Mirrored from perplexity-ai/pplx-decider-v1.1-27b
All models by perplexity-ai →pplx-decider-v1.1-27b is an update of our pplx-decider-v1-27b model, that boost the Decision Index score from 56.4 to 61.56, now outperforming Jev by more than 3.5 points, while using the same backbone. Most of the gains are driven by the lifting of the causal mask and training on more data, including data from tasksource.
The overall score uses the suite's weighting, not a simple category average.
This is the native decision-checkpoint layout: a Qwen3_5Model backbone
plus readout.safetensors containing a BF16 [255, 5120] decision head.
The Transformers class name does not change the base model identity, Qwen3.8-27B.
Use the included inference implementation to preserve the evaluated behavior:
DecisionModel.predict applies the saved calibration temperature.
When using logits directly, apply it exactly once and normalize over the valid
candidates for the current decision.lm_head.
A serving system requiring Qwen3_5ForConditionalGeneration needs a separate
export, mapping readout row i into vocabulary row
decision_config.json["token_ids"][i]. Such an export must also preserve the
attention behavior. If the frontend applies temperature, use the value above.Use Python 3.12+, authenticated Hugging Face access to this private repository,
and a CUDA GPU with room for roughly 49 GiB of weights plus working memory.
Download the repository and install its pinned requirements.txt dependencies.
Pass images as file paths, PIL images, or data:image/...;base64 URLs in
row["images"]. The processor resizes them while preserving aspect ratio.
Each side is rounded to a multiple of 32 px, and the total area is kept between
256×256 (65,536 px) and max_image_pixels. Each 32×32 patch costs one visual token,
so an image uses about width * height / 1024 tokens.
max_image_pixels defaults to 2048×2048 (4,194,304 px, ≤4,096 tokens per image).
Larger images are downscaled automatically, so you don't need to resize them
yourself. To lower or raise the cap:
Keep in mind:
prepare's 8,192-token
limit. Inputs are rejected rather than truncated. With several images per
decision, lower max_image_pixels or downscale the images first.max_image_pixels=512 * 512.min(1, sqrt(max_image_pixels / (width * height))) to keep the aspect ratio.source/src/autojev/model.py is the evaluated training-run code, with one change:
the image-size cap is now the configurable max_image_pixels. It was previously
hardcoded at 512×512 pixels. release-manifest.json records file checksums and
checkpoint provenance.
A large part of the gains are thanks to the tasksource data. If you also use this data, consider citing the corresponding article
import sysfrom pathlib import Pathfrom huggingface_hub import snapshot_download
checkpoint = Path(snapshot_download("perplexity-ai/pplx-decider-v1.1-27b"))sys.path.insert(0, str(checkpoint / "source" / "src"))from autojev.model import DecisionModel, answer
model = DecisionModel(checkpoint, device="cuda")question = { "type": "choice", "instructions": "Which team should handle this request?", "criteria": { "billing": "Charges and refunds", "support": "Technical integration errors", "sales": "Questions about buying a product", },}row = {"state": "My integration keeps failing. Please help.", "question": question}probabilities = model.predict([row])[0]print(answer(question, probabilities))model = DecisionModel(checkpoint, device="cuda", max_image_pixels=1600 * 1000)@inproceedings{sileo-2024-tasksource, title = "tasksource: A Large Collection of {NLP} tasks with a Structured Dataset Preprocessing Framework", author = "Sileo, Damien", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.1361", pages = "15655--15684",}