Instructions to use magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8 with Transformers:

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

pipe = pipeline("text-generation", model="magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8")

# Load model directly
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8")
model = AutoModel.from_pretrained("magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8

SGLang

How to use magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8 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 "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8" \
    --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": "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8",
		"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 "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8" \
        --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": "magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8 with Docker Model Runner:
```
docker model run hf.co/magiccodingman/Qwen3.8-27B-heretic-ara-DFlash2-fp8
```

Qwen3.8-27B-heretic-ara-DFlash2-fp8 / fp8_structural_validation.json

magiccodingman

Upload folder using huggingface_hub

b6bbe8a verified about 15 hours ago

Raw

History Blame Contribute Delete

5.31 kB

	{
	"status": "pass",
	"profile": "mlp-only",
	"source_tensor_entries": 81,
	"candidate_tensor_entries": 101,
	"quantized_tensors": 15,
	"preserved_tensors_byte_equivalent": 66,
	"aggregate_rmse": 0.00343747144092223,
	"aggregate_relative_rmse": 0.026455517670677747,
	"worst_relative_rmse_tensors": [
	{
	"tensor": "layers.3.mlp.gate_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0032719125529877802,
	"relative_rmse": 0.026461520725813415,
	"maximum_absolute_error": 0.045758962631225586,
	"scale_min": 0.0009940011659637094,
	"scale_max": 0.0035923549439758062
	},
	{
	"tensor": "layers.0.mlp.gate_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.003276958150145669,
	"relative_rmse": 0.026461051027726844,
	"maximum_absolute_error": 0.03962057828903198,
	"scale_min": 0.0010157993528991938,
	"scale_max": 0.003383091650903225
	},
	{
	"tensor": "layers.2.mlp.up_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0035662104279010677,
	"relative_rmse": 0.02646011386372461,
	"maximum_absolute_error": 0.04241067171096802,
	"scale_min": 0.0010681153507903218,
	"scale_max": 0.004202706273645163
	},
	{
	"tensor": "layers.4.mlp.gate_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.00329046166073514,
	"relative_rmse": 0.026459302783789097,
	"maximum_absolute_error": 0.03348219394683838,
	"scale_min": 0.0009722027461975813,
	"scale_max": 0.003383091650903225
	},
	{
	"tensor": "layers.4.mlp.up_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0035572689212522556,
	"relative_rmse": 0.02645898389165941,
	"maximum_absolute_error": 0.03627234697341919,
	"scale_min": 0.0010637555969879031,
	"scale_max": 0.003069196594879031
	},
	{
	"tensor": "layers.1.mlp.up_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0035208753277019625,
	"relative_rmse": 0.02645866599872885,
	"maximum_absolute_error": 0.037946462631225586,
	"scale_min": 0.0010550363222137094,
	"scale_max": 0.0031912666745483875
	},
	{
	"tensor": "layers.0.mlp.up_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0032898814757157777,
	"relative_rmse": 0.026457355227407442,
	"maximum_absolute_error": 0.03236609697341919,
	"scale_min": 0.0010157993528991938,
	"scale_max": 0.002947126282379031
	},
	{
	"tensor": "layers.2.mlp.down_proj.weight",
	"shape": [
	5120,
	17408
	],
	"rmse": 0.003520015099063686,
	"relative_rmse": 0.026456624226654197,
	"maximum_absolute_error": 0.10491061210632324,
	"scale_min": 0.0010811942629516125,
	"scale_max": 0.0110909603536129
	},
	{
	"tensor": "layers.3.mlp.up_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0035803991978373777,
	"relative_rmse": 0.026455870434037503,
	"maximum_absolute_error": 0.04464280605316162,
	"scale_min": 0.0010506765684112906,
	"scale_max": 0.00537109375
	},
	{
	"tensor": "layers.0.mlp.down_proj.weight",
	"shape": [
	5120,
	17408
	],
	"rmse": 0.003413225732495426,
	"relative_rmse": 0.026454801653551588,
	"maximum_absolute_error": 0.20089268684387207,
	"scale_min": 0.0010114397155120969,
	"scale_max": 0.01736886240541935
	},
	{
	"tensor": "layers.1.mlp.gate_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.003344292283514903,
	"relative_rmse": 0.02645373161442601,
	"maximum_absolute_error": 0.032924115657806396,
	"scale_min": 0.0009852817747741938,
	"scale_max": 0.0031738278921693563
	},
	{
	"tensor": "layers.2.mlp.gate_proj.weight",
	"shape": [
	17408,
	5120
	],
	"rmse": 0.0033150167877642202,
	"relative_rmse": 0.026452570119492928,
	"maximum_absolute_error": 0.03459817171096802,
	"scale_min": 0.0009809221373870969,
	"scale_max": 0.003313336754217744
	},
	{
	"tensor": "layers.3.mlp.down_proj.weight",
	"shape": [
	5120,
	17408
	],
	"rmse": 0.0035649403766182847,
	"relative_rmse": 0.026452033474464075,
	"maximum_absolute_error": 0.07589280605316162,
	"scale_min": 0.0010811942629516125,
	"scale_max": 0.01318359375
	},
	{
	"tensor": "layers.1.mlp.down_proj.weight",
	"shape": [
	5120,
	17408
	],
	"rmse": 0.003493047765345981,
	"relative_rmse": 0.02644884647564942,
	"maximum_absolute_error": 0.16741085052490234,
	"scale_min": 0.0010463169310241938,
	"scale_max": 0.01311383955180645
	},
	{
	"tensor": "layers.4.mlp.down_proj.weight",
	"shape": [
	5120,
	17408
	],
	"rmse": 0.0035265503863780205,
	"relative_rmse": 0.02644285322798197,
	"maximum_absolute_error": 0.30803585052490234,
	"scale_min": 0.0010375975398346782,
	"scale_max": 0.0203683041036129
	}
	],
	"elapsed_seconds": 8.807376893004403
	}