What are the hardware requirements for Chinese Llama 2 7B 4bit?

Chinese Llama 2 7B 4bit requires approximately 13.5 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Chinese Llama 2 7B 4bit and how large is it?

Chinese Llama 2 7B 4bit is developed by LinkSoul, a model with 7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Chinese Llama 2 7B 4bit perform on standard benchmarks?

Chinese Llama 2 7B 4bit has the following published scores: MMLU 37.42. Compare against reference models on this page or on the LLM Explorer leaderboards.

Chinese Llama 2 7B 4bit by LinkSoul — VRAM 13.5GB, 2K context

Name: Chinese Llama 2 7B 4bit
Author: LinkSoul

Chinese Llama 2 7B 4bit is an open-source language model by LinkSoul. Features: 7b LLM, VRAM: 13.5GB, Context: 2K, License: openrail, Quantized, Instruction-Based, Merged, LLM Explorer Score: 0.12, Arc: 44.5, HellaSwag: 69.5, MMLU: 37.4, GSM8K: 3.8.

Merged Model 4bit Dataset:linksoul/instruction m... En Endpoints compatible Instruct Llama Pytorch Quantized Region:us Sharded Zh

Model Card on HF 🤗: https://huggingface.co/LinkSoul/Chinese-Llama-2-7b-4bit

Chinese Llama 2 7B 4bit Parameters and Internals

Model Type

text generation, multimodal

Use Cases

Areas:

research, commercial

Applications:

text generation

Primary Use Cases:

multilingual chatbots in Chinese and English

Additional Notes

Compatible with all optimizations for the original Llama-2-chat model

Supported Languages

zh (high), en (moderate)

Training Details

Data Sources:

LinkSoul/instruction_merge_set

Data Volume:

10 million

Model Architecture:

Llama-2-chat with optimizations

Input Output

Input Format:

Llama-2-chat format with instructions

Accepted Modalities:

text

Output Format:

text

LLM Name	Chinese Llama 2 7B 4bit
Repository 🤗	https://huggingface.co/LinkSoul/Chinese-Llama-2-7b-4bit
Base Model(s)	Chinese Llama 2 7B 64K hfl/chinese-llama-2-7b-64k
Merged Model	Yes
Model Size	7b
Required VRAM	13.5 GB
Updated	2026-07-23
Maintainer	LinkSoul
Model Type	llama
Instruction-Based	Yes
Model Files	10.0 GB: 1-of-2 3.5 GB: 2-of-2
Supported Languages	zh en
Quantization Type	4bit
Model Architecture	LlamaForCausalLM
License	openrail
Context Length	2048
Model Max Length	2048
Transformers Version	4.30.2
Tokenizer Class	LlamaTokenizer
Beginning of Sentence Token	<s>
End of Sentence Token	</s>
Unk Token	<unk>
Vocabulary Size	32000
Torch Data Type	float16

Rank the Chinese Llama 2 7B 4bit 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 Chinese Llama 2 7B 4bit

Best Alternatives	Context / RAM	Downloads	Likes
...p 0.05 Max Grad1.0 Grad Accu32	32K / 14.4 GB	9	0
...p 0.05 Max Grad1.0 Grad Accu32	32K / 14.4 GB	6	0
...ruct Solidity Bnb 4bit Smashed	16K / 4.2 GB	6	0
...B Instruct Hf Bnb 4bit Smashed	16K / 4.2 GB	6	0
CodelLama7B Inst DPO 7K Mlx	16K / 4.2 GB	49	4
...eLlama 7B Instruct Hf 4bit MLX	16K / 4.2 GB	84	3
...6.7B Instruct 8.0bpw H8 EXL2 2	16K / 6.8 GB	7	2
...6.7B Instruct 3.0bpw H6 EXL2 2	16K / 2.8 GB	8	1
...coder S CL 7B 5.0bpw H6 EXL2 2	16K / 4.4 GB	6	1
...coder S CL 7B 3.0bpw H6 EXL2 2	16K / 2.8 GB	2	1

Note: green Score (e.g. "73.2") means that the model is better than LinkSoul/Chinese-Llama-2-7b-4bit.