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

Phi SoSerious Mini V1 by concedo

 »  All LLMs  »  concedo  »  Phi SoSerious Mini V1   URL Share it on

  12   
  Autotrain compatible   Custom code   Endpoints compatible   Phi3   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/concedo/Phi-SoSerious-Mini-V1 

Phi SoSerious Mini V1 Benchmarks

ARC: 62.37 vs 96.7 (so35)-35.5%
HellaSwag: 77.74 vs 95.3 (gpt4)-18.4%
MMLU: 69.24 vs 88.3 (so35)-21.6%
TruthfulQA: 50.92 vs 59 (gpt4)-13.7%
WinoGrande: 74.19 vs 87.5 (gpt4)-15.2%
GSM8K: 69.22 vs 96.4 (so35)-28.2%
LLME Score: 0.20155
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Phi SoSerious Mini V1 (concedo/Phi-SoSerious-Mini-V1)

Phi SoSerious Mini V1 Parameters and Internals

Model Type 
text generation, storywriting, conversation, instruct
Use Cases 
Applications:
storywriting, conversations, instructions
Primary Use Cases:
storywriting, conversations, instructions
Limitations:
Excess tendency for uncensored and unrestricted responses
Considerations:
For use with KoboldAI software and Kobold Lite
Training Details 
Data Sources:
Kobble Dataset, synthetic data
Methodology:
qLora finetuning (LR 1.2e-4, rank 16, alpha 16, batch size 3, gradient acc. 3, 2048 ctx)
Context Length:
2048
Training Time:
under 4 hours
Hardware Used:
single Nvidia RTX 3090 GPU
Input Output 
Input Format:
Alpaca
Accepted Modalities:
text
LLM NamePhi SoSerious Mini V1
Repository 🤗https://huggingface.co/concedo/Phi-SoSerious-Mini-V1 
Model Size3.8b
Required VRAM7.7 GB
Updated2025-02-22
Maintainerconcedo
Model Typephi3
Model Files  5.0 GB: 1-of-2   2.7 GB: 2-of-2
Model ArchitecturePhi3ForCausalLM
Licenseother
Context Length4096
Model Max Length4096
Transformers Version4.41.0
Tokenizer ClassLlamaTokenizer
Padding Token<|endoftext|>
Vocabulary Size32064
Torch Data Typefloat16

Best Alternatives to Phi SoSerious Mini V1

Best Alternatives
Context / RAM
Downloads
Likes
Phi 3.5 Mini Instruct128K / 7.7 GB721615819
Phi 3 Mini 128K Instruct128K / 7.7 GB1741761636
NuExtract 1.5128K / 7.7 GB115378198
NuExtract V1.5128K / 7.7 GB10851189
Phi 3.5 Mini TitanFusion 0.1128K / 7.7 GB1650
Glider128K / 15.4 GB150436
Saka 3.8B128K / 7.7 GB3091
ECE EIFFEL 3Bv2128K / 7.7 GB100
Samantha2.0 Phi 3.5 Mini ITA128K / 7.7 GB41210
Phi3 Medmcqa Finetuned128K / 15.4 GB440
Note: green Score (e.g. "73.2") means that the model is better than concedo/Phi-SoSerious-Mini-V1.

Rank the Phi SoSerious Mini V1 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