Unrecorded-weight models from Google
32
Models
1 organisations
—
Open weights
0 open, 0 closed
8
Frontier
By Epoch's flag
2B
Largest traceable
Seq2Seq LSTM
2007–2018
Span
6 with a traceable size
Organisation
Country
Records
32| Model | Organisation | Released | Parameters | Weights |
|---|---|---|---|---|
| GPipe (Transformer) | 2018-11 | 6B * | Not recorded | |
| MetaMimic | 2018-10 | 22M * | Not recorded | |
| MobileNetV2 | 2018-06 | 3M * | Not recorded | |
| DeepLabV3+ | 2018-02 | — | Not recorded | |
| DeepLabV3 | 2017-06 | — | Not recorded | |
| MobileNet | 2017-04 | 4M * | Not recorded | |
| Xception | 2016-10 | 23M | Not recorded | |
| Youtube recommendation model | 2016-09 | — | Not recorded | |
| A3C FF hs | 2016-02 | — | Not recorded | |
| Inception-ResNet-V2 | 2016-02 | 56M * | Not recorded | |
| Inceptionv4 | 2016-02 | 43M * | Not recorded | |
| Inception v3 | 2015-12 | 24M * | Not recorded | |
| BatchNorm | 2015-06 | 14M | Not recorded | |
| DQN-2015 | 2015-02 | 2M * | Not recorded | |
| SNM-skip | 2014-12 | 62B * | Not recorded | |
| GoogLeNet / InceptionV1 | 2014-09 | 7M | Not recorded | |
| Seq2Seq LSTM | 2014-09 | 2B | Not recorded | |
| Multiresolution CNN | 2014-06 | 126M * | Not recorded | |
| Paragraph Vector | 2014-05 | 32M | Not recorded | |
| DeViSE | 2013-12 | — | Not recorded | |
| RNN for 1B words | 2013-12 | 20B * | Not recorded | |
| Word2Vec (large) | 2013-10 | 692M * | Not recorded | |
| Word2Vec (small) | 2013-10 | 208M * | Not recorded | |
| Multilingual DNN | 2013-05 | 207M | Not recorded | |
| ReLU-Speech | 2013-05 | 102M * | Not recorded | |
| DistBelief NNLM | 2013-01 | — | Not recorded | |
| DistBelief Speech | 2012-12 | 47M * | Not recorded | |
| DistBelief Vision | 2012-12 | 2B * | Not recorded | |
| Unsupervised High-level Feature Learner | 2012-07 | 1B * | Not recorded | |
| YouTube Video Recommendation System | 2010-09 | — | Not recorded | |
| KN-LM | 2007-06 | 21B * | Not recorded | |
| SB-LM | 2007-06 | 300B * | Not recorded |
An asterisk marks a parameter figure the source does not consider traceable to the people who built the model. That judgement is the source’s and it has false negatives: Kimi K3 is marked speculative at 2.8T while Moonshot’s own model card states the figure. Why that distinction carries the whole exhibit.
Counts are of models Epoch AI records as notable, not of every model released. The most recent month is always incomplete, because a model enters the dataset when it is reviewed rather than when it ships.
Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models-documentation' Used under CC BY 4.0.