Nemotron-H-56B-Base-8K

Home/Models/nvidia/Nemotron-H-56B-Base-8K

nvidia/Nemotron-H-56B-Base-8K

kind: original
quant: bf16 (16-bit)
family: nemotron
modality: text-generation
root model: nvidia/Nemotron-H-56B-Base-8K
provider: nvidia
license: other
language: en, de, es, fr, it, ko, pt, ru, jp, zh
context: 8K
library: transformers
vRAM (GB): —
size on disk: —
downloads: 16,431
likes: 33
created: —
last modified: 2025-08-21T15:45:48+00:00 (2025)
huggingface: https://huggingface.co/nvidia/Nemotron-H-56B-Base-8K
tags: transformerssafetensorsnvidiapytorchnemotron-htext-generationendeesfritkoptrujpzharxiv:2504.03624license:othereval-resultsendpoints_compatibleregion:us

benchmark	category	score	context	conf	evidence
GSM8K	math	93.71	GSM8K 8-shot CoT \| MATH 4-shot CoT \| MATH-Lvl 5 4-shot CoT \| MATH-500 4-shot CoT	low	`\| 93.71 \| 59.42 \| 35.19 \| 57.37 \|`
CommonsenseQA	knowledge	86.73		low	`\| 94.97 \| 89.00 \| 84.45 \| 86.73 \|`
WinoGrande	knowledge	84.45		low	`\| 94.97 \| 89.00 \| 84.45 \| 86.73 \|`
MMLU	knowledge	84.21	MMLU-Pro 5-shot CoT	low	`\| 60.51 \| 84.21 \|`
MBPP	coding	77.82		low	`\| 77.82 \| 67.20 \| 60.37 \| 54.27 \|`
MMLU-Pro	knowledge	60.51	MMLU-Pro 5-shot CoT	low	`\| 60.51 \| 84.21 \|`
HumanEval	coding	60.37		low	`\| 77.82 \| 67.20 \| 60.37 \| 54.27 \|`
MATH	math	59.42	GSM8K 8-shot CoT \| MATH 4-shot CoT \| MATH-Lvl 5 4-shot CoT \| MATH-500 4-shot CoT	low	`\| 93.71 \| 59.42 \| 35.19 \| 57.37 \|`
MATH-500	math	57.37	GSM8K 8-shot CoT \| MATH 4-shot CoT \| MATH-Lvl 5 4-shot CoT \| MATH-500 4-shot CoT	low	`\| 93.71 \| 59.42 \| 35.19 \| 57.37 \|`
ARC-Challenge	knowledge	25.0		medium	`ARC Challenge 25`
HellaSwag	knowledge	10.0		medium	`Hellaswag 10`

none scraped.