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

GALAXY XB V.03 by TeeZee

 »  All LLMs  »  TeeZee  »  GALAXY XB V.03   URL Share it on

  72   
  Autotrain compatible   Endpoints compatible   Llama   Model-index   Moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/TeeZee/GALAXY-XB-v.03 

GALAXY XB V.03 Benchmarks

ARC: 61.77 vs 96.7 (so35)-36.1%
HellaSwag: 83.59 vs 95.3 (gpt4)-12.3%
MMLU: 64.55 vs 88.3 (so35)-26.9%
TruthfulQA: 44.19 vs 59 (gpt4)-25.1%
WinoGrande: 81.06 vs 87.5 (gpt4)-7.4%
GSM8K: 45.03 vs 96.4 (so35)-53.3%
LLME Score: 0.20316
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
🌟 Advertise your project 🚀

GALAXY XB V.03 Parameters and Internals

Model Type 
text-generation
Additional Notes 
Small quality loss observed when compared to the base model as described in the DUS paper. This version is a merge of 12 layers removed from both models. Future steps involve finetuning on specific datasets to recover performance post-merge.
Training Details 
Data Sources:
openorca dataset, DPO datasets used by SOLAR
Methodology:
DUS - 12 layers removed from both models
LLM NameGALAXY XB V.03
Repository 🤗https://huggingface.co/TeeZee/GALAXY-XB-v.03 
Model Size16b
Required VRAM31.9 GB
Updated2024-12-02
MaintainerTeeZee
Model Typellama
Model Files  10.0 GB: 1-of-4   9.9 GB: 2-of-4   10.0 GB: 3-of-4   2.0 GB: 4-of-4
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length4096
Model Max Length4096
Transformers Version4.35.2
Tokenizer ClassLlamaTokenizer
Vocabulary Size32000
Torch Data Typebfloat16
GALAXY XB V.03 (TeeZee/GALAXY-XB-v.03)

Best Alternatives to GALAXY XB V.03

Best Alternatives
Context / RAM
Downloads
Likes
Phi 3.5 Mini Investigator 16B128K / 7.6 GB150
...o Sft Ties Post Merge Auto DPO8K / 141.2 GB60
Nanbeige2 16B Chat4K / 31.6 GB11990
...ALAXY V03 Slimorca 1 Epoch 50k4K / 31.8 GB900
...ca 1 Epoch 50k DPO 1 Epoch 30k4K / 31.8 GB680
Nanbeige 16B Base Llama4K / 31.6 GB12763
FusionNet SOLAR4K / 31.9 GB13031
Llama 2 16B Nastychat4K / 32.4 GB13008
Vinallama 16B Chat Franken4K / 31.6 GB111
Nanbeige 16B Chat4K / 31.6 GB2412
Note: green Score (e.g. "73.2") means that the model is better than TeeZee/GALAXY-XB-v.03.

Rank the GALAXY XB V.03 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? 38765 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 v20241124
  Support LLM Explorer