper · level 0 · unverified · f61dd94ea06b9f92
  4. load_mm_projector helper · level 1 · ran · our draft was wrong · d8e25fe3aa75d2ce
  5. IdentityMap helper · level 1 · ran · 9f0f5d62445adabe
  6. STCConnector model · level 1 · ran · 143af1a0e1416655
  7. STCConnectorV35 helper · level 1 · ran · d25dbb06b89a36bd
  8. STPConnector helper · level 1 · ran · a1260df691548981
  9. SpatialConv helper · level 1 · ran · 75d05c81526b6ca0
  10. SpatialPool helper · level 1 · ran · 27e6bfb755d3e81b
  11. CLIPVisionTower model · level 0 · unverified · 351648024debfdef
  12. SiglipVisionTower model · level 0 · unverified · 93c7869a74572a70
  13. parse_snapshot_folder helper · level 1 · ran · our draft was wrong · 04129d1d261c322a
  14. build_mlp helper · level 1 · ran · our draft was wrong · d62ac690ebadd7fa

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2306.02858") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Language ModelingLanguage ModellingText GenerationVideo Question AnsweringVideo UnderstandingVideo-Text RetrievalVideo-based Generative Performance BenchmarkingVideo-based Generative Performance Benchmarking (Consistency)Video-based Generative Performance Benchmarking (Contextual Understanding)Video-based Generative Performance Benchmarking (Correctness of Information)Video-based Generative Performance Benchmarking (Detail Orientation))Video-based Generative Performance Benchmarking (Temporal Understanding)Zero-Shot Video Question Answer

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Question Answering MVBench VideoLLaMA Avg. 34.1 #19 of 22 Archive leaderboard report
Video-Text Retrieval Test-of-Time Video-LLAMA 2-Class Accuracy 88.33 #1 of 4 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA Consistency 1.79 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA Contextual Understanding 2.16 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA Correctness of Information 1.96 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA Detail Orientation 2.18 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA Temporal Understanding 1.82 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct Video LLaMA mean 1.98 #23 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking (Consistency) VideoInstruct Video LLaMA gpt-score 1.79 #18 of 18 Archive leaderboard report
Video-based Generative Performance Benchmarking (Contextual Understanding) VideoInstruct Video LLaMA gpt-score 2.16 #18 of 18 Archive leaderboard report
Video-based Generative Performance Benchmarking (Correctness of Information) VideoInstruct Video LLaMA gpt-score 1.96 #18 of 18 Archive leaderboard report
Video-based Generative Performance Benchmarking (Detail Orientation)) VideoInstruct Video LLaMA gpt-score 2.18 #18 of 18 Archive leaderboard report
Video-based Generative Performance Benchmarking (Temporal Understanding) VideoInstruct Video LLaMA gpt-score 1.82 #18 of 18 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA Video LLaMA Accuracy 12.4 #28 of 28 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA Video LLaMA Confidence Score 1.1 #28 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA Video LLaMA-7B Accuracy 29.6 #30 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA Video LLaMA-7B Confidence Score 1.8 #30 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA Video LLaMA-7B Accuracy 51.6 #27 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA Video LLaMA-7B Confidence Score 2.5 #27 of 28 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.

Methods

ALIGN

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