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

Phi 1 5 by microsoft

 »  All LLMs  »  microsoft  »  Phi 1 5   URL Share it on

  141563   
  Arxiv:2309.05463   Autotrain compatible   Code   En   Endpoints compatible   Phi   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/microsoft/phi-1_5 

Phi 1 5 Benchmarks

MMLU Pro: 7.68
GPQA: 2.35
MUSR: 3.39
BBH: 7.47
IFEval: 20.33 vs 88 (so35)-76.9%
ARC: 55.4 vs 96.7 (so35)-42.7%
HellaSwag: 57.1 vs 95.3 (gpt4)-40.1%
MMLU: 44.3 vs 88.3 (so35)-49.8%
TruthfulQA: 40.89 vs 59 (gpt4)-30.7%
WinoGrande: 72.22 vs 87.5 (gpt4)-17.5%
GSM8K: 32.5 vs 96.4 (so35)-66.3%
OpenCompas: 39.3 vs 73.3 (gpt4)-46.4%
Exam: 40.7 vs 77.2 (gpt4)-47.3%
Language: 29.7 vs 62 (gpt4)-52.1%
Knowledge: 35.4 vs 73.5 (gpt4)-51.8%
Understand: 51.5 vs 70 (gpt4)-26.4%
Reasoning: 38.5 vs 74.4 (gpt4)-48.3%
MATH Lvl 5: 1.13
LLME Score: 0.35169
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Phi 1 5 (microsoft/phi-1_5)

Phi 1 5 Parameters and Internals

Model Type 
Transformer, text generation, NLP, code
Use Cases 
Areas:
research
Applications:
text generation, code generation
Primary Use Cases:
poem writing, email drafting, story creation, text summarization, Python code writing
Limitations:
Potential to generate harmful content, Generate inaccurate code and facts, Unreliable responses to instruction, Limited scope for code
Considerations:
Users should be cautious and critically evaluate outputs.
Additional Notes 
Phi-1.5-generated text/code should be treated as a starting point. Users should verify API uses manually where uncommon packages are involved.
Supported Languages 
en (standard English)
Training Details 
Data Sources:
same data sources as phi-1, various NLP synthetic texts
Data Volume:
150B tokens
Training Time:
8 days
Hardware Used:
32xA100-40G GPUs
Model Architecture:
Transformer-based with next-word prediction objective
Responsible Ai Considerations 
Transparency:
The model has not undergone instruction fine-tuning.
Mitigation Strategies:
Model is intended for research to help develop methods to reduce toxicity directly after pretraining.
Input Output 
Input Format:
QA format, Chat format, Code format
Accepted Modalities:
text
Output Format:
text
Performance Tips:
Users should update to `transformers` version 4.37.0 or higher.
LLM NamePhi 1 5
Repository 🤗https://huggingface.co/microsoft/phi-1_5 
Model Size1.4b
Required VRAM2.8 GB
Updated2024-12-22
Maintainermicrosoft
Model Typephi
Model Files  2.8 GB
Supported Languagesen
Model ArchitecturePhiForCausalLM
Licensemit
Context Length2048
Model Max Length2048
Transformers Version4.37.0
Tokenizer ClassCodeGenTokenizer
Vocabulary Size51200
Torch Data Typefloat16

Quantized Models of the Phi 1 5

Model
Likes
Downloads
VRAM
Phi 1 5 Q4170 GB

Best Alternatives to Phi 1 5

Best Alternatives
Context / RAM
Downloads
Likes
Phi 1 5 Instruct V0.12K / 2.8 GB2161
Phi 12K / 2.8 GB8125207
...i 1 5 Hinglish Text Pretrained2K / 5.7 GB3930
Phi 1 5 Tldr Sft2K / 5.7 GB110
...1 5 FULL Arithmetic Curriculum2K / 0 GB140
...tic Curriculum Subjects 1 To 52K / 0 GB90
...ath Phi 1 5 FULL Arithmetic 2K2K / 0 GB100
Phibode 1 5 Ultraalpaca2K / 5.7 GB283
Phi Sentiment Analysis Model2K / 2.8 GB181
Tofu Ft Phi 1.52K / 2.8 GB9180

Rank the Phi 1 5 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? 40066 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 v20241217
  Support LLM Explorer