Nemotron-Research-GooseReason-4B-Instruct

Home/Models/nvidia/Nemotron-Research-GooseReason-4B-Instruct

nvidia/Nemotron-Research-GooseReason-4B-Instruct

kind: original
quant: bf16 (16-bit)
family: nemotron
modality: text-generation
root model: Qwen/Qwen3-4B-Instruct-2507
provider: nvidia
license: cc-by-nc-4.0
language: en
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 103
likes: 8
created: —
last modified: 2026-03-01T13:58:55+00:00 (2026)
huggingface: https://huggingface.co/nvidia/Nemotron-Research-GooseReason-4B-Instruct
tags: transformerssafetensorsqwen3text-generationreasoningrlvrmathcodestemnvidiaconversationalenarxiv:2601.22975arxiv:2505.24864base_model:Qwen/Qwen3-4B-Instruct-2507base_model:finetune:Qwen/Qwen3-4B-Instruct-2507license:cc-by-nc-4.0text-generation-inferenceendpoints_compatibleregion:us

benchmark	category	score	conf	evidence
MATH	math	94.66	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 64.79 \| 48.75 \| 85.17 \| 94.66 \| 50.09 \| 65.83 \| 68.21 \|`
HumanEval	coding	77.56	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 47.01 \| 42.08 \| 33.69 \| 23.69 \| 77.56 \| 31.74 \| 42.63 \|`
IFEval	instruction-following	72.36	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 60.26 \| 72.36 \| 43.69 \| 19.46 \| 34.92 \| 57.26 \| 33.98 \|`
AIME 2024	math	64.79	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 64.79 \| 48.75 \| 85.17 \| 94.66 \| 50.09 \| 65.83 \| 68.21 \|`
GPQA	reasoning	60.26	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 60.26 \| 72.36 \| 43.69 \| 19.46 \| 34.92 \| 57.26 \| 33.98 \|`
AIME 2025	math	48.75	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 64.79 \| 48.75 \| 85.17 \| 94.66 \| 50.09 \| 65.83 \| 68.21 \|`
CodeContests	coding	42.08	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 47.01 \| 42.08 \| 33.69 \| 23.69 \| 77.56 \| 31.74 \| 42.63 \|`
Codeforces	coding	33.69	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 47.01 \| 42.08 \| 33.69 \| 23.69 \| 77.56 \| 31.74 \| 42.63 \|`
LiveCodeBench	coding	31.74	low	`\| Qwen3‑4B‑Instruct \| — \| — \| 47.01 \| 42.08 \| 33.69 \| 23.69 \| 77.56 \| 31.74 \| 42.63 \|`

none scraped.