OtterHD-8B
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Organisation
- Nanyang Technological University
- Organisation type
- Academia
- Country
- Singapore
- Published
- 7 November 2023
- Authors
- Bo Li, Peiyuan Zhang, Jingkang Yang, Yuanhan Zhang, Fanyi Pu, Ziwei Liu
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Multimodal, Vision, Language
- Task
- Chat, Visual question answering
- Base model
- Fuyu-8B
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Parameters
- 8B
- Training data
- 200,540,000 tokens
- Epochs
- 3
8B
"We compiled a total of 370K instruction/response pairs sourced from the follow- ing public datasets: LLaVA-Instruct [ 30], VQAv2 [2], GQA [ 23 ], OKVQA [ 36 ], OCRVQA [38 ], A-OKVQA [ 45], COCO-GOI [33 ], COCO-Caption [ 10], TextQA [ 48], RefCOCO [58], COCO-ITM [ 28 ], ImageNet [17 ], and LLaVA-RLHF [ 51 ]."
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Fine-tuning compute
- 2.4 × 10¹⁹ FLOP
flops = (8) * (312 * 10**12) * (3 * 3 * 3600) * (0.3) = 2.4e19 (num gpu) * (peak flops) * (time in seconds) * (assumed utilization rate) 'Our implementation permits the completion of full-parameter training within 3 hours per epoch on 8×A100 GPUs. ' Table 4 indicates 3 epochs for the full finetune.
The training run
What it physically took to train: which chips, how many, for how long, and what that drew from the wall.
- Training hardware
- NVIDIA A100
- Chips used
- 8
- Chip-hours
- 24
- Wall-clock time
- 9 hours
- Power draw
- 6.3 kW
3 hours per epoch over 3 epochs: 9 hours
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
- Citations
- 85
Sources
Where this record came from and when it was last checked.
- Reference
- OtterHD: A High-Resolution Multi-modality Model
- Last updated
- 25 May 2026
What the numbers mean
Background
OtterHD-8B was published by Nanyang Technological University, in the country recorded as Singapore, during November 2023. The publishing organisation is categorised as academia.
It works in the domain of Multimodal, Vision, Language, and is recorded as performing the task of chat, Visual question answering.
Its starting point was an existing base model, Fuyu-8B. That is why it shares the base model's general shape and size.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
Training and provenance
It was trained on a corpus of about 200,540,000 tokens of text.
Answers
OtterHD-8B — common questions
OtterHD-8B— is it open source?
The licensing was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.
OtterHD-8B— how many parameters does it have?
It has a parameter count of 8B. 8B. That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
OtterHD-8B— who created it?
It was published by Nanyang Technological University, based in Singapore, an organisation categorised as academia.
OtterHD-8B— when was it released?
It was published in November 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
OtterHD-8B— what is it used for?
It works in the domain of Multimodal, Vision, Language, and is recorded as handling the task of chat, Visual question answering. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
OtterHD-8B— what GPU do I need to run it?
None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.