Instructions to use jdopensource/JoyAI-LLM-Flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jdopensource/JoyAI-LLM-Flash with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="jdopensource/JoyAI-LLM-Flash", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("jdopensource/JoyAI-LLM-Flash", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use jdopensource/JoyAI-LLM-Flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jdopensource/JoyAI-LLM-Flash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jdopensource/JoyAI-LLM-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jdopensource/JoyAI-LLM-Flash

SGLang

How to use jdopensource/JoyAI-LLM-Flash with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "jdopensource/JoyAI-LLM-Flash" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jdopensource/JoyAI-LLM-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "jdopensource/JoyAI-LLM-Flash" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jdopensource/JoyAI-LLM-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jdopensource/JoyAI-LLM-Flash with Docker Model Runner:
```
docker model run hf.co/jdopensource/JoyAI-LLM-Flash
```

🚀 JoyAI-LLM-Flash Official Collection Notice

#17

pinned

by Mingke977 - opened Feb 27

Discussion

Mingke977

JD Open Source org Feb 27

We are excited to announce that the official JoyAI-LLM-Flash collection by jdopensource now provides a complete lineup of models designed to support diverse hardware environments — from high-performance GPUs to memory-constrained edge devices.

All models below are part of the official Hugging Face collection.

🔷 Core Models

jdopensource/JoyAI-LLM-Flash
Full-precision flagship model for high-quality inference and advanced reasoning.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash
jdopensource/JoyAI-LLM-Flash-Base
Streamlined base version for rapid experimentation and foundational deployment.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-Base

🔶 Optimized & Quantized Variants

jdopensource/JoyAI-LLM-Flash-FP8
FP8 quantization — strong balance between performance and efficiency.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-FP8
jdopensource/JoyAI-LLM-Flash-INT4
Ultra-compact INT4 version for extremely limited VRAM environments.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-INT4
jdopensource/JoyAI-LLM-Flash-Block-INT8
Block-wise INT8 quantization for improved memory efficiency.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-Block-INT8
jdopensource/JoyAI-LLM-Flash-Channel-INT8
Channel-wise INT8 optimization for balanced throughput and accuracy.
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-Channel-INT8
jdopensource/JoyAI-LLM-Flash-GGUF
GGUF format for broad compatibility (e.g., llama.cpp, Ollama, CPU inference).
👉 https://huggingface.co/jdopensource/JoyAI-LLM-Flash-GGUF

🎯 Deploy Anywhere

Whether you are running:

High-throughput GPU servers
Consumer-grade GPUs
Low-memory edge devices
CPU-only local inference

The official JoyAI-LLM-Flash collection provides a tailored option for your hardware stack.

Choose the precision. Match your memory budget. Maximize performance.

Mingke977 pinned discussion Feb 27

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment