Text Classification
Transformers
Safetensors
qwen3_5
image-text-to-text
neojev
jev
system-one
typed-decisions
classification
calibration
vision
fp8
Instructions to use dowands/neojev with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use dowands/neojev with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="dowands/neojev")# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("dowands/neojev") model = AutoModelForMultimodalLM.from_pretrained("dowands/neojev", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download jevbench_via_server.py from dowands/neojev: direct link, hf CLI and curl.
- Browser
- Download file 1.33 kB
-
https://huggingface.co/dowands/neojev/resolve/main/jevbench_via_server.py
- Command line
-
hf download hf://dowands/neojev/jevbench_via_server.py
-
curl -L -o jevbench_via_server.py https://huggingface.co/dowands/neojev/resolve/main/jevbench_via_server.py
1.33 kB
| """Run JevBench's own TypeSafe adapter (unmodified) against a NeoJev server on the 231 public items. | |
| usage: jevbench_via_server.py <endpoint> <out.jsonl>""" | |
| import json | |
| import os | |
| import sys | |
| ROOT = os.path.join(os.path.dirname(__file__), "..") | |
| JB = os.environ.get("JB_DIR", os.path.join(ROOT, "reviews", "jevbench")) | |
| sys.path.insert(0, JB) | |
| from jevbench.adapters.typesafe import TypeSafeAdapter # noqa: E402 | |
| from jevbench.scoring import score_task # noqa: E402 | |
| class T: | |
| def __init__(self, r): | |
| self.state, self.question, self.labels, self.expected = r["state"], r["question"], r["labels"], r["expected"] | |
| ep, out = sys.argv[1], sys.argv[2] | |
| ad = TypeSafeAdapter(endpoint=ep, model="neojev", key_env=None) | |
| rows = [] | |
| for tier in ["easy", "original", "hard"]: | |
| rows += [json.loads(l) for l in open(os.path.join(JB, "datasets", "public", f"{tier}.jsonl"))] | |
| ok = 0 | |
| with open(out, "w") as f: | |
| for r in rows: | |
| res = ad.run(T(r)) | |
| s = score_task(res.probs, T(r)) if res.ok else {"correct": False, "valid": False} | |
| ok += res.ok | |
| f.write(json.dumps({"id": r["id"], "tier": r["id"].split("-")[0], "ok": res.ok, "error": res.error, | |
| "latency_s": round(res.latency_s, 4), "correct": s["correct"], "probs": res.probs}) + "\n") | |
| print("requests ok", ok, "/", len(rows)) | |