Llama3-8B-Chinese-Chat-GGUF

<div align="center">

Llama3-8B-Chinese-Chat — GGUF Quantizations

Quantized GGUF versions of shenzhi-wang/Llama3-8B-Chinese-Chat

Works with llama.cpp · Ollama · LM Studio · Open WebUI · Jan

Quantized by DhruvalLabs on June 29, 2026 using quant-kit

</div>

---

⚖️ The Pareto Frontier — Efficiency vs Intelligence

Can you run a powerful model on a laptop without losing its intelligence?

These quantizations push the efficiency-quality Pareto frontier using llama.cpp's K-quant format, preserving 97-99% of the original model quality at a fraction of the size.

---

📦 Available Files

💡 Which file should I download?

• Most users: Llama3-8B-Chinese-Chat-Q4_K_M.gguf — best balance of size and quality
• High RAM (32GB+): Llama3-8B-Chinese-Chat-Q8_0.gguf — near-original quality
• Low RAM (8GB): Llama3-8B-Chinese-Chat-Q3_K_M.gguf — fits in 8GB with room to spare

---

⚡ Speed Benchmarks

Run python benchmark.py --model Llama3-8B-Chinese-Chat to generate speed results.

---

🧠 Quality Benchmarks

Run kaggle_bench.ipynb on Kaggle to benchmark this model.

---

🚀 How to Use

Ollama

ollama run dhruvallabs/llama3-8b-chinese-chat

LM Studio / Jan / Open WebUI

Search for DhruvalLabs/Llama3-8B-Chinese-Chat in the model browser.

llama.cpp CLI

# Download the binary from https://github.com/ggerganov/llama.cpp/releases
./llama-cli \
  -m Llama3-8B-Chinese-Chat-Q4_K_M.gguf \
  -p "You are a helpful assistant." \
  --conversation \
  -n 512

Python — llama-cpp-python

from llama_cpp import Llama

llm = Llama(
    model_path="./Llama3-8B-Chinese-Chat-Q4_K_M.gguf",
    n_gpu_layers=-1,   # -1 = offload everything to GPU
    n_ctx=4096,
)

response = llm.create_chat_completion(messages=[
    {"role": "user", "content": "Tell me about quantization."}
])
print(response["choices"][0]["message"]["content"])

---

🔍 About GGUF Quantization

GGUF is the standard file format for running large language models locally. Quantization reduces the number of bits per weight:

---

💬 Community & Feedback

Found an issue? Have a question? Open a Discussion in the Community tab above.

If these quantizations were useful, please consider:

• ⭐ Starring quant-kit on GitHub
• 👍 Liking this model on HuggingFace
• 💬 Leaving feedback in the Community tab