Text Classification
Transformers
Safetensors
English
qwen3_5_text
text-generation
decider
systemone
jev
decision-model
qwen3.5
post-train
text-to-typed-output
structured-output
classification
tool-calling
routing
moderation
custom-code
Eval Results (legacy)
Instructions to use Jeetu456/decider-2b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Jeetu456/decider-2b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="Jeetu456/decider-2b")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Jeetu456/decider-2b") model = AutoModelForCausalLM.from_pretrained("Jeetu456/decider-2b", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download training/clef_engine.py from Jeetu456/decider-2b: direct link, hf CLI and curl.
- Browser
- Download file 1.33 kB
-
https://huggingface.co/Jeetu456/decider-2b/resolve/main/training/clef_engine.py
- Command line
-
hf download hf://Jeetu456/decider-2b/training/clef_engine.py
-
curl -L -o clef_engine.py https://huggingface.co/Jeetu456/decider-2b/resolve/main/training/clef_engine.py
1.33 kB
| """Decision Index engine for Cloudflare/clef-flash, so it runs on our exact suite and hardware. | |
| Run with: python -m decision_index run --engine clef_engine:ClefEngine ... | |
| The model card reports Decision Index 0.2.1 from Cloudflare's own run; our suite is 0.3 | |
| without HLE. Comparing on the same harness is the only fair way to say "we match clef-flash". | |
| """ | |
| import sys | |
| import torch | |
| from decision_index.engines import Engine | |
| class ClefEngine(Engine): | |
| name = "clef-flash" | |
| def __init__(self, repo="Cloudflare/clef-flash", **options): | |
| super().__init__(**options) | |
| from huggingface_hub import snapshot_download | |
| path = snapshot_download(repo) | |
| sys.path.insert(0, path) | |
| from joint_schema_model import load_release_model, systemone | |
| self.model, self.processor = load_release_model(path, device="cuda") | |
| self.systemone = systemone | |
| self.provenance = {"repo": repo, "path": path} | |
| def synchronize(self): | |
| torch.cuda.synchronize() | |
| def __call__(self, state, questions): | |
| with torch.inference_mode(): | |
| response = self.systemone(self.model, self.processor, | |
| {"model": "clef-flash", "state": state, "questions": questions}) | |
| return {"model": self.name, "answers": response["answers"]}, None | |