Instructions to use game2player/Game2Player-4B-Thinking with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use game2player/Game2Player-4B-Thinking with Transformers:

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

pipe = pipeline("image-text-to-text", model="game2player/Game2Player-4B-Thinking")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("game2player/Game2Player-4B-Thinking")
model = AutoModelForMultimodalLM.from_pretrained("game2player/Game2Player-4B-Thinking", device_map="auto")
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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use game2player/Game2Player-4B-Thinking with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "game2player/Game2Player-4B-Thinking"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "game2player/Game2Player-4B-Thinking",
		"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/game2player/Game2Player-4B-Thinking

SGLang

How to use game2player/Game2Player-4B-Thinking 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 "game2player/Game2Player-4B-Thinking" \
    --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": "game2player/Game2Player-4B-Thinking",
		"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 "game2player/Game2Player-4B-Thinking" \
        --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": "game2player/Game2Player-4B-Thinking",
		"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 game2player/Game2Player-4B-Thinking with Docker Model Runner:
```
docker model run hf.co/game2player/Game2Player-4B-Thinking
```

Game2Player-4B-Thinking / README.md

feng0929

Add files using upload-large-folder tool

01709fb verified 2 days ago

preview code

Raw History Blame Contribute Delete

1.77 kB

---
license: apache-2.0
base_model: Qwen/Qwen3.5-4B
base_model_relation: finetune
library_name: transformers
pipeline_tag: image-text-to-text
tags:
- game2player
- sft
- thinking
---
# Game2Player-4B-Thinking

Final checkpoint after one epoch of SFT on game examples containing nonempty
assistant thinking blocks. Training mixes **DeepSeek and MiniMax** data:
14,336 selected DeepSeek-v4.1-flash records and 8,138 MiniMax-m3.1-flash records.
After preprocessing, 22,425 records remain (49 filtered). This is not a
DeepSeek-only model. Final training step: 157.

Full language-model fine-tuning with vision encoder and aligner frozen;
4 H200 GPUs, BF16, ZeRO-2, gradient checkpointing, 25K packing,
global batch 48, LR 1e-5, cosine schedule, 3% warmup, seed 3407.
Base revision: `851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a`.

The 4B run completed one epoch without recorded overflow warnings.
No downstream benchmark evaluation has been performed.

The repository contains inference weights and all tokenizer/processor assets,
including supplemental base tokenizer/video assets where absent from the
checkpoint. Optimizer states, RNG states, training data and images are not
included. Full training checkpoints remain stored locally.
See `training_summary.json` for provenance.

Load using a Transformers version supporting Qwen3.5 (the training environment
uses Transformers 5.16.1). Authenticate with an account authorized for this
private repository.

```python
import torch
from transformers import AutoProcessor, Qwen3_5ForConditionalGeneration
model_id = "game2player/Game2Player-4B-Thinking"
processor = AutoProcessor.from_pretrained(model_id)
model = Qwen3_5ForConditionalGeneration.from_pretrained(
    model_id, dtype=torch.bfloat16, device_map="auto"
)
```

	---
	license: apache-2.0
	base_model: Qwen/Qwen3.5-4B
	base_model_relation: finetune
	library_name: transformers
	pipeline_tag: image-text-to-text
	tags:
	- game2player
	- sft
	- thinking
	---

	# Game2Player-4B-Thinking

	Final checkpoint after one epoch of SFT on game examples containing nonempty
	assistant thinking blocks. Training mixes DeepSeek and MiniMax data:
	14,336 selected DeepSeek-v4.1-flash records and 8,138 MiniMax-m3.1-flash records.
	After preprocessing, 22,425 records remain (49 filtered). This is not a
	DeepSeek-only model. Final training step: 157.

	Full language-model fine-tuning with vision encoder and aligner frozen;
	4 H200 GPUs, BF16, ZeRO-2, gradient checkpointing, 25K packing,
	global batch 48, LR 1e-5, cosine schedule, 3% warmup, seed 3407.
	Base revision: `851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a`.

	The 4B run completed one epoch without recorded overflow warnings.
	No downstream benchmark evaluation has been performed.

	The repository contains inference weights and all tokenizer/processor assets,
	including supplemental base tokenizer/video assets where absent from the
	checkpoint. Optimizer states, RNG states, training data and images are not
	included. Full training checkpoints remain stored locally.
	See `training_summary.json` for provenance.

	Load using a Transformers version supporting Qwen3.5 (the training environment
	uses Transformers 5.16.1). Authenticate with an account authorized for this
	private repository.

	```python
	import torch
	from transformers import AutoProcessor, Qwen3_5ForConditionalGeneration

	model_id = "game2player/Game2Player-4B-Thinking"
	processor = AutoProcessor.from_pretrained(model_id)
	model = Qwen3_5ForConditionalGeneration.from_pretrained(
	model_id, dtype=torch.bfloat16, device_map="auto"
	)
	```