What are the hardware requirements for CodeFuse DeepSeek 33B 3.0bpw H6 EXL2?

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 requires approximately 13 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 CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 and how large is it?

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 is developed by LoneStriker, a model with 33b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate CodeFuse DeepSeek 33B 3.0bpw H6 EXL2?

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 by LoneStriker — VRAM 13GB, 16K context

Name: CodeFuse DeepSeek 33B 3.0bpw H6 EXL2
Author: LoneStriker

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 is an open-source language model by LoneStriker. Features: 33b LLM, VRAM: 13GB, Context: 16K, License: other, Quantized, LLM Explorer Score: 0.11.

Conversational Endpoints compatible Exl2 Llama Pytorch Quantized Region:us Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/LoneStriker/CodeFuse-DeepSeek-33B-3.0bpw-h6-exl2

CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 Parameters and Internals

Model Type

code-generation

Training Details

Methodology:

QLoRA

Input Output

Input Format:

Concatenated string in training data format

Accepted Modalities:

text

Output Format:

Generated code

Performance Tips:

Ensure input string ends with '\ bot' for generating answers.

Release Notes

Date:

2024-01-12

Notes:

Released with pass@1 score of 78.65% on HumanEval.

LLM Name	CodeFuse DeepSeek 33B 3.0bpw H6 EXL2
Repository 🤗	https://huggingface.co/LoneStriker/CodeFuse-DeepSeek-33B-3.0bpw-h6-exl2
Model Size	33b
Required VRAM	13 GB
Updated	2026-08-07
Maintainer	LoneStriker
Model Type	llama
Model Files	8.6 GB: 1-of-2 4.4 GB: 2-of-2
Quantization Type	exl2
Model Architecture	LlamaForCausalLM
License	other
Context Length	16384
Model Max Length	16384
Transformers Version	4.34.1
Tokenizer Class	LlamaTokenizerFast
Beginning of Sentence Token	<｜begin▁of▁sentence｜>
End of Sentence Token	<｜end▁of▁sentence｜>
Vocabulary Size	32256
Torch Data Type	bfloat16

Rank the CodeFuse DeepSeek 33B 3.0bpw H6 EXL2 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 CodeFuse DeepSeek 33B 3.0bpw H6 EXL2

Best Alternatives	Context / RAM	Downloads	Likes
...Interpreter DS 33B Hf 4bit Mlx	16K / 19.1 GB	270	7
...erpreter DS 33B 4.0bpw H6 EXL2	16K / 17.1 GB	8	4
...rpreter DS 33B 4.65bpw H6 EXL2	16K / 19.8 GB	5	3
...erpreter DS 33B 5.0bpw H6 EXL2	16K / 21.2 GB	6	1
...erpreter DS 33B 6.0bpw H6 EXL2	16K / 25.3 GB	8	0
...erpreter DS 33B 8.0bpw H8 EXL2	16K / 33.5 GB	5	1
...e DeepSeek 33B 4.65bpw H6 EXL2	16K / 19.8 GB	7	2
...se DeepSeek 33B 4.0bpw H6 EXL2	16K / 17.1 GB	5	2
...se DeepSeek 33B 8.0bpw H8 EXL2	16K / 33.5 GB	3	2
...se DeepSeek 33B 6.0bpw H6 EXL2	16K / 25.3 GB	5	1

Note: green Score (e.g. "73.2") means that the model is better than LoneStriker/CodeFuse-DeepSeek-33B-3.0bpw-h6-exl2.