SPN-4+KN5
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
- Singapore University of Technology & Design,DSO National Laboratories
- Organisation type
- Academia,Government
- Country
- Singapore
- Published
- 14 September 2014
- Authors
- W. Cheng, Stanley Kok, Hoai Vu Pham, Hai Leong Chieu, K. M. A. Chai
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Language
- Task
- Language modeling
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
- 5M
- Training data
- 929,000 tokens
Estimate from table 2 of https://arxiv.org/abs/1609.07843 The authors of the linked paper draw on estimates from table 3 of https://arxiv.org/pdf/1508.06615.pdf
seems like the authors use a non-standard split for the dataset "We performed our experiments on the commonly used Penn Treebank corpus [15], and adhered to the experimental setup used in previous work [6, 9]. We used sections 0-20, sections 21-22, and sections 23-24 respectively as training, validation and test sets" apparently the most common split is "In the most common split of this corpus, sections from 0 to 18 are used for training (38 219 sentences, 912 344 tokens), sections from 19 to …
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.
- Training compute
- 4.4 × 10¹⁶ FLOP
- How it was established
- Hardware
40h, 1 GPU, 1028e9 Peak FLOP/s, 30% 1028000000000 FLOP/s/GPU * 1GPU * 40 hours * 3600 s/hour * 0.3 [assumed utilization] = 4.44096e+16 FLOP
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 Tesla C2075
- Chips used
- 1
- Wall-clock time
- 40 hours
- Power draw
- 290 W
""e stopped training our SPN when its performance on the validation set stops improving at two consecutive evaluation points, or when it has run for 40 hours, whichever occurred first. (It turned out that both SPN-3 and SPN-4 ran for the maximum of 40 hours.) We parallelized our SPN code2 to run on a GPU, and ran our experiments on a machine with a 2.4 GHz CPU and an NVIDIA Tesla C2075 GPU (448 CUDA cores, 5GB of device memory)."
Availability
Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.
- Weights
- Closed — provider access only
- Model access
- Unreleased
- Training code
- Open (non-commercial)
code, no license specified: https://github.com/stakok/lmspn/tree/master/SPNLM training code: https://github.com/stakok/lmspn/blob/master/SPNLM/README.doc
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Why it is tracked
- SOTA improvement
- Record confidence
- Likely
- Citations
- 102
- Benchmark data
- SPN-4+KN5
"Our empirical comparisons with six previous language models indicate that our SPN has superior performance"
Sources
Where this record came from and when it was last checked.
- Reference
- Language modeling with sum-product networks
- Last updated
- 28 November 2025
What the numbers mean
What this model is
SPN-4+KN5 was published by Singapore University of Technology & Design,DSO National Laboratories, in Singapore, in September 2014. The organisation is categorised as academia,Government.
It works in Language, and is recorded as doing language modeling.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
How it was trained
Training it took roughly 4.4 × 10¹⁶ FLOP of computation, on NVIDIA Tesla C2075 — a measure of what producing the model cost, not of how fast it answers.
It was trained on about 929,000 tokens of text.
It is tracked in the underlying dataset for one reason in particular: sOTA improvement.
Answers
SPN-4+KN5 — common questions
How many parameters does SPN-4+KN5 have?
SPN-4+KN5 has 5M parameters. Estimate from table 2 of https://arxiv.org/abs/1609.07843 The authors of the linked paper draw on estimates from table 3 of https://arxiv.org/pdf/1508.06615.pdf. 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.
Who created SPN-4+KN5?
SPN-4+KN5 was published by Singapore University of Technology & Design,DSO National Laboratories, based in Singapore, categorised as academia,Government.
When was SPN-4+KN5 released?
SPN-4+KN5 was published in September 2014. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is SPN-4+KN5 used for?
SPN-4+KN5 works in Language, and is recorded as handling language modeling. These are the areas it was designed around; they describe intent rather than a hard boundary.
How much compute was used to train SPN-4+KN5?
Around 4.4 × 10¹⁶ FLOP, on NVIDIA Tesla C2075. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
What GPU do I need to run SPN-4+KN5?
None. SPN-4+KN5 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.
Is SPN-4+KN5 open source?
No. SPN-4+KN5 has not had its weights published, so it exists only as a service controlled by its owner.
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.