SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2411.14522 · 2024

GMAI-VL & GMAI-VL-5.5M: A Large Vision-Language Model and A Comprehensive Multimodal Dataset Towards General Medical AI

arXiv · PDF · Open in the Atlas

Code that ran

10 of 10 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 9 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/2411.14522.json · badge: /api/ran/2411.14522.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
uni-medical/gmai-vl canonical (classifier label)
classifier basis: census
10 ran · 0 no recorded run
FunctionStatusWhere it lives
build_model Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/xtuner/apis/model.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from echo840/LIRA, is pointer only (licence: NONE) · get_code("dcdc7ef233f3e1e8")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from echo840/LIRA, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

convert_state_dict_to_hf Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/xtuner/model/llava.py
code served (permissive licence) · get_code("f8035a195671164a")
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, through the batch API) 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

torch.manual_seed(0)

# The function takes a state_dict (dictionary of tensors/values)
# and a mapping dictionary that maps old key patterns to new key patterns.
# It iterates through state_dict items, skips those ending in '.inv_freq',
# and applies the mapping to rename keys.

state_dict = {
    'layer.0.weight': torch.randn(4, 3),
    'layer.0.bias': torch.randn(4),
    'layer.1.weight': torch.randn(5, 4),
    'layer.1.bias': torch.randn(5),
    'rope.inv_freq': torch.randn(2),  # This will be skipped
    'attention.query.weight': torch.randn(3, 2),
}

mapping = {
    'layer': 'transformer.layer',
    'attention.query': 'attention.q_proj',
}

ARGS = {
    'state_dict': state_dict,
    'mapping': mapping,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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
create_reference_model Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/xtuner/model/dpo.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from echo840/LIRA, is pointer only (licence: NONE) · get_code("269c10dd16c744e8")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from echo840/LIRA, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

download_model_from_hub Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
uni-medical/gmai-vl/xtuner/_lite/auto.py
code served (permissive licence) · get_code("444b8623d0de4b5a")
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, through the batch API) and returned str. 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 os
import tempfile

# Create a temporary directory to serve as a local model path
temp_model_dir = tempfile.mkdtemp()

# Create a minimal model directory structure
os.makedirs(temp_model_dir, exist_ok=True)
with open(os.path.join(temp_model_dir, "model.txt"), "w") as f:
    f.write("dummy model")

ARGS = {
    "model_name_or_path": temp_model_dir,
    "from_hub": "huggingface",
    "cache_dir": None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
huggingface_hub, modelscope — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
find_all_linear_names Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/xtuner/model/utils.py
code served (permissive licence) · get_code("53281c424cf5891f")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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
from torch import nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
        self.nested = nn.Sequential(
            nn.Linear(2, 5),
            nn.ReLU(),
            nn.Linear(5, 1)
        )
        self.lm_head = nn.Linear(1, 1)
        self.output_layer = nn.Linear(1, 1)

model = SimpleModel()

ARGS = {
    "model": model
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
get_main_modality Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/demo/gmai_vl_demo/static/get_modality_num.py
code served (permissive licence) · get_code("5d8fd42d109d9939")
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, through the batch API) and returned str. 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
import numpy as np

# Setup for get_main_modality function
ARGS = {
    'modality': 'MR-FLAIR'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
get_peft_model_state_dict Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
uni-medical/gmai-vl/xtuner/model/utils.py
code served (permissive licence) · get_code("9356d2ed9b73134b")
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, through the batch API) 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
import torch.nn as nn
from peft import PeftType

torch.manual_seed(0)

# Create a minimal mock PeftConfig
class MockPeftConfig:
    def __init__(self, peft_type, bias='none'):
        self.peft_type = peft_type
        self.bias = bias

# Create a minimal mock model with required attributes
class MockPeftModel(nn.Module):
    def __init__(self):
        super().__init__()
        # Create some dummy parameters with lora_ in the name
        self.linear = nn.Linear(4, 3)
        self.lora_A = nn.Parameter(torch.randn(4, 2))
        self.lora_B = nn.Parameter(torch.randn(2, 3))
        
        # Setup peft_config
        self.peft_config = {
            'default': MockPeftConfig(peft_type=PeftType.LORA, bias='none')
        }
        
        # Setup modules_to_save
        self.modules_to_save = None

# Create mock model instance
model = MockPeftModel()

# Create state dict with lora parameters and bias
state_dict = {
    'lora_A': torch.randn(4, 2),
    'lora_B': torch.randn(2, 3),
    'linear.weight': torch.randn(3, 4),
    'linear.bias': torch.randn(3),
}

