Instructions to use Hcompany/Holo2-235B-A22B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Hcompany/Holo2-235B-A22B with Transformers:

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

pipe = pipeline("image-text-to-text", model="Hcompany/Holo2-235B-A22B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("Hcompany/Holo2-235B-A22B")
model = AutoModelForImageTextToText.from_pretrained("Hcompany/Holo2-235B-A22B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Hcompany/Holo2-235B-A22B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Hcompany/Holo2-235B-A22B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Hcompany/Holo2-235B-A22B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Hcompany/Holo2-235B-A22B

SGLang

How to use Hcompany/Holo2-235B-A22B 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 "Hcompany/Holo2-235B-A22B" \
    --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": "Hcompany/Holo2-235B-A22B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "Hcompany/Holo2-235B-A22B" \
        --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": "Hcompany/Holo2-235B-A22B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Hcompany/Holo2-235B-A22B with Docker Model Runner:
```
docker model run hf.co/Hcompany/Holo2-235B-A22B
```

Holo2-235B-A22B / .eval_results /screenspot_pro_agentic.yaml

marc-thibault-h

Add ScreenSpot-Pro evaluation result (Holo2-235B-A22B (Agentic)) (#1)

f7f85b1 3 months ago

raw

history blame contribute delete

6.19 kB

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: overall
	value: 78.5
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: android_studio_macos
	value: 53.8
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: autocad_windows
	value: 76.5
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: blender_windows
	value: 70.4
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: davinci_macos
	value: 72.7
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: eviews_windows
	value: 82.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: excel_macos
	value: 76.6
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: fruitloops_windows
	value: 66.7
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: illustrator_windows
	value: 83.9
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: inventor_windows
	value: 80.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: linux_common_linux
	value: 82.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: macos_common_macos
	value: 87.7
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: matlab_macos
	value: 86.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: origin_windows
	value: 75.8
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: photoshop_windows
	value: 88.2
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: powerpoint_windows
	value: 86.6
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: premiere_windows
	value: 65.4
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: pycharm_macos
	value: 80.8
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: quartus_windows
	value: 71.1
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: solidworks_windows
	value: 81.8
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: stata_windows
	value: 71.4
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: unreal_engine_windows
	value: 60.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: vivado_windows
	value: 81.2
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: vmware_macos
	value: 78.0
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: vscode_macos
	value: 96.4
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: windows_common_windows
	value: 80.2
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"

	- dataset:
	id: likaixin/ScreenSpot-Pro
	task_id: word_macos
	value: 90.5
	source:
	url: https://gui-agent.github.io/grounding-leaderboard/
	name: ScreenSpot-Pro Leaderboard
	user: merve
	notes: "Agentic"