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

Fireball 12B V1.0 by EpistemeAI

 »  All LLMs  »  EpistemeAI  »  Fireball 12B V1.0   URL Share it on

  18   
  Autotrain compatible Base model:epistemeai/fireball... Base model:finetune:epistemeai...   En   Endpoints compatible   Mistral   Pytorch   Region:us   Sharded   Trl   Unsloth
Model Card on HF 🤗: https://huggingface.co/EpistemeAI/Fireball-12B-v1.0 

Fireball 12B V1.0 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").
Fireball 12B V1.0 (EpistemeAI/Fireball-12B-v1.0)

Fireball 12B V1.0 Parameters and Internals

Model Type 
text-generation, transformers
Additional Notes 
Fireball-12B-V1.0 is a heavily fine-tuned Mistral-Nemo-Base-2407 LLM, outperforming existing 12B parameter models. Requires smaller temperatures (e.g., 0.3) for output generation.
Training Details 
Data Sources:
candenizkocak/code-alpaca-297k, yahma/alpaca-cleaned
Methodology:
Supervised fine-tuning with ORPO method
Context Length:
128000
Model Architecture:
transformer, Layers: 40, Dim: 5,120, Head dim: 128, Hidden dim: 14,436, Activation Function: SwiGLU, Number of heads: 32, Number of kv-heads: 8 (GQA), Vocabulary size: 128k. Rotary embeddings (theta = 1M)
LLM NameFireball 12B V1.0
Repository 🤗https://huggingface.co/EpistemeAI/Fireball-12B-v1.0 
Base Model(s)  Fireball 12B   EpistemeAI/Fireball-12B
Model Size12b
Required VRAM24.5 GB
Updated2025-02-22
MaintainerEpistemeAI
Model Typemistral
Model Files  4.9 GB: 1-of-5   4.9 GB: 2-of-5   4.9 GB: 3-of-5   4.9 GB: 4-of-5   4.9 GB: 5-of-5
Supported Languagesen
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length1024000
Model Max Length1024000
Transformers Version4.44.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<pad>
Vocabulary Size131072
Torch Data Typefloat16

Best Alternatives to Fireball 12B V1.0

Best Alternatives
Context / RAM
Downloads
Likes
Captain Eris Violet V0.420 12B1000K / 24.5 GB36872229
...r Nemo 12B Instruct R 21 09 241000K / 24.5 GB8787111
...s PersonalityEngine V1.1.0 12B1000K / 24.5 GB55732
Mistral Nemo Kartoffel 12B1000K / 24.5 GB3114
MN 12B Mag Mell R11000K / 24.5 GB6336115
MN 12B Mimicore GreenSnake1000K / 24.5 GB1682
SauerkrautLM Nemo 12B Instruct1000K / 24.5 GB2272122
Saiga Nemo 12b1000K / 24.5 GB25332239
MN 12B Mimicore WhiteSnake1000K / 24.5 GB883
Dolphin 2.9.3 Mistral Nemo 12B1000K / 24.5 GB725799
Note: green Score (e.g. "73.2") means that the model is better than EpistemeAI/Fireball-12B-v1.0.

Rank the Fireball 12B V1.0 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? 43470 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