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 / config.json

magiccodingman

Upload folder using huggingface_hub

b6bbe8a verified about 1 month ago

Raw History Blame Contribute Delete

3.19 kB

	{
	"architectures": [
	"DFlash2DraftModel"
	],
	"attention_bias": false,
	"attention_dropout": 0.0,
	"bos_token_id": null,
	"dflash_config": {
	"block_size": 8,
	"conv_group_size": 16,
	"conv_kernel_size": 2,
	"mask_token_id": 248070,
	"selector_rank": 256,
	"selector_top_k": 16,
	"target_layer_ids": [
	5,
	19,
	33,
	47,
	61
	]
	},
	"draft_vocab_size": 248320,
	"dtype": "bfloat16",
	"eos_token_id": 248044,
	"head_dim": 128,
	"hidden_act": "silu",
	"hidden_size": 5120,
	"initializer_range": 0.02,
	"intermediate_size": 17408,
	"is_causal": false,
	"layer_types": [
	"sliding_attention",
	"sliding_attention",
	"sliding_attention",
	"sliding_attention",
	"sliding_attention"
	],
	"max_position_embeddings": 262144,
	"max_window_layers": 5,
	"model_type": "qwen3",
	"num_attention_heads": 32,
	"num_hidden_layers": 5,
	"num_key_value_heads": 8,
	"num_target_layers": 64,
	"pad_token_id": 248044,
	"rms_norm_eps": 1e-06,
	"rope_parameters": {
	"rope_theta": 10000000,
	"rope_type": "default"
	},
	"rope_theta": 10000000,
	"sliding_window": 2048,
	"tie_word_embeddings": false,
	"transformers_version": "5.8.1",
	"use_cache": true,
	"use_sliding_window": true,
	"vocab_size": 248320,
	"quantization_config": {
	"quant_method": "fp8",
	"fmt": "e4m3",
	"activation_scheme": "dynamic",
	"weight_block_size": [
	128,
	128
	],
	"modules_to_not_convert": [
	"fc",
	"candidate_selector.hidden_projection",
	"layers.0.self_attn.q_proj",
	"layers.0.self_attn.k_proj",
	"layers.0.self_attn.v_proj",
	"layers.0.attention_conv.kernel_projection",
	"layers.0.mlp_conv.kernel_projection",
	"layers.0.self_attn.o_proj",
	"layers.1.self_attn.q_proj",
	"layers.1.self_attn.k_proj",
	"layers.1.self_attn.v_proj",
	"layers.1.attention_conv.kernel_projection",
	"layers.1.mlp_conv.kernel_projection",
	"layers.1.self_attn.o_proj",
	"layers.2.self_attn.q_proj",
	"layers.2.self_attn.k_proj",
	"layers.2.self_attn.v_proj",
	"layers.2.attention_conv.kernel_projection",
	"layers.2.mlp_conv.kernel_projection",
	"layers.2.self_attn.o_proj",
	"layers.3.self_attn.q_proj",
	"layers.3.self_attn.k_proj",
	"layers.3.self_attn.v_proj",
	"layers.3.attention_conv.kernel_projection",
	"layers.3.mlp_conv.kernel_projection",
	"layers.3.self_attn.o_proj",
	"layers.4.self_attn.q_proj",
	"layers.4.self_attn.k_proj",
	"layers.4.self_attn.v_proj",
	"layers.4.attention_conv.kernel_projection",
	"layers.4.mlp_conv.kernel_projection",
	"layers.4.self_attn.o_proj",
	"model.fc",
	"model.layers.64.self_attn.qkv_proj",
	"model.layers.65.self_attn.qkv_proj",
	"model.layers.66.self_attn.qkv_proj",
	"model.layers.67.self_attn.qkv_proj",
	"model.layers.68.self_attn.qkv_proj",
	"model.layers.64.self_attn.o_proj",
	"model.layers.65.self_attn.o_proj",
	"model.layers.66.self_attn.o_proj",
	"model.layers.67.self_attn.o_proj",
	"model.layers.68.self_attn.o_proj"
	]
	}
	}