Instructions to use jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027 with Transformers:

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

pipe = pipeline("image-text-to-text", model="jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027")
model = AutoModelForMultimodalLM.from_pretrained("jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027", 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=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027",
		"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/jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027

SGLang

How to use jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027 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 "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027" \
    --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": "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027",
		"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 "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027" \
        --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": "jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027",
		"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 jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027 with Docker Model Runner:
```
docker model run hf.co/jjakgui/gemma3-4b-v-KoV_2.0.1_fft_20251027
```

gemma3-4b-v-KoV_2.0.1_fft_20251027 / README.md

yangdonghun3

Upload folder using huggingface_hub

0ff7602 verified 11 months ago

preview code

Raw

History Blame Contribute Delete

4.36 kB

	---
	library_name: transformers
	license: gemma
	base_model: google/gemma-3-4b-it
	tags:
	- generated_from_trainer
	datasets:
	- vlm_data_2025101_1/gemma3-4b-v-KoV_2.0.1.jsonl
	model-index:
	- name: outputs/gemma3-4b-v-KoV_2.0.1_fft.jsonl
	results: []
	---

	<!-- This model card has been generated automatically according to the information the Trainer had access to. You
	should probably proofread and complete it, then remove this comment. -->

	[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
	<details><summary>See axolotl config</summary>

	axolotl version: `0.12.2`
	```yaml
	# ===== Model =====

	base_model: google/gemma-3-4b-it
	processor_type: AutoProcessor

	chat_template: gemma3

	# 멀티모달(비전-챗) 필수 플래그
	skip_prepare_dataset: true
	remove_unused_columns: false
	sample_packing: false

	#shuffle_merged_datasets: false
	#shuffle_before_merging_datasets: false # (기본 false지만 명시 추천)

	ddp_find_unused_parameters: true

	dataloader_num_workers: 0

	# ===== Data =====
	eot_tokens:
	- <end_of_turn>
	datasets:
	- path: vlm_data_2025101_1/gemma3-4b-v-KoV_2.0.1.jsonl
	type: chat_template
	field_messages: messages
	split: null

	val_set_size: 0.0
	dataset_prepared_path:

	# ===== Output / Logging =====
	output_dir: ./outputs/gemma3-4b-v-KoV_2.0.1_fft.jsonl
	logging_steps: 1

	# wandb 연동(원하면 변경/주석)
	wandb_entity: minkyun1
	wandb_project: kisti_vlm_axo
	wandb_name: gemma3-4b-v-KoV_2.0.1_fft.jsonl

	# ===== LoRA / Quantization =====
	#adapter: lora
	# LLaVA에서 언어모델 쪽 프로젝션에만 LoRA(안전 기본값)
	#lora_r: 256
	#lora_alpha: 512
	#lora_dropout: 0.05
	#lora_target_modules: "model.language_model.layers.[\\d]+.(mlp\|cross_attn\|self_attn).(up\|down\|gate\|q\|k\|v\|o)_proj"

	# 메모리 여유 충분하지만, 시작은 4bit 로 안정적으로
	load_in_4bit: false
	load_in_8bit: false
	bf16: true
	tf32: true
	gradient_checkpointing: true
	gradient_checkpointing_kwargs:
	use_reentrant: false
	flash_attention: true
	eager_attention:

	# ===== Optim & Train =====
	optimizer: adamw_torch_fused
	learning_rate: 4e-5
	lr_scheduler: cosine
	warmup_ratio: 0.05
	weight_decay: 0.01
	max_grad_norm: 1.0
	seed: 42
	sequence_len: 8192
	pad_to_sequence_len: false
	excess_length_strategy: drop

	# GPU당 마이크로 배치/누적 → 유효 배치 = 1 * 8 * 2GPU = 16
	micro_batch_size: 1
	gradient_accumulation_steps: 16

	num_epochs: 3
	evals_per_epoch: 1
	saves_per_epoch: 1
	# save_first_step: true

	# ===== Multi-GPU: DeepSpeed (추천) =====
	# deepspeed 프리셋을 받아서 사용:
	# axolotl fetch deepspeed_configs
	# 2×A100 80GB + 7B에는 zero2가 빠르고 안정적
	deepspeed: ds_zero2.json

	# ===== 디버그/재현성(선택) =====
	# 데이터 전처리 멀티프로세스가 문제 생기면 1로 낮춰서 원인 파악
	# dataset_processes: 1

	# ===== [대안] FSDP2 설정(DeepSpeed 대신 쓰고 싶을 때) =====
	# fsdp_version: 2
	# fsdp_config:
	# offload_params: false
	# cpu_ram_efficient_loading: true
	# auto_wrap_policy: TRANSFORMER_BASED_WRAP
	# transformer_layer_cls_to_wrap: LlamaDecoderLayer
	# state_dict_type: FULL_STATE_DICT
	# reshard_after_forward: true

	```

	</details><br>

	# outputs/gemma3-4b-v-KoV_2.0.1_fft.jsonl

	This model is a fine-tuned version of [google/gemma-3-4b-it](https://huggingface.co/google/gemma-3-4b-it) on the vlm_data_2025101_1/gemma3-4b-v-KoV_2.0.1.jsonl dataset.

	## Model description

	More information needed

	## Intended uses & limitations

	More information needed

	## Training and evaluation data

	More information needed

	## Training procedure

	### Training hyperparameters

	The following hyperparameters were used during training:
	- learning_rate: 4e-05
	- train_batch_size: 1
	- eval_batch_size: 1
	- seed: 42
	- distributed_type: multi-GPU
	- num_devices: 8
	- gradient_accumulation_steps: 16
	- total_train_batch_size: 128
	- total_eval_batch_size: 8
	- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
	- lr_scheduler_type: cosine
	- lr_scheduler_warmup_steps: 578
	- training_steps: 11567

	### Training results



	### Framework versions

	- Transformers 4.55.2
	- Pytorch 2.6.0+cu124
	- Datasets 4.0.0
	- Tokenizers 0.21.4