Image-Text-to-Text
Transformers
Safetensors
English
qwen3_5
piko
piko-9b
multimodal
vision-language
hybrid-attention
linear-attention
ocr
document-understanding
conversational
Instructions to use Dexy2/Piko-9b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Dexy2/Piko-9b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="Dexy2/Piko-9b") 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("Dexy2/Piko-9b") model = AutoModelForMultimodalLM.from_pretrained("Dexy2/Piko-9b", 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 Dexy2/Piko-9b with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Dexy2/Piko-9b" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Dexy2/Piko-9b", "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/Dexy2/Piko-9b
- SGLang
How to use Dexy2/Piko-9b 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 "Dexy2/Piko-9b" \ --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": "Dexy2/Piko-9b", "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 "Dexy2/Piko-9b" \ --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": "Dexy2/Piko-9b", "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 Dexy2/Piko-9b with Docker Model Runner:
docker model run hf.co/Dexy2/Piko-9b
File size: 3,886 Bytes
12cd919 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 | #!/usr/bin/env python3
"""Strip machine-specific filesystem paths from anything destined for publication.
The provenance information in these reports is worth publishing; the author's
directory layout is not. This replaces concrete paths with neutral placeholders
while leaving checkpoint *names* intact, since those are the actual provenance
identifiers and are already public in config.json.
python scripts/sanitize_paths.py --root . --check
python scripts/sanitize_paths.py --root . --apply
"""
from __future__ import annotations
import argparse
import re
import sys
from pathlib import Path
# Ordered: longest / most specific first, so a broad rule cannot eat a narrow one.
REPLACEMENTS: list[tuple[re.Pattern[str], str]] = [
# Windows form, both single and JSON-escaped backslashes.
(
re.compile(r"E:\\{1,2}New folder \(4\)\\{1,2}wraithfast-9b\\{1,2}wraith-finetune"),
"<workspace>",
),
(re.compile(r"E:\\{1,2}New folder \(4\)"), "<workspace>"),
# POSIX / WSL form.
(re.compile(r"/mnt/e/New folder \(4\)/wraithfast-9b/wraith-finetune"), "<workspace>"),
(re.compile(r"/mnt/e/New folder \(4\)"), "<workspace>"),
(re.compile(r"/mnt/c/piko9b-weights"), "<local-checkpoint>"),
# Home directories and Windows user profiles, whoever they belong to.
(re.compile(r"/home/[A-Za-z0-9_.-]+"), "<home>"),
(re.compile(r"C:\\{1,2}Users\\{1,2}[A-Za-z0-9_.-]+"), "<home>"),
]
# Leftover bare drive/folder references that survive the rules above.
RESIDUAL = re.compile(r"New folder \(4\)|/mnt/[a-z]/|C:\\Users|/home/[a-z]")
SUFFIXES = {".md", ".json", ".py", ".yaml", ".yml", ".txt", ".cff", ".toml", ".jinja"}
SKIP_DIRS = {".git", "__pycache__", ".ruff_cache", ".pytest_cache", ".venv"}
def iter_files(root: Path):
for path in sorted(root.rglob("*")):
if not path.is_file() or path.suffix not in SUFFIXES:
continue
if any(part in SKIP_DIRS for part in path.parts):
continue
# Never rewrite this script's own rules.
if path.name == "sanitize_paths.py":
continue
yield path
def sanitize(text: str) -> tuple[str, int]:
total = 0
for pattern, replacement in REPLACEMENTS:
text, count = pattern.subn(replacement, text)
total += count
return text, total
def main() -> None:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--root", type=Path, default=Path("."))
parser.add_argument("--apply", action="store_true", help="Rewrite files in place.")
parser.add_argument("--check", action="store_true", help="Exit 1 if anything remains.")
args = parser.parse_args()
changed: list[tuple[Path, int]] = []
residual: list[str] = []
for path in iter_files(args.root):
original = path.read_text(encoding="utf-8", errors="replace")
cleaned, count = sanitize(original)
if count:
changed.append((path.relative_to(args.root), count))
if args.apply:
path.write_text(cleaned, encoding="utf-8")
final = cleaned if args.apply or count else original
for line_number, line in enumerate(final.splitlines(), start=1):
if RESIDUAL.search(line):
residual.append(
f"{path.relative_to(args.root)}:{line_number}: {line.strip()[:110]}"
)
verb = "sanitised" if args.apply else "would sanitise"
for path, count in changed:
print(f" {verb} {path} ({count} occurrence(s))")
print(f"\n{len(changed)} file(s) {verb}, {sum(c for _, c in changed)} replacement(s)")
if residual:
print(f"\n{len(residual)} residual reference(s) needing manual review:")
for entry in residual[:40]:
print(f" {entry}")
if args.check and residual:
sys.exit(1)
if __name__ == "__main__":
main()
|