Image-Text-to-Text
Transformers
Safetensors
qwen2_5_vl
qwen3_vl
vision-language
visual-token-pruning
visionzip
scopd
Instructions to use enmingzhangzz/SCOPD with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use enmingzhangzz/SCOPD with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="enmingzhangzz/SCOPD")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("enmingzhangzz/SCOPD", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use enmingzhangzz/SCOPD with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "enmingzhangzz/SCOPD" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "enmingzhangzz/SCOPD", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/enmingzhangzz/SCOPD
- SGLang
How to use enmingzhangzz/SCOPD 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 "enmingzhangzz/SCOPD" \ --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": "enmingzhangzz/SCOPD", "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 "enmingzhangzz/SCOPD" \ --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": "enmingzhangzz/SCOPD", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use enmingzhangzz/SCOPD with Docker Model Runner:
docker model run hf.co/enmingzhangzz/SCOPD
Download evaluation/scripts/merge_adapter.py from enmingzhangzz/SCOPD: direct link, hf CLI and curl.
- Browser
- Download file 3.3 kB
-
https://huggingface.co/enmingzhangzz/SCOPD/resolve/main/evaluation/scripts/merge_adapter.py
- Command line
-
hf download hf://enmingzhangzz/SCOPD/evaluation/scripts/merge_adapter.py
-
curl -L -o merge_adapter.py https://huggingface.co/enmingzhangzz/SCOPD/resolve/main/evaluation/scripts/merge_adapter.py
3.3 kB
| """Merge a released adapter locally for evaluation; no training is performed.""" | |
| from __future__ import annotations | |
| import argparse | |
| import hashlib | |
| import json | |
| from pathlib import Path | |
| import sys | |
| ROOT = Path(__file__).resolve().parents[1] | |
| sys.path.insert(0, str(ROOT)) | |
| def main(): | |
| parser = argparse.ArgumentParser(description=__doc__) | |
| parser.add_argument('--adapter', type=Path, required=True) | |
| parser.add_argument('--base-model', help='Override with the matching pinned local Hub snapshot') | |
| parser.add_argument('--base-revision', help='Defaults to the adapter provenance revision') | |
| parser.add_argument('--output', type=Path, required=True) | |
| parser.add_argument('--device', choices=['cpu', 'cuda'], default='cpu') | |
| args = parser.parse_args() | |
| if args.output.exists() and any(args.output.iterdir()): | |
| raise FileExistsError('Refusing to overwrite an existing model') | |
| provenance = json.loads((args.adapter / 'provenance.json').read_text()) | |
| args.base_model = args.base_model or provenance['base_model'] | |
| args.base_revision = args.base_revision or provenance['base_revision'] | |
| digest = hashlib.sha256((args.adapter / 'adapter_model.safetensors').read_bytes()).hexdigest() | |
| if digest != provenance['adapter_sha256']: | |
| raise ValueError('Adapter SHA256 mismatch') | |
| if args.base_revision != provenance['base_revision']: | |
| raise ValueError('Base revision differs from adapter provenance') | |
| qwen3 = provenance['base_model'] == 'Qwen/Qwen3-VL-4B-Instruct' | |
| if qwen3: | |
| from scopd_eval.qwen3_runtime import activate | |
| else: | |
| from scopd_eval.runtime import activate | |
| activate() | |
| from huggingface_hub import snapshot_download | |
| import torch | |
| from peft import PeftModel | |
| from transformers import AutoProcessor | |
| if qwen3: | |
| from transformers import Qwen3VLForConditionalGeneration as Model | |
| else: | |
| from transformers import Qwen2_5_VLForConditionalGeneration as Model | |
| base = Path(args.base_model) | |
| if not base.is_dir(): | |
| if args.base_model != provenance['base_model']: | |
| raise ValueError('Base model differs from adapter provenance') | |
| base = Path(snapshot_download(args.base_model, revision=args.base_revision, | |
| allow_patterns=['*.json', '*.safetensors', '*.txt', '*.jinja'])) | |
| elif base.name != args.base_revision: | |
| raise ValueError('Local base must be the pinned Hub snapshot directory named by its revision') | |
| model = Model.from_pretrained( | |
| base, torch_dtype=torch.bfloat16, attn_implementation='eager', | |
| device_map={'': args.device}, local_files_only=True) | |
| model = PeftModel.from_pretrained(model, args.adapter, is_trainable=False) | |
| model = model.merge_and_unload(safe_merge=True) | |
| model.save_pretrained(args.output, safe_serialization=True, max_shard_size='5GB') | |
| AutoProcessor.from_pretrained(base, local_files_only=True).save_pretrained(args.output) | |
| (args.output / 'merge_manifest.json').write_text(json.dumps({ | |
| 'adapter_sha256': digest, 'base_model': provenance['base_model'], | |
| 'base_revision': args.base_revision, 'safe_merge': True, | |
| 'dtype': 'bfloat16'}, indent=2) + '\n') | |
| print(f'Merged model ready: {args.output}') | |
| if __name__ == '__main__': | |
| main() | |