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

Jais Adapted 13B by inceptionai

 »  All LLMs  »  inceptionai  »  Jais Adapted 13B   URL Share it on

  31   
  Arxiv:2307.09288   Arxiv:2308.16149   Arxiv:2402.12840   Ar   Arabic Base model:finetune:meta-llama... Base model:meta-llama/llama-2-...   Decoder   En   English   Jais-family   Llama   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/inceptionai/jais-adapted-13b 

Jais Adapted 13B Benchmarks

LLME Score: 0.20963
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Jais Adapted 13B (inceptionai/jais-adapted-13b)

Jais Adapted 13B Parameters and Internals

Model Type 
LLM, Decoder, causal-lm
Use Cases 
Areas:
Research, Commercial applications
Applications:
Natural language understanding and generation, Mechanistic interpretability, Sentiment analysis, Summarization
Primary Use Cases:
Research purposes for Arabic NLP, Commercial chat applications, Sentiment analysis, Academic research
Limitations:
Prohibited from generating harmful content, Sensitive information handling, Generalization across non-supported languages, High-stakes decision making
Considerations:
Efforts to ensure cultural adaptation and diverse topic range in fine-tuning datasets.
Additional Notes 
Techniques used for Arabic model augmentation applicable to other low-resource languages.
Supported Languages 
Arabic (MSA) (Strong capabilities), English (Strong capabilities)
Training Details 
Data Sources:
Web pages, Wikipedia articles, News articles, Social network content, Code data, Books, Scientific papers, Synthetic data (English to Arabic translations)
Data Volume:
Up to 1.6 Trillion tokens
Methodology:
Documents packed with EOS tokens for pre-training and frozen backbone during adapted pre-training. Instructional fine-tuning for chat models.
Context Length:
16384
Hardware Used:
Condor Galaxy supercomputer, 64 Cerebras CS-2 Wafer-Scale Engines
Model Architecture:
Auto-regressive Transformer-based, decoder-only architecture with support for long context lengths.
Responsible Ai Considerations 
Mitigation Strategies:
Minimized biases; AI assistant role limited to Arabic and English for fine-tuned models.
Input Output 
Input Format:
Text inputs
Accepted Modalities:
text
Output Format:
Generated text
LLM NameJais Adapted 13B
Repository 🤗https://huggingface.co/inceptionai/jais-adapted-13b 
Base Model(s)  Llama 2 13B   meta-llama/Llama-2-13b
Model Size13b
Required VRAM53.5 GB
Updated2024-11-04
Maintainerinceptionai
Model Typellama
Model Files  4.8 GB: 1-of-11   4.8 GB: 2-of-11   4.8 GB: 3-of-11   5.0 GB: 4-of-11   5.0 GB: 5-of-11   5.0 GB: 6-of-11   5.0 GB: 7-of-11   4.8 GB: 8-of-11   4.8 GB: 9-of-11   4.8 GB: 10-of-11   4.7 GB: 11-of-11
Supported Languagesar en
Gated ModelYes
Model ArchitectureLlamaForCausalLM
Licenseproprietary
Context Length4096
Model Max Length4096
Transformers Version4.38.2
Tokenizer ClassLlamaTokenizer
Vocabulary Size64000
Torch Data Typefloat32

Best Alternatives to Jais Adapted 13B

Best Alternatives
Context / RAM
Downloads
Likes
Yarn Llama 2 13B 128K128K / 26 GB4778113
Luminaura RP 13B128K / 26 GB90
Agent Llama2 13B 80K80K / 26.4 GB140
Chat Llama2 13B 80K80K / 52.8 GB110
LongAlign 13B 64K64K / 26 GB1713
LongAlign 13B 64K Base64K / 26 GB143
Yarn Llama 2 13B 64K64K / 26 GB466017
Openbuddy Llama2 13B V15p1 64K64K / 26.1 GB64
Openbuddy Llama2 13b64k V1564K / 26.1 GB131
Airoboros L2 13B 2.1 YaRN 64K64K / 26 GB117

Rank the Jais Adapted 13B 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? 42565 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