Modelpublic

second-state/Qwen3-Coder-30B-A3B-Instruct-GGUF

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes3.7kdownloads

Model Card

<div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">

Qwen3-Coder-30B-A3B-Instruct-GGUF

Original Model

Qwen/Qwen3-Coder-30B-A3B-Instruct

Run with LlamaEdge

—LlamaEdge version: coming soon

<!-- - LlamaEdge version:

—Thinking: v0.17.0 and above
—No Thinking: v0.18.2 -->

—Prompt template

—Prompt type: chatml

—Prompt string

text

    <|im_start|>system
    {system_message}<|im_end|>
    <|im_start|>user
    {prompt}<|im_end|>
    <|im_start|>assistant

<!-- - Prompt type: qwen3-no-think

—Prompt string

text

    <|im_start|>system
    {system_message}<|im_end|>
    <|im_start|>user
    {user_message_1}<|im_end|>
    <|im_start|>assistant
    {assistant_message_1}<|im_end|>
    <|im_start|>user
    {user_message_2}<|im_end|>
    <|im_start|>assistant
    <think>

    </think>

—Context size: 256000

—Run as LlamaEdge service

bash

  wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen3-Coder-30B-A3B-Instruct-Q5_K_M.gguf \
    llama-api-server.wasm \
    --model-name Qwen3-Coder-30B-A3B-Instruct \
    --prompt-template chatml \
    --ctx-size 256000

<!-- ## Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Qwen3-Coder-30B-A3B-Instruct-Q2_K.gguf	Q2_K	2	3.28 GB	smallest, significant quality loss - not recommended for most purposes
Qwen3-Coder-30B-A3B-Instruct-Q3_K_L.gguf	Q3KL	3	4.43 GB	small, substantial quality loss
Qwen3-Coder-30B-A3B-Instruct-Q3_K_M.gguf	Q3KM	3	4.12 GB	very small, high quality loss
Qwen3-Coder-30B-A3B-Instruct-Q3_K_S.gguf	Q3KS	3	3.77 GB	very small, high quality loss
Qwen3-Coder-30B-A3B-Instruct-Q4_0.gguf	Q4_0	4	4.77 GB	legacy; small, very high quality loss - prefer using Q3KM
Qwen3-Coder-30B-A3B-Instruct-Q4_K_M.gguf	Q4KM	4	5.03 GB	medium, balanced quality - recommended
Qwen3-Coder-30B-A3B-Instruct-Q4_K_S.gguf	Q4KS	4	4.80 GB	small, greater quality loss
Qwen3-Coder-30B-A3B-Instruct-Q5_0.gguf	Q5_0	5	5.72 GB	legacy; medium, balanced quality - prefer using Q4KM
Qwen3-Coder-30B-A3B-Instruct-Q5_K_M.gguf	Q5KM	5	5.85 GB	large, very low quality loss - recommended
Qwen3-Coder-30B-A3B-Instruct-Q5_K_S.gguf	Q5KS	5	5.72 GB	large, low quality loss - recommended
Qwen3-Coder-30B-A3B-Instruct-Q6_K.gguf	Q6_K	6	6.73 GB	very large, extremely low quality loss
Qwen3-Coder-30B-A3B-Instruct-Q8_0.gguf	Q8_0	8	8.71 GB	very large, extremely low quality loss - not recommended
Qwen3-Coder-30B-A3B-Instruct-f16.gguf	f16	16	16.4 GB		-->

Quantized with llama.cpp b6209