System
stringlengths 4
29
| Training compute (FLOP)
float64 10,500,000,000,000B
50,000,000,001,000,000B
| Task
stringlengths 4
90
| Training hardware
stringclasses 10
values | Publication date
stringlengths 10
10
| Country (from Organization)
stringlengths 5
643
| Main region
stringclasses 7
values | kgco2/kwh
float64 0.08
0.67
| low_alpha
float64 -2.56
100
| high_alpha
float64 -0.17
110
| TDP
int64 250
700
| Carbon_emission(tCO2eq)
float64 10.1
31.8k
|
|---|---|---|---|---|---|---|---|---|---|---|---|
Llama 3-70B | 6,300,000,000,010,000,000,000,000 | Chat,Language modelling/generation,Code generation | NVIDIA H100 SXM5 | 2024-04-18 | United States of America | United States of America | 0.409 | 100 | 110 | 700 | 3,901.33162 |
Inflection-2.5 | 10,001,000,000,000,000,000,000,000 | Chat | NVIDIA H100 SXM5 | 2024-03-07 | United States of America | United States of America | 0.409 | 100 | 110 | 700 | 6,171.141322 |
MegaScale (Production) | 12,000,000,000,000,000,000,000,000 | Language modelling/generation | NVIDIA A100 | 2024-02-23 | China,China | China | 0.581 | 31.930312 | 44.714006 | 400 | 12,361.093788 |
Gemini Ultra | 50,000,000,001,000,000,000,000,000 | Language modelling,Visual question answering,Chat,Translation | Google TPU v4 | 2023-12-06 | Multinational | Multinational | 0.436 | 26.081536 | 36.128637 | 300 | 31,773.76719 |
Inflection-2 | 10,010,000,000,000,000,000,000,000 | Language modelling | NVIDIA H100 SXM5 | 2023-11-22 | United States of America | United States of America | 0.409 | 100 | 110 | 700 | 6,176.651955 |
Nemotron-3-8B | 180,000,000,000,000,000,000,000 | Chat,Language generation | NVIDIA A100 | 2023-11-15 | United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 139.838517 |
Skywork-13B | 250,000,000,000,000,000,000,000 | Language modelling | NVIDIA A800 | 2023-10-30 | China | China | 0.581 | 53.860444 | 53.860444 | 250 | 134.867506 |
Falcon-180B | 3,760,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 40 GB | 2023-09-06 | United Arab Emirates | United Arab Emirates | 0.673 | 24.588855 | 37.971324 | 400 | 5,163.793906 |
Llama 2-70B | 810,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2023-07-18 | United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 568.175867 |
Llama 2-7B | 84,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2023-07-18 | United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 61.049052 |
xTrimoPGLM -100B | 600,010,000,000,000,000,000,000 | Proteins | NVIDIA A100 SXM4 40 GB | 2023-07-06 | China,China | China | 0.581 | 24.588855 | 37.971324 | 400 | 736.076286 |
Pangu-Weather | 39,800,000,000,000,000,000,000 | Weather prediction | NVIDIA V100 | 2023-07-05 | China | China | 0.581 | -0.172627 | 10.469236 | 300 | 79.264265 |
PaLM 2 | 7,340,000,000,000,000,000,000,000 | Language modelling | Google TPU v4 | 2023-05-10 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 4,525.857806 |
StarCoder | 84,600,000,000,000,000,000,000 | Code generation | NVIDIA A100 SXM4 80 GB | 2023-05-09 | Multinational,United States of America,United States of America,Canada,United States of America,United States of America,Germany,Germany,United Kingdom of Great Britain and Northern Ireland,United States of America,Singapore,Israel,Australia,Australia,Australia,Canada,United States of America,Canada,United States of America,Germany,United States of America,United States of America,Sweden,Multinational,United States of America,United States of America,United States of America,India,India,Multinational,Spain,United States of America,Israel,United Kingdom of Great Britain and Northern Ireland,United States of America,Multinational,Germany | Multinational | 0.436 | 36.112884 | 50.505783 | 400 | 65.536605 |
