This project collects various of large-scale models as follows:
- NLP
- CV
Links to resource: Github, Paper, Hugging Face Etc.
All models are not sorted by any items, may be sorted by date or parameter size, etc.
Model Name | Release Date | Developer/Institute | Size of Parameter | Github | Hugging Face | modelscope(魔搭) | Framework | Paper | Closed/Open source |
---|---|---|---|---|---|---|---|---|---|
Transformer | 2017.06 | N / A | [Link] | [Link] | [Link] (Alibaba DAMO) |
---- | [Link] | Open | |
GPT 1.0 | 2018.06 | OpenAI | 117M | [Link] | [Link] | ---- | PyTorch | [Link] | Open |
Bert | 2018.10 | 110M/340M | [Link] | [Link] | [Link] (Alibaba DAMO) |
TF | [Link] | Open | |
GPT-2 | 2019.02 | OpenAI | 124M/1158M | [Link] | [Link] | ---- | PyTorch | [Link] | Open |
XLNet | 2019.06 | CMU & Google | 110M/240M | [Link] | [Link] | ---- | TF | [Link] | Open |
T5 | 2019.10 | 60M/220M/770M | [Link] | [Link] | ---- | TF / JAX | [Link] | Open | |
mT5 | 2020.10 | 13B | [Link] | [Link] | ---- | TF | [Link] | Open | |
GPT-3 | 2020.05 | OpenAI | 175B | [Link] | ---- | ---- | PyTorch | [Link] | Closed |
Pangu-Alpha | 2020.07 | Huawei & Peng Cheng Lab | 2.6B | [Link] | [Link] | [Link] | mindspore | [Link] | Open |
CPM-2 | 2021.06 | Tsinghua University & BAAI(北京智源AI研究院) | 11B/198B | [Link] | ---- | ---- | PyTorch | [Link] | Open |
T0 | 2021.03 | Hugging Face | 11B | [Link] | [Link] | ---- | PyTorch | [Link] | Open |
PLUG | 2021.04 | Alibaba DAMO | 27B | [Link] | ---- | [Link] | PyTorch | ---- | Open |
Bloom | 2021.08 | Bloom | 176B | ---- | [Link] | [Link] (langboat Tech) |
PyTorch | [Link] | Closed |
Codex (based on GPT3) | 2021.07 | OpenAI | ---- | ---- | ---- | ---- | ---- | [Link] | Closed |
LaMDA | 2022.01 | 2B | [Link] | ---- | ---- | ---- | [Link] | Open | |
OPT | 2022.01 | FaceBook(Meta) | 125M ~ 175M | [Link] | [Link] | ---- | PyTorch | [Link] | Closed |
MT-NLG | 2022.01 | Microsoft | 530B | ---- | ---- | ---- | PyTorch | [Link] | Closed |
FLAN-T5v1.1 | 2021.09 | 245B | [Link] | [Link] | ---- | TF | [Link] | Open | |
LLaMA | 2023.02 | FaceBook(Meta) | 7B ~ 65B | [Link] | [Link] | [Link (Fengshenbang)] |
PyTorch | [Link] | Open |
WebGPT | 2021.12 | OpenAI | 175B | ---- | ---- | ---- | ---- | [Link] | Closed |
PaLM | 2022.04 | 540B | [Link] | ---- | ---- | PyTorch | [Link] | Open | |
Gopher | 2021.12 | DeepMind | 280B | ---- | ---- | ---- | ---- | [Link] | Closed |
PALM | 2020.04 | Alibaba DAMO | 257M/483M | [Link] | ---- | [Link] (Alibaba DAMO) |
PyTorch | [Link] | Open |
GPT-NeoX | 2022.04 | EleutherAI | 20B | [Link] | [Link] | ---- | PyTorch | [Link] | Open |
AlphaCode | 2021.01 | DeepMind | ---- | ---- | ---- | ---- | ---- | [Link] | Closed |
InstructGPT | 2022.01 | OpenAI | 1.3B | ---- | ---- | ---- | ---- | [Link] | Closed |
CodeGen | 2022.01 | SaleForce Research | 350M/1B/3B/7B/16B | [Link] | [Link] | ---- | PyTorch | [Link] | Open |
Model Name | Release Date | Developer/Firms | Size of Parameter | Domain | Github | Hugging Face | Supported Framework | Paper | Closed / Open source | FLOPS | Top-1 Error | Top-5 Error |
---|---|---|---|---|---|---|---|---|---|---|---|---|
ResNet | ||||||||||||
- NLP models
- CV models
- Hybrid models
- Other