Zero-Shot Classification
Transformers
Safetensors
qwen3_5
feature-extraction
decision-model
classification
system-one
multimodal
vision
video
custom_code
Eval Results
Instructions to use vllm-sr/d3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use vllm-sr/d3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("zero-shot-classification", model="vllm-sr/d3", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModel processor = AutoProcessor.from_pretrained("vllm-sr/d3", trust_remote_code=True) model = AutoModel.from_pretrained("vllm-sr/d3", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Add LEXam (mcq_4_choices) eval result
#5
by HappyWindman555 - opened
- .eval_results/lexam.yaml +10 -0
.eval_results/lexam.yaml
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
- dataset:
|
| 2 |
+
id: LEXam-Benchmark/LEXam
|
| 3 |
+
task_id: mcq_4_choices
|
| 4 |
+
revision: 31e57ae395f92ed7284a4c28d278c55da892b898
|
| 5 |
+
value: 61.75
|
| 6 |
+
date: "2026-10-17"
|
| 7 |
+
source:
|
| 8 |
+
url: https://huggingface.co/datasets/HappyWindman555/d3-hub-evals
|
| 9 |
+
name: Decision 3.0 Hub benchmark run 2026-10-17
|
| 10 |
+
notes: "Decision-model protocol, not generative: one /v1/systemone request per item with a single choice question (lettered options with their text); prediction = argmax of the returned option probabilities, no sampling, no CoT. All 1655 items, 1655 answered, 0 errors. vLLM nightly (0.32.0rc1.dev190+g3f1bcfd82), --max-model-len 8192 --enforce-eager, 1x A100 80GB, BF16."
|