WizardLM-7B | 40,200,000,000,000,000,000,000 | Language modelling | NVIDIA V100 | 2023-04-24 | United States of America,China | United States of America | 0.409 | -0.172627 | 10.469236 | 300 | 56.336465 |
BloombergGPT | 236,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 | 2023-03-30 | United States of America,United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 182.503868 |
PanGu-Σ | 466,999,999,999,999,900,000,000 | Code generation,Language modelling | Huawei Ascend 910 | 2023-03-20 | China | China | 0.581 | 36.636204 | 36.636204 | 310 | 399.814081 |
GPT-4 | 21,000,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 40 GB | 2023-03-15 | United States of America | United States of America | 0.409 | 24.588855 | 37.971324 | 400 | 16,992.765273 |
Falcon-40B | 240,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 | 2023-03-15 | United Arab Emirates | United Arab Emirates | 0.673 | 31.930312 | 44.714006 | 400 | 305.309019 |
LLaMA-65B | 550,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 | 2023-02-24 | United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 419.326422 |
LLaMA-7B | 40,200,000,000,000,000,000,000 | Language modelling | NVIDIA A100 | 2023-02-24 | United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 32.04692 |
ViT-22B | 400,010,000,000,000,100,000,000 | Object detection,Image classification | Google TPU v4 | 2023-02-10 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 260.200516 |
GPT-3.5 (text-davinci-003) | 2,578,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 40 GB | 2022-11-28 | United States of America | United States of America | 0.409 | 24.588855 | 37.971324 | 400 | 2,166.601168 |
Galactica | 324,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2022-11-16 | United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 230.51761 |
BLOOM-176B | 577,000,000,000,010,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2022-11-08 | Multinational,Multinational | Multinational | 0.436 | 36.112884 | 50.505783 | 400 | 433.717871 |
Taiyi-Stable Diffusion | 51,000,000,000,000,000,000,000 | Image generation,Text-to-image | NVIDIA A100 | 2022-10-31 | China | China | 0.581 | 31.930312 | 44.714006 | 400 | 57.515509 |
U-PaLM (540B) | 2,530,000,000,000,000,000,000,000 | Language generation | Google TPU v4 | 2022-10-20 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 1,590.32924 |
Flan-T5 11B | 33,000,000,000,000,000,000,000 | Language modelling/generation | Google TPU v4 | 2022-10-20 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 22.527348 |
Flan-PaLM 540B | 2,500,000,000,000,000,000,000,000 | Language modelling/generation | Google TPU v4 | 2022-10-20 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 1,571.813776 |
BlenderBot 3 | 430,000,000,000,000,000,000,000 | Chat | NVIDIA A100 SXM4 40 GB | 2022-08-10 | Canada,United States of America,Canada | Canada | 0.076 | 24.588855 | 37.971324 | 400 | 69.441097 |
AlexaTM 20B | 204,374,016,000,000,000,000,000 | Language modelling | NVIDIA A100 | 2022-08-02 | United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 158.432355 |
OmegaPLM | 13,801,881,600,000,000,000,000 | Proteins,Protein folding prediction | NVIDIA A100 SXM4 80 GB | 2022-07-22 | United States of America,China | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 10.328053 |
ESM2-15B | 73,500,000,000,099,990,000,000 | Proteins,general-purpose protein or nucleotide language model (pLM/nLM) | NVIDIA V100 | 2022-07-21 | United States of America,United States of America,United States of America,United States of America | United States of America | 0.409 | -0.172627 | 10.469236 | 300 | 100.518447 |
NLLB | 17,511,137,280,000,000,000,000 | Translation | NVIDIA A100 SXM4 80 GB | 2022-07-06 | United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 13.052741 |
Minerva (540B) | 2,741,500,000,000,000,000,000,000 | Quantitative reasoning | Google TPU v4 | 2022-06-29 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 1,720.754326 |
ProGen2-xlarge | 13,500,000,000,000,000,000,000 | Proteins,Protein generation,general-purpose protein or nucleotide language model (pLM/nLM) | Google TPU v3 | 2022-06-27 | United States of America,United States of America,United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 14.496563 |
Parti | 396,289,537,619,263,500,000,000 | Text-to-image | Google TPU v4 | 2022-06-22 | Multinational | Multinational | 0.436 | 26.081536 | 36.128637 | 300 | 274.846201 |
CoCa | 73,000,000,000,000,000,000,000 | Image classification | Google TPU v4 | 2022-06-14 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 49.061343 |
Imagen | 14,600,000,000,000,002,000,000 | Text-to-image | Google TPU v4 | 2022-05-23 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 10.130342 |
OPT-175B | 430,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2022-05-02 | United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 304.589491 |
Flamingo | 270,000,000,000,000,000,000,000 | Visual question answering,Image captioning | Google TPU v4 | 2022-04-29 | United Kingdom of Great Britain and Northern Ireland | United Kingdom of Great Britain and Northern Ireland | 0.214 | 26.081536 | 36.128637 | 300 | 92.582149 |
Stable Diffusion (LDM-KL-8-G) | 50,000,000,000,000,000,000,000 | Image generation,Text-to-image | NVIDIA A100 | 2022-04-13 | United States of America,Germany | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 39.708304 |
PaLM (540B) | 2,527,200,000,000,000,000,000,000 | Language modelling | Google TPU v4 | 2022-04-04 | Multinational | Multinational | 0.436 | 26.081536 | 36.128637 | 300 | 1,693.472287 |
LaMDA | 355,000,000,000,000,000,000,000 | Language modelling | Google TPU v3 | 2022-02-10 | United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 339.646554 |
XGLM-7.5B | 43,475,927,040,000,000,000,000 | Translation,Question answering,Language modelling/generation | NVIDIA A100 | 2021-12-20 | United States of America,United States of America | United States of America | 0.409 | 31.930312 | 44.714006 | 400 | 34.611167 |
Gopher (280B) | 631,000,000,000,000,000,000,000 | Language modelling | Google TPU v3 | 2021-12-08 | United Kingdom of Great Britain and Northern Ireland | United Kingdom of Great Britain and Northern Ireland | 0.214 | 3.330417 | 13.139611 | 250 | 309.936124 |
BASIC-L | 41,200,000,000,000,000,000,000 | Image classification | Google TPU v4 | 2021-11-19 | United States of America | United States of America | 0.409 | 26.081536 | 36.128637 | 300 | 28.001908 |
T0-XXL | 17,920,000,000,000,000,000,000 | Language modelling | Google TPU v3 | 2021-10-15 | Multinational,United States of America | Multinational | 0.436 | 3.330417 | 13.139611 | 250 | 20.297888 |
Megatron-Turing NLG 530B | 1,170,000,000,000,000,000,000,000 | Language modelling | NVIDIA A100 SXM4 80 GB | 2021-10-11 | United States of America,United States of America | United States of America | 0.409 | 36.112884 | 50.505783 | 400 | 816.085169 |
FLAN 137B | 48,960,000,000,000,000,000,000 | Language modelling | Google TPU v3 | 2021-09-03 | Multinational | Multinational | 0.436 | 3.330417 | 13.139611 | 250 | 53.466124 |
GOAT | 78,000,000,000,000,000,000,000 | Open ended play | Google TPU v3 | 2021-07-27 | United Kingdom of Great Britain and Northern Ireland | United Kingdom of Great Britain and Northern Ireland | 0.214 | 3.330417 | 13.139611 | 250 | 41.124183 |
