Instructions to use amd/tiny-qwen3-moe-w8a8-int8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use amd/tiny-qwen3-moe-w8a8-int8 with Transformers:

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

pipe = pipeline("text-generation", model="amd/tiny-qwen3-moe-w8a8-int8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("amd/tiny-qwen3-moe-w8a8-int8")
model = AutoModelForCausalLM.from_pretrained("amd/tiny-qwen3-moe-w8a8-int8", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use amd/tiny-qwen3-moe-w8a8-int8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "amd/tiny-qwen3-moe-w8a8-int8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "amd/tiny-qwen3-moe-w8a8-int8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/amd/tiny-qwen3-moe-w8a8-int8

SGLang

How to use amd/tiny-qwen3-moe-w8a8-int8 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 "amd/tiny-qwen3-moe-w8a8-int8" \
    --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": "amd/tiny-qwen3-moe-w8a8-int8",
		"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 "amd/tiny-qwen3-moe-w8a8-int8" \
        --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": "amd/tiny-qwen3-moe-w8a8-int8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use amd/tiny-qwen3-moe-w8a8-int8 with Docker Model Runner:
```
docker model run hf.co/amd/tiny-qwen3-moe-w8a8-int8
```

Model Overview

Model Architecture: Qwen3MoeForCausalLM (tiny, randomly initialized)
- Input: Text
- Output: Text
Supported Hardware Microarchitecture: AMD MI300 / MI350 / MI355 (gfx942 / gfx950), NVIDIA (SM ≥ 7.5)
Inference Engine: vLLM
Model Optimizer: AMD-Quark
- Weight quantization: INT8, per-channel, static, symmetric
- Activation quantization: INT8, per-token, dynamic, symmetric
- Quantized layers: attention (self_attn) and the routed MoE experts. The MoE router (mlp.gate) and lm_head are kept in the original precision.

This is a tiny, randomly-initialized Qwen3-MoE model quantized to W8A8-INT8, used purely as vLLM CI coverage for the Quark INT8 fused-MoE path (QuarkW8A8Int8MoEMethod), which dispatches through the modular Triton INT8 MoE kernel. It is not intended to produce meaningful text.

Note: the router (mlp.gate) is intentionally left in bf16. Its weight is [num_experts, hidden] and num_experts (8) is not a multiple of 16, which would trip AITER's INT8 linear-kernel shape assertion. Routers are normally left in high precision anyway.

Model Creation

Built and quantized with AMD-Quark: per-channel INT8 weights (static) + per-token INT8 activations (dynamic), symmetric, on a tiny qwen3_moe model. Attention and routed experts are quantized; mlp.gate and lm_head are excluded. vocab_size matches the tokenizer so token ids stay within the embedding table.

Usage in vLLM

vllm serve amd/tiny-qwen3-moe-w8a8-int8 --enforce-eager

Because the weights are random, outputs are not meaningful — this model is a structural / smoke-test fixture only.

License

Apache-2.0. The tiny model is randomly initialized and derives no weights from any base model.

Downloads last month: 51

Safetensors

Model size

0.7B params

Tensor type

BF16