What are the hardware requirements for Codellama 13B Oasst Sft V10?

Codellama 13B Oasst Sft V10 requires approximately 25.9 GB of VRAM and supports a context window of 16K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Codellama 13B Oasst Sft V10 and how large is it?

Codellama 13B Oasst Sft V10 is developed by OpenAssistant, a model with 13b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Codellama 13B Oasst Sft V10 perform on standard benchmarks?

Codellama 13B Oasst Sft V10 has the following published scores: MMLU 35.36. Compare against reference models on this page or on the LLM Explorer leaderboards.

Codellama 13B Oasst Sft V10 by OpenAssistant — VRAM 25.9GB, 16K context

Name: Codellama 13B Oasst Sft V10
Author: OpenAssistant

Codellama 13B Oasst Sft V10 is an open-source language model by OpenAssistant. Features: 13b LLM, VRAM: 25.9GB, Context: 16K, License: llama2, Code Generating, LLM Explorer Score: 0.12, Arc: 45.4, HellaSwag: 62.4, MMLU: 35.4, GSM8K: 13.2.

Codegen Custom code Dataset:openassistant/oasst1 Dataset:shahules786/orca-best En Endpoints compatible Llama Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/OpenAssistant/codellama-13b-oasst-sft-v10

Codellama 13B Oasst Sft V10 Parameters and Internals

Model Type

Causal decoder-only transformer language model

Additional Notes

Demo and training logs available, showcasing 6123 steps with batch size 64. Partially trained with orca system messages.

Supported Languages

en (English)

Training Details

Data Sources:

OpenAssistant/oasst1, shahules786/orca-best

Methodology:

Fine-tuned using orca system messages and OpenAI's chatml standard prompt format.

Hardware Used:

EPFL's Machine Learning and Optimization Laboratory, EPFL's Natural Language Processing Lab

Model Architecture:

RoPE Theta value (1e6 instead of 1e4)

Responsible Ai Considerations

Fairness:

Testing conducted in English and may not cover all scenarios, model may produce inaccurate, biased or objectionable responses.

Accountability:

Developers should perform safety testing and tuning for specific applications of the model.

Mitigation Strategies:

Refer to Meta's Responsible Use Guide

Input Output

Input Format:

OpenAI's chatml standard format

Accepted Modalities:

text

Output Format:

text responses

Performance Tips:

Use the official Llama2 system message for optimal inference.

LLM Name	Codellama 13B Oasst Sft V10
Repository 🤗	https://huggingface.co/OpenAssistant/codellama-13b-oasst-sft-v10
Model Size	13b
Required VRAM	25.9 GB
Updated	2026-07-08
Maintainer	OpenAssistant
Model Type	llama
Model Files	1.9 GB: 1-of-14 1.9 GB: 2-of-14 1.9 GB: 3-of-14 1.9 GB: 4-of-14 1.9 GB: 5-of-14 1.9 GB: 6-of-14 1.9 GB: 7-of-14 1.9 GB: 8-of-14 1.9 GB: 9-of-14 1.9 GB: 10-of-14 1.9 GB: 11-of-14 1.9 GB: 12-of-14 1.9 GB: 13-of-14 1.2 GB: 14-of-14
Supported Languages	en
Generates Code	Yes
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	16384
Model Max Length	16384
Transformers Version	4.31.0
Tokenizer Class	LlamaTokenizer
Beginning of Sentence Token	<s>
End of Sentence Token	</s>
Unk Token	<unk>
Vocabulary Size	32032
Torch Data Type	bfloat16

Rank the Codellama 13B Oasst Sft V10 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

Quantized Models of the Codellama 13B Oasst Sft V10

Model	Likes	Downloads	VRAM
...deLlama 13B Oasst Sft V10 GGUF	14	3894	5 GB
...odeLlama 13B Oasst Sft V10 AWQ	1	7	7 GB
...deLlama 13B Oasst Sft V10 GPTQ	13	43	7 GB
...deLlama 13B Oasst Sft V10 GGML	6	11	5 GB

Best Alternatives to Codellama 13B Oasst Sft V10

Best Alternatives	Context / RAM	Downloads	Likes
NexusRaven V2 13B	16K / 26 GB	105	472
CodeLlama 13B Instruct Hf	16K / 26 GB	2469	160
CodeLlama 13B MORepair	16K / 26 GB	9	2
CodeLlama 13B Hf	16K / 26 GB	3466	117
Tora Code 13B V1.0	16K / 26 GB	878	15
CodeLlama 13B Python Hf	16K / 26 GB	558	51
...ma 13B Hf Truncated Embeddings	16K / 52.3 GB	5	0
...ma Airoboros Orca Platypus 13B	16K / 26 GB	117	0
CodeLlama 13B Hf	16K / 26 GB	224	0
CodeLlama 13B Instruct Hf	16K / 26 GB	369	29

Note: green Score (e.g. "73.2") means that the model is better than OpenAssistant/codellama-13b-oasst-sft-v10.