# Define arguments for the function
ARGS = {
    'model': model,
    'state_dict': state_dict,
    'adapter_name': 'default'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
peft, 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
load_json Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/demo/gmai_vl_demo/static/get_modality_num.py
code served (permissive licence) · get_code("501a90fb110e541b")
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, through the batch API) 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 json
from pathlib import Path
import tempfile

# Create a temporary directory and file for testing
temp_dir = tempfile.TemporaryDirectory()
temp_path = Path(temp_dir.name) / "test.json"

# Write a valid JSON file
test_data = {"key": "value", "number": 42, "list": [1, 2, 3]}
with temp_path.open('w', encoding='utf-8') as f:
    json.dump(test_data, f)

ARGS = {
    "path": temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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
parse_samed2d_name Ran
ran; our contract draft was wrong, not the code
uni-medical/gmai-vl/demo/gmai_vl_demo/static/get_modality_num.py
code served (permissive licence) · get_code("5fa73aa676ca240b")
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, through the batch API) 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
import numpy as np

# Construct arguments for parse_samed2d_name
# The function takes a single parameter: path (str)
# It extracts the filename, checks for .png extension, splits by '--', and parses modality

ARGS = {
    'path': 'datasets/sample_data/MR_T1--dataset1--patient001--slice042.png'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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
reduce_mean Ran
ran, honoured the contract we drafted
uni-medical/gmai-vl/xtuner/model/reward.py
code served (permissive licence) · get_code("a0a407c2a1c1f7d3")
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, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 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 torch.distributed as dist

torch.manual_seed(0)

# Create a small tensor argument
tensor = torch.randn(2, 4)

# The function only uses tensor.clone() and tensor.div_() methods,
# so any tensor will work. We don't need to initialize distributed
# training since the function checks dist.is_initialized() first.

ARGS = {
    "tensor": tensor
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 2, dtype family float, shape (2, 4) — shape (2, 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

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 6 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 6 later) and 4 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
reduce_mean 5 other papers
4 dated earlier, 1 later
2010.15831 (2020-10)
RelationNet++: Bridging Visual Representations for Object Detection via Transformer Decoder
2308.05581 (2023-08)
Category Feature Transformer for Semantic Segmentation
2402.02500 (2024-02)
Point Cloud Matters: Rethinking the Impact of Different Observation Spaces on Robot Learning
and 2 more not named here
build_model 4 other papers
2 dated earlier, 2 later
2401.10229 (2024-01)
OMG-Seg: Is One Model Good Enough For All Segmentation?
2406.19389 (2024-06)
OMG-LLaVA : Bridging Image-level, Object-level, Pixel-level Reasoning and Understanding
2505.14640 (2025-05)
VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
and 1 more not named here
convert_state_dict_to_hf 3 other papers
3 later
2505.14640 (2025-05)
VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
2507.06272 (2025-07)
LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region Assistance
2508.03284 (2025-08)
ToolVQA: A Dataset for Multi-step Reasoning VQA with External Tools
create_reference_model 2 other papers
2 later
2505.14640 (2025-05)
VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
2507.06272 (2025-07)
LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region Assistance
find_all_linear_names 2 other papers
1 dated earlier, 1 later
2106.09685 (2021-06)
LoRA: Low-Rank Adaptation of Large Language Models
2505.14640 (2025-05)
VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
get_peft_model_state_dict 1 other paper
1 later
2505.14640 (2025-05)
VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation

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

Despite significant advancements in general AI, its effectiveness in the medical domain is limited by the lack of specialized medical knowledge. To address this, we formulate GMAI-VL-5.5M, a multimodal medical dataset created by converting hundreds of specialized medical datasets with various annotations into high-quality image-text pairs. This dataset offers comprehensive task coverage, diverse modalities, and rich image-text data. Building upon this dataset, we develop GMAI-VL, a general medical vision-language model, with a three-stage training strategy that enhances the integration of visual and textual information. This approach significantly improves the model's ability to process multimodal data, supporting accurate diagnoses and clinical decision-making. Experiments show that GMAI-VL achieves state-of-the-art performance across various multimodal medical tasks, including visual question answering and medical image diagnosis.

For agents

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

get_harvested_code_for_paper("2411.14522")
get_code_for_paper("2411.14522")
have("2411.14522")

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

curl https://syntology.ai/api/ran/2411.14522.json

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

[![Syntology run record](https://syntology.ai/api/ran/2411.14522.svg)](https://syntology.ai/paper/2411.14522)

Connect an agent — have() is free.