Text Generation
Transformers
Safetensors
qwen3_5
image-text-to-text
qwen3.5
korean
essay-evaluation
rationale-generation
conversational
Instructions to use davemaxuellkr/260814-writer-model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use davemaxuellkr/260814-writer-model with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="davemaxuellkr/260814-writer-model") 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("davemaxuellkr/260814-writer-model") model = AutoModelForMultimodalLM.from_pretrained("davemaxuellkr/260814-writer-model", 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 davemaxuellkr/260814-writer-model with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "davemaxuellkr/260814-writer-model" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "davemaxuellkr/260814-writer-model", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/davemaxuellkr/260814-writer-model
- SGLang
How to use davemaxuellkr/260814-writer-model 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 "davemaxuellkr/260814-writer-model" \ --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": "davemaxuellkr/260814-writer-model", "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 "davemaxuellkr/260814-writer-model" \ --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": "davemaxuellkr/260814-writer-model", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use davemaxuellkr/260814-writer-model with Docker Model Runner:
docker model run hf.co/davemaxuellkr/260814-writer-model
| """Parse the writer's first balanced JSON object without repairing its text. | |
| This mirrors the parser used for the checkpoint's reported 4.3926 evaluation: | |
| Markdown or prose around a complete JSON object is tolerated, but malformed or | |
| truncated JSON is not repaired and is not retried. | |
| """ | |
| from __future__ import annotations | |
| import argparse | |
| import json | |
| import math | |
| import sys | |
| from pathlib import Path | |
| from typing import Any | |
| CATEGORIES = ("content", "organization", "expression") | |
| def extract_first_json_object(text: str) -> str | None: | |
| """Return the first balanced JSON-object substring in *text*.""" | |
| start = text.find("{") | |
| if start == -1: | |
| return None | |
| depth = 0 | |
| in_string = False | |
| escaped = False | |
| for index in range(start, len(text)): | |
| char = text[index] | |
| if in_string: | |
| if escaped: | |
| escaped = False | |
| elif char == "\\": | |
| escaped = True | |
| elif char == '"': | |
| in_string = False | |
| continue | |
| if char == '"': | |
| in_string = True | |
| elif char == "{": | |
| depth += 1 | |
| elif char == "}": | |
| depth -= 1 | |
| if depth == 0: | |
| return text[start : index + 1] | |
| return None | |
| def coerce_score(value: object) -> int: | |
| """Return a 1..5 integer, applying historical half-up float rounding.""" | |
| if isinstance(value, bool): | |
| raise ValueError(f"invalid boolean score: {value!r}") | |
| if isinstance(value, int): | |
| score = value | |
| elif isinstance(value, float) and math.isfinite(value): | |
| score = math.floor(value + 0.5) | |
| else: | |
| raise ValueError(f"score is not numeric: {value!r}") | |
| if not 1 <= score <= 5: | |
| raise ValueError(f"score is outside 1..5: {value!r}") | |
| return score | |
| def parse_writer_output(text: str) -> dict[str, dict[str, Any]]: | |
| """Extract and validate the writer's nested content/organization/expression JSON.""" | |
| candidate = extract_first_json_object(text) | |
| if candidate is None: | |
| raise ValueError("no balanced JSON object found") | |
| try: | |
| parsed = json.loads(candidate) | |
| except json.JSONDecodeError as error: | |
| raise ValueError(f"invalid JSON: {error}") from error | |
| if not isinstance(parsed, dict): | |
| raise ValueError("top-level JSON is not an object") | |
| validated: dict[str, dict[str, Any]] = {} | |
| for category in CATEGORIES: | |
| block = parsed.get(category) | |
| if not isinstance(block, dict): | |
| raise ValueError(f"{category}: expected an object") | |
| if "score" not in block or "rationale" not in block: | |
| raise ValueError(f"{category}: missing score/rationale") | |
| rationale = block["rationale"] | |
| if not isinstance(rationale, str) or not rationale.strip(): | |
| raise ValueError(f"{category}: rationale must be a non-empty string") | |
| validated[category] = { | |
| "score": coerce_score(block["score"]), | |
| "rationale": rationale, | |
| } | |
| return validated | |
| def main() -> None: | |
| parser = argparse.ArgumentParser(description=__doc__) | |
| parser.add_argument( | |
| "path", | |
| nargs="?", | |
| type=Path, | |
| help="raw model-output file; omit to read UTF-8 text from stdin", | |
| ) | |
| args = parser.parse_args() | |
| raw = args.path.read_text(encoding="utf-8") if args.path else sys.stdin.read() | |
| try: | |
| parsed = parse_writer_output(raw) | |
| except ValueError as error: | |
| raise SystemExit(f"parse failure: {error}") from error | |
| print(json.dumps(parsed, ensure_ascii=False, indent=2)) | |
| if __name__ == "__main__": | |
| main() | |