Nemotron-H-4B-Instruct-128K

Home/Models/nvidia/Nemotron-H-4B-Instruct-128K

nvidia/Nemotron-H-4B-Instruct-128K

kind: original
quant: bf16 (16-bit)
family: nemotron
modality: text-generation
root model: nvidia/Nemotron-H-4B-Base-8K
provider: nvidia
license: other
language: en
context: 128K
library: transformers
vRAM (GB): —
size on disk: —
downloads: 1,535
likes: 10
created: —
last modified: 2025-10-24T00:34:31+00:00 (2025)
huggingface: https://huggingface.co/nvidia/Nemotron-H-4B-Instruct-128K
tags: transformerssafetensorsnvidiapytorchtext-generationconversationalenarxiv:2504.11409base_model:nvidia/Nemotron-H-4B-Base-8Kbase_model:finetune:nvidia/Nemotron-H-4B-Base-8Klicense:otherendpoints_compatibleregion:us

benchmark	category	score	conf	evidence
GSM8K	math	88.93	low	`\| 88.93 \| 76.4 \|`
MBPP	coding	78.6	low	`\| 78.6 \| 68.25 \| 76.2 \| 70.85 \|`
MATH-500	math	76.4	low	`\| 88.93 \| 76.4 \|`
IFEval	instruction-following	76.24	low	`\| 7.9 \| 76.24 \|`
HumanEval	coding	76.2	low	`\| 78.6 \| 68.25 \| 76.2 \| 70.85 \|`
MT-Bench	chat	7.9	low	`\| 7.9 \| 76.24 \|`

none scraped.