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

Cosmosage V3 by Tijmen2

 »  All LLMs  »  Tijmen2  »  Cosmosage V3   URL Share it on

  21   
  Arxiv:2407.04420   Autotrain compatible Base model:finetune:meta-llama... Base model:meta-llama/meta-lla...   Conversational   Cosmology   Dataset:teknium/openhermes-2.5   En   Endpoints compatible   Llama   Physics   Pytorch   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/Tijmen2/cosmosage-v3 

Cosmosage V3 Benchmarks

MMLU Pro: 16.51
GPQA: 4.36
MUSR: 10.69
BBH: 22.69
IFEval: 44.82 vs 88 (so35)-49.1%
MATH Lvl 5: 5.06
LLME Score: 0.2287
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Cosmosage V3 (Tijmen2/cosmosage-v3)

Cosmosage V3 Parameters and Internals

Model Type 
text-generation
Additional Notes 
cosmosage-v3 is a natural-language cosmology assistant that excels in Q&A mode.
Training Details 
Data Sources:
cosmology papers and textbooks, synthetically-generated question-answer pairs, OpenHermes 2.5 dataset
Methodology:
Continued pretraining and fine-tuning on synthetic data
Hardware Used:
4xA100 (40 GB) GPUs at Gadi supercomputer
LLM NameCosmosage V3
Repository 🤗https://huggingface.co/Tijmen2/cosmosage-v3 
Base Model(s)  Meta Llama 3 8B   meta-llama/Meta-Llama-3-8B
Model Size8b
Required VRAM16.1 GB
Updated2025-05-20
MaintainerTijmen2
Model Typellama
Model Files  16.1 GB   16.1 GB
Supported Languagesen
Model ArchitectureLlamaForCausalLM
Licensemit
Context Length8192
Model Max Length8192
Transformers Version4.41.1
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|end_of_text|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Cosmosage V3

Best Alternatives
Context / RAM
Downloads
Likes
...otron 8B UltraLong 4M Instruct4192K / 32.1 GB4677105
UltraLong Thinking4192K / 16.1 GB2482
...a 3.1 8B UltraLong 4M Instruct4192K / 32.1 GB17624
...a 3.1 8B UltraLong 2M Instruct2096K / 32.1 GB8759
...otron 8B UltraLong 2M Instruct2096K / 32.1 GB22615
Zero Llama 3.1 8B Beta61048K / 16.1 GB5271
...otron 8B UltraLong 1M Instruct1048K / 32.1 GB306941
...a 3.1 8B UltraLong 1M Instruct1048K / 32.1 GB138729
...dger Nu Llama 3.1 8B UltraLong1048K / 16.2 GB432
....1 1million Ctx Dark Planet 8B1048K / 32.3 GB322
Note: green Score (e.g. "73.2") means that the model is better than Tijmen2/cosmosage-v3.

Rank the Cosmosage V3 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? 47466 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