ValiantLabs/gpt-oss-20b-ShiningValiant3

Support our open-source dataset and model releases!

Shining Valiant 3: Qwen3-1.7B, Qwen3-4B, Qwen3-8B, gpt-oss-20b

Shining Valiant 3 is a science, AI design, and general reasoning specialist built on gpt-oss-20b.

• Finetuned on our newest science reasoning data generated with Deepseek R1 0528!
• AI to build AI: our high-difficulty AI reasoning data makes Shining Valiant 3 your friend for building with current AI tech and discovering new innovations and improvements!
• Improved general and creative reasoning to supplement problem-solving and general chat performance.
• Small model sizes allow running on local desktop and mobile, plus super-fast server inference!

Prompting Guide

Shining Valiant 3 uses the gpt-oss-20b prompt format.

Shining Valiant 3 is a reasoning finetune; reasoning level high is generally recommended.

NOTE: This release of Shining Valiant 3 currently uses bf16 for all parameters. Consider quantized models if you're not looking to use bf16.

Example inference script provided by gpt-oss-20b to get started:

from transformers import pipeline
import torch

model_id = "ValiantLabs/gpt-oss-20b-ShiningValiant3"

pipe = pipeline(
    "text-generation",
    model=model_id,
    torch_dtype="auto",
    device_map="auto",
)

messages = [
    {"role": "user", "content": "Reversible Cellular Automata (RCAs) are CAs that have an inverse rule, allowing the simulation to run backward in time. Explain the theoretical significance of RCAs in the context of modeling physical laws that are time-symmetric. Describe the additional constraints that must be placed on a rule set to ensure it is reversible and discuss the challenges in constructing non-trivial reversible rules."},
]

outputs = pipe(
    messages,
    max_new_tokens=12000,
)
print(outputs[0]["generated_text"][-1])

Shining Valiant 3 is created by Valiant Labs.

Check out our HuggingFace page to see all of our models!

We care about open source. For everyone to use.