NVLM-D-72B

Home/Models/nvidia/NVLM-D-72B

nvidia/NVLM-D-72B

kind: original
quant: bf16 (16-bit)
family: misc
modality: image-text-to-text
root model: nvidia/NVLM-D-72B
provider: nvidia
license: cc-by-nc-4.0
language: en
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 178,174
likes: 776
created: —
last modified: 2025-01-14T18:34:52+00:00 (2025)
huggingface: https://huggingface.co/nvidia/NVLM-D-72B
tags: transformerssafetensorsNVLM_DnvidiaNVLMpytorchmultimodalconversationalimage-text-to-textcustom_codeenarxiv:2409.11402license:cc-by-nc-4.0endpoints_compatibleregion:us

benchmark	category	score	conf	evidence
DocVQA	multimodal	92.6	low	`\| NVLM-D 1.0 72B (Huggingface) \| 58.7 / 54.9 \| 65.2 \| 852 \| 94.2 \| 86.0 \| 92.6 \| 82.6 \| 69.5 \| 85.4 \|`
GSM8K	math	90.8	low	`\| Gemini Pro 1.5 (Aug 2024) \| N/A \| 85.9 \| 90.8 \| 67.7 \| 84.1 \| 82.1 \|`
HumanEval	coding	90.2	low	`\| GPT-4.0 \| N/A \| 88.7 \| - \| 76.6 \| 90.2 \| - \|`
MMLU	knowledge	88.7	low	`\| GPT-4.0 \| N/A \| 88.7 \| - \| 76.6 \| 90.2 \| - \|`
ChartQA	multimodal	86.0	low	`\| NVLM-D 1.0 72B (Huggingface) \| 58.7 / 54.9 \| 65.2 \| 852 \| 94.2 \| 86.0 \| 92.6 \| 82.6 \| 69.5 \| 85.4 \|`
MATH	math	76.6	low	`\| GPT-4.0 \| N/A \| 88.7 \| - \| 76.6 \| 90.2 \| - \|`
MathVista	multimodal	65.2	low	`\| NVLM-D 1.0 72B (Huggingface) \| 58.7 / 54.9 \| 65.2 \| 852 \| 94.2 \| 86.0 \| 92.6 \| 82.6 \| 69.5 \| 85.4 \|`

none scraped.