What are the hardware requirements for Octocoder?

Octocoder requires approximately 62.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Octocoder and how large is it?

Octocoder is developed by bigcode, a model with 15.5b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Octocoder perform on standard benchmarks?

Octocoder has the following published scores: HumanEval (Python) 45.3. Compare against reference models on this page or on the LLM Explorer leaderboards.

Octocoder by bigcode — VRAM 62.1GB

Name: Octocoder
Author: bigcode

Octocoder is an open-source language model by bigcode. Features: 15.5b LLM, VRAM: 62.1GB, License: bigcode-openrail-m, Code Generating, LLM Explorer Score: 0.1, HumanEval: 45.3.

Arxiv:2308.07124 Code Codegen Dataset:bigcode/commitpackft Dataset:bigcode/oasst-octopack Endpoints compatible Model-index Pytorch Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/bigcode/octocoder

Octocoder Benchmarks

HumanEvalA: 24.01

Python Eva: 45.3 vs 92 (so35)^-50.8%

Racket Eva: 10.61

Rust Eval: 24.26

Swift Eval: 18.24

LLME Score: 0.09748

^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 🚀

Octocoder Parameters and Internals

Model Type

text generation

Training Details

Data Sources:

CommitPack, CommitPackFT, OASST

Data Volume:

1 trillion pretraining tokens and 2M instruction tuning

Methodology:

Instruction tuning with Fill-in-the-Middle objective

Training Time:

24 days for pretraining, 4 hours for instruction tuning

Hardware Used:

512 Tesla A100 for pretraining, 8 Tesla A100 for instruction tuning

Model Architecture:

GPT-2 model with multi-query attention

LLM Name	Octocoder
Repository 🤗	https://huggingface.co/bigcode/octocoder
Model Size	15.5b
Required VRAM	62.1 GB
Updated	2026-05-06
Maintainer	bigcode
Model Type	gpt_bigcode
Model Files	9.9 GB: 1-of-7 9.9 GB: 2-of-7 9.8 GB: 3-of-7 9.9 GB: 4-of-7 9.8 GB: 5-of-7 9.9 GB: 6-of-7 2.9 GB: 7-of-7 9.9 GB: 1-of-7 9.9 GB: 2-of-7 9.8 GB: 3-of-7 9.9 GB: 4-of-7 9.8 GB: 5-of-7 9.9 GB: 6-of-7 2.9 GB: 7-of-7
Generates Code	Yes
Model Architecture	GPTBigCodeForCausalLM
License	bigcode-openrail-m
Transformers Version	4.31.0.dev0
Tokenizer Class	GPT2Tokenizer
Vocabulary Size	49152
Torch Data Type	float32
Activation Function	gelu

Quantized Models of the Octocoder

Model	Likes	Downloads	VRAM
Octocoder GGML	15	16	10 GB
Octocoder GPTQ	8	31	9 GB

Best Alternatives to Octocoder

Best Alternatives	Context / RAM	Downloads	Likes
Starchat Alpha	0K / 31.2 GB	1165	231
Starchat Beta	0K / 31.2 GB	960	263
...der15B Personal Copilot Merged	0K / 30.9 GB	6	1
Starcoderbase Triviaqa	0K / 62.1 GB	8	0

Note: green Score (e.g. "73.2") means that the model is better than bigcode/octocoder.

Rank the Octocoder 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? 53972 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

Octocoder by bigcode