Instructions to use h-tonywu/qwen3.5-moe-debug with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use h-tonywu/qwen3.5-moe-debug with Transformers:

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

pipe = pipeline("image-text-to-text", model="h-tonywu/qwen3.5-moe-debug")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("h-tonywu/qwen3.5-moe-debug")
model = AutoModelForMultimodalLM.from_pretrained("h-tonywu/qwen3.5-moe-debug", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use h-tonywu/qwen3.5-moe-debug with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "h-tonywu/qwen3.5-moe-debug"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "h-tonywu/qwen3.5-moe-debug",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/h-tonywu/qwen3.5-moe-debug

SGLang

How to use h-tonywu/qwen3.5-moe-debug 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 "h-tonywu/qwen3.5-moe-debug" \
    --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": "h-tonywu/qwen3.5-moe-debug",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "h-tonywu/qwen3.5-moe-debug" \
        --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": "h-tonywu/qwen3.5-moe-debug",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use h-tonywu/qwen3.5-moe-debug with Docker Model Runner:
```
docker model run hf.co/h-tonywu/qwen3.5-moe-debug
```

h-tonywu commited on Mar 6

Commit

2a25012

verified ·

1 Parent(s): bca2b77

Update README.md

Browse files

Files changed (1) hide show

README.md +48 -0

README.md CHANGED Viewed

@@ -92,6 +92,54 @@ output_text = processor.batch_decode(generated_ids[0][inputs['input_ids'].shape[
 print(output_text)
 ```
 ### Codes to create this repo:
 <details>

 print(output_text)
 ```
+### Comparison
+| Field                | Qwen/Qwen3.5-397B-A17B | h-tonywu/qwen3.5-moe-debug |
+| -------------------- | ---------------------- | -------------------------- |
+| transformers_version | 4.57.0.dev0            | 5.2.0.dev0                 |
+| top-level `dtype`    | —                      | bfloat16                   |
+#### Text model
+| Field                           | 397B         | Debug     |
+| ------------------------------- | ------------ | --------- |
+| hidden_size                     | 4096         | 8         |
+| head_dim                        | 256          | 32        |
+| num_hidden_layers               | 60           | 4         |
+| num_attention_heads             | 32           | 8         |
+| num_key_value_heads             | 2            | 4         |
+| num_experts                     | 512          | 128       |
+| num_experts_per_tok             | 10           | 10        |
+| moe_intermediate_size           | 1024         | 32        |
+| shared_expert_intermediate_size | 1024         | 32        |
+| intermediate_size               | —            | 32        |
+| linear_key_head_dim             | 128          | 32        |
+| linear_value_head_dim           | 128          | 32        |
+| linear_num_key_heads            | 16           | 4         |
+| linear_num_value_heads          | 64           | 8         |
+| layer_types length              | 60           | 4         |
+| rope_parameters.mrope_section   | [11, 11, 10] | [1, 1, 2] |
+#### Text config fields only in debug
+| Field                                            | Value |
+| ------------------------------------------------ | ----- |
+| bos_token_id                                     | null  |
+| pad_token_id                                     | null  |
+| output_router_logits                             | false |
+| text_config.tie_word_embeddings                  | false |
+| partial_rotary_factor (top-level in text_config) | 0.25  |
+#### Vision model
+| Field             | 397B | Debug |
+| ----------------- | ---- | ----- |
+| depth             | 27   | 2     |
+| hidden_size       | 1152 | 64    |
+| intermediate_size | 4304 | 128   |
+| num_heads         | 16   | 2     |
+| out_hidden_size   | 4096 | 8     |
 ### Codes to create this repo:
 <details>