Phi 3 Mini 128K Instruct is an open-source language model by tuong-nguyen-prd. Features: 3.8b LLM, VRAM: 7.7GB, Context: 128K, License: mit, Instruction-Based, LLM Explorer Score: 0.24, ELO: 1129.
Phi 3 Mini 128K Instruct Benchmarks
nn.n%
Phi 3 Mini 128K Instruct Parameters and Internals
Model Type Use Cases
Areas: Applications: Memory/compute constrained environments, Latency bound scenarios, Strong reasoning tasks like code, math and logic
Primary Use Cases: language and multimodal research, generative AI-powered features
Limitations: not specifically evaluated for all downstream purposes, performance varies across different modalities
Considerations: Developers should apply debiasing and further mitigate for accuracy, safety, and fairness.
Supported Languages languages_supported (en), proficiency_levels ()
Training Details
Data Sources: Publicly available documents filtered for quality, high-quality educational data, code
Data Volume: Methodology: Supervised fine-tuning (SFT) and Direct Preference Optimization (DPO)
Context Length: Training Time: Hardware Used: Model Architecture: 3.8B parameter dense decoder-only Transformer model
Safety Evaluation
Methodologies: Supervised fine-tuning (SFT), Direct Preference Optimization (DPO)
Findings: Can potentially behave unfairly or offend, Possibility of generating nonsensical content, Quality of Service may vary based on language variety
Risk Categories: misinformation, stereotype perpetuation
Ethical Considerations: Developers must adhere to responsible AI practices and ensure compliance with laws and regulations.
Responsible Ai Considerations
Fairness: Models may under/over-represent groups and decisions on use-cases should be sensitive to model limitations.
Transparency: Detailed transparency related to the training and evaluation process is provided.
Accountability: Developers are responsible for ensuring fair and compliant use.
Mitigation Strategies: Supervised fine-tuning and direct preference optimizations are used to align with human preferences and safety guidelines.
Input Output
Input Format: Chat format. E.g. <|user|>Question<|end|><|assistant|>...
Accepted Modalities: Output Format: Generated text in response to inputs
Rank the Phi 3 Mini 128K Instruct capabilities
Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.
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
Best Alternatives to Phi 3 Mini 128K Instruct
Note: green Score (e.g. "73.2 ") means that the model is better than tuong-nguyen-prd/Phi-3-mini-128k-instruct .
Expand