Octocoder GPTQ is an open-source language model by TheBloke. Features: 15.8b LLM, VRAM: 9.2GB, License: bigcode-openrail-m, Quantized, Code Generating, LLM Explorer Score: 0.08.
Octocoder GPTQ Parameters and Internals
Model Type Use Cases
Areas: Software development, Code generation
Applications: Programming assistance, Coding instruction
Primary Use Cases: Assisting programmers in writing code, Providing coding solutions based on instructions
Limitations: May not provide optimal solutions for complex problems., Performance is dependent on the quality of the input prompt.
Considerations: Preface input with 'Question: ' and finish it with 'Answer:'
Supported Languages programming_languages (80+ Programming languages)
Training Details
Data Sources: bigcode/commitpackft, bigcode/oasst-octopack
Data Volume: 1 trillion pretraining & 2M instruction tuning tokens
Methodology: Instruction Tuning on CommitPackFT and OASST
Training Time: Pretraining: 24 days, Instruction tuning: 4 hours
Hardware Used: Pretraining: 512 Tesla A100 GPUs, Instruction tuning: 8 Tesla A100 GPUs
Model Architecture: GPT-2 model with multi-query attention and Fill-in-the-Middle objective
Input Output
Input Format: Preface input with 'Question:' and finish with 'Answer:'
Accepted Modalities: Output Format: Performance Tips: Use quality prompts to improve output relevance.
Release Notes
Version: Notes: Introduction of OctoCoder with instruction tuning based on StarCoder.
Rank the Octocoder GPTQ 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 Octocoder GPTQ
Note: green Score (e.g. "73.2 ") means that the model is better than TheBloke/Octocoder-GPTQ .
Expand