Phi 3 Mini 4K Instruct By microsoft: Benchmarks, Features and Detailed Analysis. Insights on Phi 3 Mini 4K Instruct.

Autotrain compatible Code Conversational Custom code En Endpoints compatible Fr Instruct Phi3 Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/microsoft/Phi-3-mini-4k-instruct

Phi 3 Mini 4K Instruct Benchmarks

LMSys ELO: 1081 vs 1272 (so35)^-15%

IFEval: 54.77 vs 88 (so35)^-37.8%

ARC: 62.97 vs 96.7 (so35)^-34.9%

HellaSwag: 80.6 vs 95.3 (gpt4)^-15.4%

MMLU: 69.08 vs 88.3 (so35)^-21.8%

TruthfulQA: 59.88 vs 59 (gpt4)^1.5%

WinoGrande: 72.38 vs 87.5 (gpt4)^-17.3%

GSM8K: 74.53 vs 96.4 (so35)^-22.7%

MATH Lvl 5: 16.39

LLME Score: 0.48726

^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)

Phi 3 Mini 4K Instruct (microsoft/Phi-3-mini-4k-instruct)

🌟 Advertise your project 🚀

Phi 3 Mini 4K Instruct Parameters and Internals

Model Type

text generation, language model

Use Cases

Areas:

Research, Commercial applications

Applications:

General purpose AI systems, Computationally constrained environments

Primary Use Cases:

Memory/computational constraint scenarios, Latency bound applications, Tasks requiring mathematical and logical reasoning

Limitations:

Limited by language/data representation bias, Requires additional debiasing techniques for high-risk use cases

Considerations:

Evaluate performance and mitigate for safety and accuracy.

Additional Notes

The model's performance improves when integrated with retrieval systems for external knowledge.

Supported Languages

en (High proficiency), fr (Moderate proficiency)

Training Details

Data Sources:

Publicly available documents, High-quality educational data, Newly created synthetic data, Chat format supervised data

Data Volume:

4.9 trillion tokens

Methodology:

Supervised fine-tuning and Direct Preference Optimization

Context Length:

4000

Training Time:

10 days

Hardware Used:

512 H100-80G GPUs

Model Architecture:

Dense decoder-only Transformer model

Safety Evaluation

Methodologies:

Supervised fine-tuning, Direct Preference Optimization

Findings:

Strong reasoning capabilities, Improved instruction following

Risk Categories:

Misinformation, Bias

Ethical Considerations:

Use responsibly and ensure compliance with laws.

Responsible Ai Considerations

Fairness:

Address bias through training data selection and filtering.

Transparency:

Encourage user feedback and continuous improvement.

Accountability:

Developers responsible for outputs and compliance.

Mitigation Strategies:

Use Retrieval Augmented Generation for grounding responses.

Input Output

Input Format:

Chat format prompts

Accepted Modalities:

text

Output Format:

Text generation outputs

Performance Tips:

Utilize chat format for best results. Consider prompt engineering for improved performance.

Release Notes

Version:

June 2024 Update

Notes:

Improved instruction following, structure output, and reasoning compared to the original release.

LLM Name	Phi 3 Mini 4K Instruct
Repository 🤗	https://huggingface.co/microsoft/Phi-3-mini-4k-instruct
Model Size	3.8b
Required VRAM	7.7 GB
Updated	2025-09-23
Maintainer	microsoft
Model Type	phi3
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 2.7 GB: 2-of-2
Supported Languages	en fr
Model Architecture	Phi3ForCausalLM
License	mit
Context Length	4096
Model Max Length	4096
Transformers Version	4.40.2
Tokenizer Class	LlamaTokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	32064
Torch Data Type	bfloat16

Quantized Models of the Phi 3 Mini 4K Instruct

Model	Likes	Downloads	VRAM
Phi 3 Mini 4K Instruct GGUF	4	1481	1 GB
...i 3 Mini 4K Instruct IMat GGUF	1	1411	0 GB
...hi 3 Mini 4K Instruct Bnb 4bit	5	26	2 GB

Best Alternatives to Phi 3 Mini 4K Instruct

Best Alternatives	Context / RAM	Downloads	Likes
Phi 4 Mini Instruct	128K / 7.7 GB	268462	600
Phi 3 Mini 128K Instruct	128K / 7.7 GB	730029	1673
Phi 3.5 Mini Instruct	128K / 7.7 GB	291298	909
MediPhi Instruct	128K / 7.7 GB	4661	44
NuExtract 1.5	128K / 7.7 GB	158853	239
MediPhi Clinical	128K / 7.7 GB	7080	9
NuExtract V1.5	128K / 7.7 GB	108511	89
Phi 3.5 Mini TitanFusion 0.1	128K / 7.7 GB	5	0
Phi 4 Mini Instruct	128K / 7.7 GB	4930	20
MediPhi MedCode	128K / 7.7 GB	845	3

Rank the Phi 3 Mini 4K Instruct 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? 51544 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 v20241124

Support LLM Explorer