ALIGN | 25,986,700,000,009,994,000,000 | Representation learning,Image classification | Google TPU v3 | 2021-06-11 | Multinational | Multinational | 0.436 | 3.330417 | 13.139611 | 250 | 29.035205 |
CoAtNet | 42,700,000,000,000,000,000,000 | Image classification | Google TPU v3 | 2021-06-09 | United States of America,Multinational,United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 43.957066 |
ByT5-XXL | 81,000,000,000,000,000,000,000 | Language modelling | Google TPU v3 | 2021-05-28 | United States of America,Multinational | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 81.51209 |
ProtT5-XXL-BFD | 37,000,000,000,000,000,000,000 | Proteins,general-purpose protein or nucleotide language model (pLM/nLM) | Google TPU v3 | 2021-05-04 | Germany,China,United States of America,United States of America,United States of America,Korea (Republic of) | Germany | 0.425 | 3.330417 | 13.139611 | 250 | 39.783858 |
ProtT5-XXL | 73,700,000,000,000,010,000,000 | Proteins,general-purpose protein or nucleotide language model (pLM/nLM) | Google TPU v3 | 2021-05-04 | Germany,China,United States of America,United States of America,United States of America,Korea (Republic of) | Germany | 0.425 | 3.330417 | 13.139611 | 250 | 77.323418 |
ProtBERT-BFD | 39,000,000,000,000,000,000,000 | Proteins,general-purpose protein or nucleotide language model (pLM/nLM) | Google TPU v3 | 2021-05-04 | Germany,United States of America,Korea (Republic of),United States of America,United States of America,China | Germany | 0.425 | 3.330417 | 13.139611 | 250 | 41.854859 |
Meta Pseudo Labels | 47,900,000,000,000,000,000,000 | Image classification | Google TPU v3 | 2021-03-01 | United States of America,Multinational | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 49.107653 |
Switch | 82,200,000,000,000,000,000,000 | Text autocompletion | Google TPU v3 | 2021-01-11 | United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 82.677036 |
CLIP (ViT L/14@336px) | 10,500,000,000,000,000,000,000 | Zero-shot image classification,Character recognition,Video description | NVIDIA V100 | 2021-01-05 | United States of America | United States of America | 0.409 | -0.172627 | 10.469236 | 300 | 15.571931 |
GShard (dense) | 12,800,000,000,000,000,000,000 | Translation | Google TPU v3 | 2020-06-30 | United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 13.77235 |
iGPT-XL | 33,000,000,000,000,000,000,000 | Image completion | NVIDIA Tesla V100 DGXS 32 GB | 2020-06-17 | United States of America | United States of America | 0.409 | -2.560243 | -0.172409 | 300 | 59.676968 |
GPT-3 175B (davinci) | 314,000,000,000,000,000,000,000 | Text autocompletion | NVIDIA Tesla V100 DGXS 32 GB | 2020-05-28 | United States of America | United States of America | 0.409 | -2.560243 | -0.172409 | 300 | 509.65795 |
Turing-NLG | 15,700,000,000,000,000,000,000 | Text autocompletion | NVIDIA Tesla V100 DGXS 32 GB | 2020-02-13 | United States of America | United States of America | 0.409 | -2.560243 | -0.172409 | 300 | 29.49847 |
Meena | 112,000,000,000,000,000,000,000 | Text autocompletion | Google TPU v3 | 2020-01-28 | United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 111.441642 |
AlphaStar | 59,250,000,000,001,000,000,000 | StarCraft | Google TPU v3 | 2019-10-30 | United Kingdom of Great Britain and Northern Ireland | United Kingdom of Great Britain and Northern Ireland | 0.214 | 3.330417 | 13.139611 | 250 | 31.543135 |
T5-11B | 33,000,000,000,000,000,000,000 | Text autocompletion | Google TPU v3 | 2019-10-23 | United States of America | United States of America | 0.409 | 3.330417 | 13.139611 | 250 | 34.288581 |