LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Yi VL 34B by 01-ai

 »  All LLMs  »  01-ai  »  Yi VL 34B   URL Share it on

  67   
  Arxiv:2403.04652   Conversational   Image-text-to-text   Llava   Pytorch   Region:us   Sharded
Model Card on HF 🤗: https://huggingface.co/01-ai/Yi-VL-34B 

Yi VL 34B Benchmarks

LLME Score: 0.17706
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Yi VL 34B (01-ai/Yi-VL-34B)

Yi VL 34B Parameters and Internals

Model Type 
Multimodal, Vision-Language
Use Cases 
Areas:
Research, Commercial applications
Applications:
Visual question answering, Image comprehension
Primary Use Cases:
Multi-round text-image conversations
Limitations:
Supports text-image conversations but not image-to-video., May hallucinate content not present in images., Resolution limited to 448x448.
Considerations:
Evaluate potential risks before adopting.
Additional Notes 
Supports image understanding at 448×448 resolution; ongoing limitations may affect certain applications.
Supported Languages 
English (proficient), Chinese (proficient)
Training Details 
Data Sources:
LAION-400M, CLLaVA, Flickr, VQAv2, RefCOCO, Visual7w, GQA, VizWiz VQA, TextCaps, OCR-VQA, Visual Genome, LAION GPT4V
Data Volume:
N/A
Methodology:
Three-stage training: image and text alignment using ViT and Yi LLM with datasets like LAION-400M and Visual Genome.
Training Time:
10 days for Yi-VL-34B, 3 days for Yi-VL-6B
Hardware Used:
128 NVIDIA A800 (80G) GPUs
Model Architecture:
Vision Transformer (ViT) initialized with CLIP ViT-H/14, a projection module, and Yi LLMs.
Input Output 
Input Format:
Text and images
Accepted Modalities:
text, image
Output Format:
Text
LLM NameYi VL 34B
Repository 🤗https://huggingface.co/01-ai/Yi-VL-34B 
Model Size34b
Required VRAM70.3 GB
Updated2025-02-05
Maintainer01-ai
Model Typellava
Model Files  10.0 GB: 1-of-8   9.9 GB: 2-of-8   9.8 GB: 3-of-8   9.8 GB: 4-of-8   9.8 GB: 5-of-8   9.9 GB: 6-of-8   10.0 GB: 7-of-8   1.1 GB: 8-of-8
Model ArchitectureLlavaLlamaForCausalLM
Licenseapache-2.0
Context Length4096
Model Max Length4096
Transformers Version4.34.0
Tokenizer ClassLlamaTokenizer
Padding Token<unk>
Vocabulary Size64000
Torch Data Typebfloat16

Best Alternatives to Yi VL 34B

Best Alternatives
Context / RAM
Downloads
Likes
Llava V1.6 34B4K / 69.9 GB13096343
HuatuoGPT Vision 34B4K / 69.9 GB56217
LLaVA NeXT Video 34B4K / 69.9 GB5716
LLaVA NeXT Video 34B DPO4K / 69.9 GB2810
Llava V1.6 34B Finetune4K / 69.9 GB94
Hf Llava V1.6 34B4K / 69.9 GB116
Note: green Score (e.g. "73.2") means that the model is better than 01-ai/Yi-VL-34B.

Rank the Yi VL 34B Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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  

What open-source LLMs or SLMs are you in search of? 42577 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227
  Support LLM Explorer