Instructions to use incoai/GLM-5.3-DFlash2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use incoai/GLM-5.3-DFlash2 with Transformers:

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

pipe = pipeline("text-generation", model="incoai/GLM-5.3-DFlash2")

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

tokenizer = AutoTokenizer.from_pretrained("incoai/GLM-5.3-DFlash2")
model = AutoModel.from_pretrained("incoai/GLM-5.3-DFlash2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use incoai/GLM-5.3-DFlash2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "incoai/GLM-5.3-DFlash2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "incoai/GLM-5.3-DFlash2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/incoai/GLM-5.3-DFlash2

SGLang

How to use incoai/GLM-5.3-DFlash2 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 "incoai/GLM-5.3-DFlash2" \
    --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": "incoai/GLM-5.3-DFlash2",
		"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 "incoai/GLM-5.3-DFlash2" \
        --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": "incoai/GLM-5.3-DFlash2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use incoai/GLM-5.3-DFlash2 with Docker Model Runner:
```
docker model run hf.co/incoai/GLM-5.3-DFlash2
```

very slow

by botsfs - opened Sep 16

Discussion

botsfs

Sep 16

(APIServer pid=1) INFO 09-16 01:03:48 [loggers.py:310] Engine 000: Avg prompt throughput: 3908.9 tokens/s, Avg generation throughput: 82.9 tokens/s, Running: 2 reqs, Waiting: 0 reqs, GPU KV cache usage: 5.6%, Prefix cache hit rate: 73.9%
(APIServer pid=1) INFO 09-16 01:03:48 [metrics.py:120] SpecDecoding metrics: Mean acceptance length: 1.13, Accepted throughput: 9.50 tokens/s, Drafted throughput: 510.87 tokens/s, Accepted: 95 tokens, Drafted: 5110 tokens, Per-position acceptance rate: 0.105, 0.023, 0.001, 0.000, 0.000, 0.000, 0.000, Avg Draft acceptance rate: 1.9%

botsfs changed discussion title from why i am slow to very slow Sep 16

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment