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

JEJUMA 002 by JEJUMA

 »  All LLMs  »  JEJUMA  »  JEJUMA 002   URL Share it on

  13   
  Conversational   Dialect   Instruct   Ko   Language   Llama   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/JEJUMA/JEJUMA-002 

JEJUMA 002 Benchmarks

LLME Score: 0.20969
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
JEJUMA 002 (JEJUMA/JEJUMA-002)

JEJUMA 002 Parameters and Internals

Model Type 
Text generation
Use Cases 
Areas:
Language preservation, Dialect translation and detection
Applications:
Maintaining regional dialects, Dialect translation services
Primary Use Cases:
Converting dialect to standard Korean and vice versa, Detecting regional dialect types, Identifying dialect gaps for further study
Limitations:
Limited to trained dialects, May not handle all regional nuances accurately
Considerations:
Ensure proper testing in varied dialects before deployment.
Additional Notes 
Data for training is available upon request via provided email address.
Supported Languages 
Korean (Various dialects, including Jeju, Chungcheong, Gangwon, Gyeongsang, and Jeonla)
Training Details 
Data Sources:
Regional dialect and standard Korean pairs
Data Volume:
200 million data pairs
Methodology:
Fine-tuning using LoRA technique
Training Time:
1 epoch
Hardware Used:
GPUs for training, Included language models: Upstage Solar, Naver HCX, GPT-4o
Model Architecture:
Derived from Llama3.1
Input Output 
Input Format:
Text sentence or phrase
Output Format:
Converted text or identified dialect category
Performance Tips:
Ensure correct identification of input dialect for accurate conversion.
LLM NameJEJUMA 002
Repository 🤗https://huggingface.co/JEJUMA/JEJUMA-002 
Model Size8b
Required VRAM16.1 GB
Updated2025-02-05
MaintainerJEJUMA
Model Typellama
Instruction-BasedYes
Model Files  5.0 GB: 1-of-4   5.0 GB: 2-of-4   4.9 GB: 3-of-4   1.2 GB: 4-of-4
Supported Languagesko
Model ArchitectureLlamaForCausalLM
Licensemit
Context Length131072
Model Max Length131072
Transformers Version4.43.3
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|eot_id|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to JEJUMA 002

Best Alternatives
Context / RAM
Downloads
Likes
...a 3 8B Instruct Gradient 1048K1024K / 16.1 GB6623678
161024K / 16.1 GB1690
Because Im Bored Nsfw11024K / 16.1 GB661
Mpasila Viking 8B1024K / 16.1 GB590
121024K / 16.1 GB600
MrRoboto ProLong 8B V4b1024K / 16.1 GB1070
MrRoboto ProLong 8B V1a1024K / 16.1 GB1080
MrRoboto ProLong 8B V2a1024K / 16.1 GB1020
MrRoboto ProLong 8B V4c1024K / 16.1 GB870
8B Unaligned BASE V2b1024K / 16.1 GB980
Note: green Score (e.g. "73.2") means that the model is better than JEJUMA/JEJUMA-002.

Rank the JEJUMA 002 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