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

SeaLLMs V3 7B Chat by SeaLLMs

 »  All LLMs  »  SeaLLMs  »  SeaLLMs V3 7B Chat   URL Share it on

  31889   
  Arxiv:2306.05179   Arxiv:2407.19672   Autotrain compatible   Conversational   En   Endpoints compatible   Id   Jv   Km   Lo   Ms   Multilingual   My   Qwen2   Region:us   Safetensors   Sea   Sharded   Ta   Tensorflow   Th   Tl   Vi   Zh
Model Card on HF 🤗: https://huggingface.co/SeaLLMs/SeaLLMs-v3-7B-Chat 

SeaLLMs V3 7B Chat Benchmarks

MMLU Pro: 32.16
GPQA: 6.49
MUSR: 10.47
BBH: 33.8
IFEval: 43.77 vs 88 (so35)-50.3%
MATH Lvl 5: 16.62
LLME Score: 0.36895
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
SeaLLMs V3 7B Chat (SeaLLMs/SeaLLMs-v3-7B-Chat)

SeaLLMs V3 7B Chat Parameters and Internals

Model Type 
multilingual, chat
Use Cases 
Areas:
research, commercial applications
Applications:
task automation in Southeast Asian languages
Primary Use Cases:
Human instruction following, Multilingual tasks, Translation
Limitations:
Risk of inaccurate or biased outputs
Considerations:
Use with caution, abide by local regulations.
Additional Notes 
Tailored for Southeast Asian language instructions and tasks.
Supported Languages 
en (advanced), zh (advanced), id (advanced), vi (advanced), th (advanced), ms (advanced), tl (advanced), ta (advanced), jv (advanced), lo (advanced), km (advanced), my (advanced)
Training Details 
Data Sources:
Southeast Asian languages data
Methodology:
Fine-tuning for chat with instruction-following enhancement
Model Architecture:
Large Language Model
Safety Evaluation 
Methodologies:
red teaming, safety fine-tuning
Findings:
reduced hallucination, safe response generation
Risk Categories:
misinformation, bias
Ethical Considerations:
Considerations for local governance and regulations.
Responsible Ai Considerations 
Fairness:
Efforts to ensure fairness across various Southeast Asian languages.
Transparency:
Open-source release with detailed evaluation.
Accountability:
Users should perform own evaluations and adhere to local laws.
Mitigation Strategies:
Red teaming and safety fine-tuning.
Input Output 
Input Format:
Chat-based prompts in supported languages
Accepted Modalities:
text
Output Format:
Text responses in similar language or translated as required
Performance Tips:
For resource-limited settings, use smaller versions.
LLM NameSeaLLMs V3 7B Chat
Repository 🤗https://huggingface.co/SeaLLMs/SeaLLMs-v3-7B-Chat 
Model Size7b
Required VRAM15.2 GB
Updated2024-12-21
MaintainerSeaLLMs
Model Typeqwen2
Model Files  4.9 GB: 1-of-4   4.9 GB: 2-of-4   4.3 GB: 3-of-4   1.1 GB: 4-of-4
Supported Languagesen zh id vi th ms tl ta jv lo km my
Model ArchitectureQwen2ForCausalLM
Licenseother
Context Length131072
Model Max Length131072
Transformers Version4.41.2
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size152064
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to SeaLLMs V3 7B Chat

Best Alternatives
Context / RAM
Downloads
Likes
Qwen2.5 7B128K / 15.4 GB11579484
Cybertron V4 Qw7B UNAMGS128K / 15.2 GB22825
Qwen2 7B128K / 15.4 GB29669147
Lexora Medium 7B128K / 15.2 GB90615
Homer V1.0 Qwen2.5 7B128K / 15.2 GB1520
Gte Qwen2 7B Instruct128K / 30.5 GB41284235
Cybertron V4 Qw7B MGS128K / 15.2 GB27411
Qwen2.5 7B Olm V1.0128K / 15.2 GB302
Rombos LLM V2.5 Qwen 7B128K / 15.2 GB30015
Qwen2 7B Instruct V0.1128K / 15.2 GB409681
Note: green Score (e.g. "73.2") means that the model is better than SeaLLMs/SeaLLMs-v3-7B-Chat.

Rank the SeaLLMs V3 7B Chat 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? 40013 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 v20241217
  Support LLM Explorer