Name: Phi 1 5
Author: microsoft

Phi 1 5 is an open-source language model by microsoft. Features: 1.4b LLM, VRAM: 2.8GB, Context: 2K, License: mit, HF Score: 47.7, LLM Explorer Score: 0.27, Arc: 55.4, HellaSwag: 57.1, MMLU: 44.3, TruthfulQA: 40.9, WinoGrande: 72.2, GSM8K: 32.5.

Arxiv:2309.05463 Code Deploy:azure En Endpoints compatible Phi Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/microsoft/phi-1_5

Phi 1 5 Benchmarks

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: 1.81

LLME Score: 0.26785

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

🌟 Advertise your project 🚀

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 Name	Phi 1 5
Repository 🤗	https://huggingface.co/microsoft/phi-1_5
Model Size	1.4b
Required VRAM	2.8 GB
Updated	2026-04-09
Maintainer	microsoft
Model Type	phi
Model Files	2.8 GB
Supported Languages	en
Model Architecture	PhiForCausalLM
License	mit
Context Length	2048
Model Max Length	2048
Transformers Version	4.37.0
Tokenizer Class	CodeGenTokenizer
Vocabulary Size	51200
Torch Data Type	float16

Quantized Models of the Phi 1 5

Model	Likes	Downloads	VRAM
Phi 1 5 Q4	1	3	0 GB

Best Alternatives to Phi 1 5

Best Alternatives	Context / RAM	Downloads	Likes
Phi 1 5 Instruct V0.1	2K / 2.8 GB	191	2
Phi 1	2K / 2.8 GB	5869	219
Phi 1 5 Tldr Sft	2K / 5.7 GB	5	0
...tic Curriculum Subjects 1 To 5	2K / 0 GB	5	0
...1 5 FULL Arithmetic Curriculum	2K / 0 GB	5	0
...ath Phi 1 5 FULL Arithmetic 2K	2K / 0 GB	5	0
Phibode 1 5 Ultraalpaca	2K / 5.7 GB	14	3
...i 1 5 Hinglish Text Pretrained	2K / 5.7 GB	5	0
Phi Sentiment Analysis Model	2K / 2.8 GB	9	1
BitLinear Phi 1.5	2K / 5.7 GB	9	1

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? 52721 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.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Phi 1 5 by microsoft

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