Instructions to use KKHYA/llavaqwen2.5-32b-finetune with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KKHYA/llavaqwen2.5-32b-finetune with Transformers:

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

pipe = pipeline("text-generation", model="KKHYA/llavaqwen2.5-32b-finetune")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("KKHYA/llavaqwen2.5-32b-finetune", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KKHYA/llavaqwen2.5-32b-finetune with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KKHYA/llavaqwen2.5-32b-finetune"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KKHYA/llavaqwen2.5-32b-finetune",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KKHYA/llavaqwen2.5-32b-finetune

SGLang

How to use KKHYA/llavaqwen2.5-32b-finetune 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 "KKHYA/llavaqwen2.5-32b-finetune" \
    --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": "KKHYA/llavaqwen2.5-32b-finetune",
		"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 "KKHYA/llavaqwen2.5-32b-finetune" \
        --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": "KKHYA/llavaqwen2.5-32b-finetune",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KKHYA/llavaqwen2.5-32b-finetune with Docker Model Runner:
```
docker model run hf.co/KKHYA/llavaqwen2.5-32b-finetune
```

llavaqwen2.5-32b-finetune

File size: 1,324 Bytes

{
  "architectures": [
    "LlavaQwen2ForCausalLM"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "eos_token_id": 151643,
  "freeze_mm_mlp_adapter": false,
  "hidden_act": "silu",
  "hidden_size": 5120,
  "image_aspect_ratio": "pad",
  "image_projector_type": "mlp2x_gelu",
  "initializer_range": 0.02,
  "intermediate_size": 27648,
  "max_position_embeddings": 131072,
  "max_window_layers": 64,
  "mm_hidden_size": 1024,
  "mm_image_tower": "openai/clip-vit-large-patch14-336",
  "mm_projector_lr": null,
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_video_tower": null,
  "mm_vision_select_feature": "patch",
  "mm_vision_select_layer": -2,
  "model_type": "llava_qwen2",
  "num_attention_heads": 40,
  "num_hidden_layers": 64,
  "num_key_value_heads": 8,
  "pad_token_id": 151665,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 1000000.0,
  "sliding_window": 131072,
  "tie_word_embeddings": false,
  "tokenizer_padding_side": "right",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.51.0",
  "tune_mm_mlp_adapter": false,
  "use_cache": true,
  "use_mm_proj": true,
  "use_sliding_window": false,
  "video_global_proj": false,
  "video_projector_type": "linear",
  "video_spatial_proj": false,
  "video_temproal_proj": false,
  "vocab_size": 152064
}