Instructions to use PhysShell/Qwen3.6-35B-A3B-DFlash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use PhysShell/Qwen3.6-35B-A3B-DFlash with Transformers:

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

pipe = pipeline("text-generation", model="PhysShell/Qwen3.6-35B-A3B-DFlash", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("PhysShell/Qwen3.6-35B-A3B-DFlash", trust_remote_code=True)
model = AutoModel.from_pretrained("PhysShell/Qwen3.6-35B-A3B-DFlash", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use PhysShell/Qwen3.6-35B-A3B-DFlash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "PhysShell/Qwen3.6-35B-A3B-DFlash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PhysShell/Qwen3.6-35B-A3B-DFlash",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/PhysShell/Qwen3.6-35B-A3B-DFlash

SGLang

How to use PhysShell/Qwen3.6-35B-A3B-DFlash 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 "PhysShell/Qwen3.6-35B-A3B-DFlash" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PhysShell/Qwen3.6-35B-A3B-DFlash",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "PhysShell/Qwen3.6-35B-A3B-DFlash" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PhysShell/Qwen3.6-35B-A3B-DFlash",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use PhysShell/Qwen3.6-35B-A3B-DFlash with Docker Model Runner:
```
docker model run hf.co/PhysShell/Qwen3.6-35B-A3B-DFlash
```

Qwen3.6-35B-A3B-DFlash / config.json

PhysShell

Duplicate from z-lab/Qwen3.6-35B-A3B-DFlash

bdbf0fa about 1 month ago

raw

history blame contribute delete

1.32 kB

	{
	"architectures": [
	"DFlashDraftModel"
	],
	"attention_bias": false,
	"attention_dropout": 0.0,
	"auto_map": {
	"AutoModel": "dflash.DFlashDraftModel"
	},
	"block_size": 16,
	"dflash_config": {
	"mask_token_id": 248070,
	"target_layer_ids": [
	1,
	10,
	19,
	28,
	37
	]
	},
	"dtype": "bfloat16",
	"eos_token_id": 248046,
	"head_dim": 128,
	"hidden_act": "silu",
	"hidden_size": 2048,
	"initializer_range": 0.02,
	"intermediate_size": 6144,
	"layer_types": [
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention"
	],
	"max_position_embeddings": 262144,
	"max_window_layers": 8,
	"model_type": "qwen3",
	"num_attention_heads": 32,
	"num_hidden_layers": 8,
	"num_key_value_heads": 4,
	"num_target_layers": 40,
	"pad_token_id": 248044,
	"rms_norm_eps": 1e-06,
	"rope_scaling": {
	"beta_fast": 32.0,
	"beta_slow": 1.0,
	"factor": 64.0,
	"original_max_position_embeddings": 4096,
	"rope_type": "yarn",
	"type": "yarn"
	},
	"rope_theta": 10000000,
	"sliding_window": null,
	"tie_word_embeddings": false,
	"transformers_version": "4.57.1",
	"use_cache": false,
	"use_sliding_window": false,
	"vocab_size": 248320
	}