What are the hardware requirements for Cogagent Vqa Hf?

Cogagent Vqa Hf requires approximately 36.8 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Cogagent Vqa Hf and how large is it?

Cogagent Vqa Hf is developed by THUDM, a model with 18.3b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Cogagent Vqa Hf?

Cogagent Vqa Hf is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Cogagent Vqa Hf by THUDM — VRAM 36.8GB, 2K context

Name: Cogagent Vqa Hf
Author: THUDM

Cogagent Vqa Hf is an open-source language model by THUDM. Features: 18.3b LLM, VRAM: 36.8GB, Context: 2K, License: apache-2.0, LLM Explorer Score: 0.11.

Arxiv:2311.03079 Arxiv:2312.08914 Autotrain compatible Custom code En Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/THUDM/cogagent-vqa-hf

Cogagent Vqa Hf Parameters and Internals

Model Type

Visual Language Model

Use Cases

Areas:

Research, Academic, Commercial Applications

Applications:

GUI Agent, Visual Dialogue, Visual Grounding

Primary Use Cases:

GUI Agent, Visual Dialogue, Visual Grounding

Additional Notes

This version is focused on single-turn visual dialogue and VQA benchmarks.

Supported Languages

English (fluent)

Training Details

Data Sources:

MiniGPT-4, LLAVA, LRV-Instruction, LLaVAR, Shikra

Methodology:

Improved based on CogVLM

Input Output

Accepted Modalities:

Text, Image

LLM Name	Cogagent Vqa Hf
Repository 🤗	https://huggingface.co/THUDM/cogagent-vqa-hf
Model Size	18.3b
Required VRAM	36.8 GB
Updated	2025-07-28
Maintainer	THUDM
Model Files	5.0 GB: 1-of-8 5.0 GB: 2-of-8 5.0 GB: 3-of-8 5.0 GB: 4-of-8 5.0 GB: 5-of-8 5.0 GB: 6-of-8 5.0 GB: 7-of-8 1.8 GB: 8-of-8
Supported Languages	en
Model Architecture	CogAgentForCausalLM
License	apache-2.0
Context Length	2048
Model Max Length	2048
Transformers Version	4.36.0.dev0
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the Cogagent Vqa Hf capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Cogagent Vqa Hf

Best Alternatives	Context / RAM	Downloads	Likes
Cogagent Vqa Hf	2K / 36.8 GB	1442	49
Cogagent Chat Hf	2K / 36.8 GB	504	69
Cogagent Chat Hf	2K / 36.8 GB	277	68

Note: green Score (e.g. "73.2") means that the model is better than THUDM/cogagent-vqa-hf.