Instructions to use RadixArk/glm47-flash-blockwise-fp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RadixArk/glm47-flash-blockwise-fp8 with Transformers:

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

pipe = pipeline("text-generation", model="RadixArk/glm47-flash-blockwise-fp8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("RadixArk/glm47-flash-blockwise-fp8")
model = AutoModelForCausalLM.from_pretrained("RadixArk/glm47-flash-blockwise-fp8", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use RadixArk/glm47-flash-blockwise-fp8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "RadixArk/glm47-flash-blockwise-fp8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "RadixArk/glm47-flash-blockwise-fp8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/RadixArk/glm47-flash-blockwise-fp8

SGLang

How to use RadixArk/glm47-flash-blockwise-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 "RadixArk/glm47-flash-blockwise-fp8" \
    --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": "RadixArk/glm47-flash-blockwise-fp8",
		"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 "RadixArk/glm47-flash-blockwise-fp8" \
        --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": "RadixArk/glm47-flash-blockwise-fp8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use RadixArk/glm47-flash-blockwise-fp8 with Docker Model Runner:
```
docker model run hf.co/RadixArk/glm47-flash-blockwise-fp8
```

GLM-4.7-Flash blockwise FP8

Blockwise FP8 quantization of zai-org/GLM-4.7-Flash: e4m3 weights with one scale per 128×128 block and dynamic activation scaling, about half the size of the BF16 original (30 GiB vs 58 GiB). See the original model card for model details.

Quantization

FP8: the linear layers of attention, the dense MLP, and all routed and shared experts, including the MTP layer.
BF16: embeddings, lm_head, norms, and the MoE router.

Evaluation

gsm8k (5-shot, full test set) on SGLang with MTP speculative decoding: 0.809 for FP8 vs 0.819 for BF16; average accept length 2.41 vs 2.43.

Usage (SGLang)

python -m sglang.launch_server --model-path RadixArk/glm47-flash-blockwise-fp8 --tp-size 2 \
  --speculative-algorithm EAGLE --speculative-num-steps 2 \
  --speculative-eagle-topk 1 --speculative-num-draft-tokens 3

Attention tensor parallelism must be 1 or 2: at 4, each rank's kv_b_proj shard (2240 rows) is not a multiple of the 128-row block. For more GPUs, use DP attention with expert parallelism, for example --tp-size 4 --dp-size 4 --enable-dp-attention --ep-size 4 --moe-a2a-backend deepep --cuda-graph-max-bs-decode 128.

License

MIT, same as the original model by Z.ai.

Downloads last month: 58

Safetensors

Model size

31B params

Tensor type

F32

BF16

F8_E4M3

Model tree for RadixArk/glm47-flash-blockwise-fp8

Base model

zai-org/GLM-4.7-Flash

Quantized

(111)

this model