SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2407.00634 · 2024

Tarsier: Recipes for Training and Evaluating Large Video Description Models

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2407.00634.json · badge: /api/ran/2407.00634.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
bytedance/tarsier canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
add_split_tokens Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
bytedance/tarsier/models/modeling_tarsier.py
code served (permissive licence) · get_code("ad08899b0ff95e16")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 3, dtype family float, shape (2, 21, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import math

torch.manual_seed(0)

# Analyze the function:
# - image_features: shape (num_images, num_image_patches, embed_dim)
#   Used with .shape, .view(), torch.cat()
# - image_newline_embed: shape (1, 1, embed_dim)
#   Used with .expand()
# - image_new_embed: shape (1, 1, embed_dim)
#   Used with .expand()

# The function expects num_image_patches to be a perfect square
# Let's use: num_images=2, num_image_patches=16 (4x4), embed_dim=4

num_images = 2
num_image_patches = 16  # 4x4 grid
embed_dim = 4

# Create image_features: (2, 16, 4)
image_features = torch.randn(num_images, num_image_patches, embed_dim)

# Create image_newline_embed: (1, 1, 4)
image_newline_embed = torch.randn(1, 1, embed_dim)

# Create image_new_embed: (1, 1, 4)
image_new_embed = torch.randn(1, 1, embed_dim)

ARGS = {
    "image_features": image_features,
    "image_newline_embed": image_newline_embed,
    "image_new_embed": image_new_embed,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 21, 4) — shape (2, 21, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
extract_item_for_eval Ran
ran; our contract draft was wrong, not the code
bytedance/tarsier/evaluation/evaluate.py
code served (permissive licence) · get_code("44a7aa16432213cb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

# Construct a valid data_dict argument for extract_item_for_eval
data_dict = {
    'messages': [
        {
            'role': 'user',
            'content': [
                {
                    'type': 'text',
                    'text': 'What is the capital of France?'
                }
            ]
        },
        {
            'role': 'assistant',
            'content': [
                {
                    'type': 'text',
                    'text': 'Paris',
                    'reference': 'Paris is the capital of France.'
                }
            ]
        }
    ],
    'dataset': 'test_dataset',
    'idx': 42,
    'vid': 'video_001',
    'extra_info': {
        'events': ['event1', 'event2'],
        'metadata': 'some_metadata'
    },
    'is_hard': False
}

ARGS = {
    'data_dict': data_dict
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
add_split_tokens 1 other paper
1 later
2501.07888 (2025-01)
Tarsier2: Advancing Large Vision-Language Models from Detailed Video Description to Comprehensive Video Understanding
extract_item_for_eval 1 other paper
1 later
2501.07888 (2025-01)
Tarsier2: Advancing Large Vision-Language Models from Detailed Video Description to Comprehensive Video Understanding

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Generating fine-grained video descriptions is a fundamental challenge in video understanding. In this work, we introduce Tarsier, a family of large-scale video-language models designed to generate high-quality video descriptions. Tarsier employs CLIP-ViT to encode frames separately and then uses an LLM to model temporal relationships. Despite its simple architecture, we demonstrate that with a meticulously designed two-stage training procedure, the Tarsier models exhibit substantially stronger video description capabilities than any existing open-source model, showing a $+51.4\%$ advantage in human side-by-side evaluation over the strongest model. Additionally, they are comparable to state-of-the-art proprietary models, with a $+12.3\%$ advantage against GPT-4V and a $-6.7\%$ disadvantage against Gemini 1.5 Pro. When upgraded to Tarsier2 by building upon SigLIP and Qwen2-7B, it further improves significantly with a $+4.8\%$ advantage against GPT-4o. Besides video description, Tarsier proves to be a versatile generalist model, achieving new state-of-the-art results across nine public benchmarks, including multi-choice VQA, open-ended VQA, and zero-shot video captioning. Our second contribution is the introduction of a new benchmark -- DREAM-1K (https://tarsier-vlm.github.io/) for evaluating video description models, consisting of a new challenging dataset featuring videos from diverse sources and varying complexity, along with an automatic method specifically designed to assess the quality of fine-grained video descriptions. We make our models and evaluation benchmark publicly available at https://github.com/bytedance/tarsier.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2407.00634")
get_code_for_paper("2407.00634")
have("2407.00634")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2407.00634.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2407.00634.svg)](https://syntology.ai/paper/2407.00634)

Connect an agent — have() is free.