title
stringlengths 1
544
โ | parent
stringlengths 0
57
โ | created
stringlengths 11
12
โ | editor
stringclasses 1
value | creator
stringclasses 4
values | edited
stringlengths 11
12
โ | refs
stringlengths 0
536
โ | text
stringlengths 1
26k
| id
stringlengths 32
32
|
---|---|---|---|---|---|---|---|---|
WizardCoder
|
AI Coders
|
Jun 20, 2023
|
Alan Jo
|
Alan Jo
|
Jun 20, 2023
|
[Wizard Vicuna](https://texonom.com/wizard-vicuna-c555316c38a24097ab53b72a9067dacc)
|
3da057829b834bbe8d0847fe91a32047
|
|
[copilot-in-the-terminal](https://github.com/maxwell-bland/copilot-in-the-terminal)
|
Github Usages
|
Jun 23, 2022
|
Alan Jo
|
Alan Jo
|
Jul 6, 2023
|
c0a18a70ef0d4274bfe4bb39e447d8a7
|
||
****FauxPilot****
|
Github Usages
|
Jun 23, 2022
|
Alan Jo
|
Alan Jo
|
Jul 6, 2023
|
[fauxpilot](https://github.com/moyix/fauxpilot)
|
c6376ebef88b4902b81d3c1b03e0e42b
|
|
Github Copilot Chat
|
Github Usages
|
Jul 28, 2023
|
Alan Jo
|
Alan Jo
|
Jul 28, 2023
|
> [GitHub Copilot Chat beta now available for every organization - The GitHub Blog](https://github.blog/2023-07-20-github-copilot-chat-beta-now-available-for-every-organization/)
|
5ea8cc826a214ea5b4b36daeef9c474e
|
|
Github Copilot X
|
Github Usages
|
Jun 23, 2022
|
Alan Jo
|
Alan Jo
|
Jul 6, 2023
|
1fbb14b647494e239edf33ae35cb781f
|
||
Monte Carlo Tree Search
|
Monte Carlo Method
| null | null | null | null | null |
[Monte Carlo Method](https://texonom.com/monte-carlo-method-9e3e9708b973488eb0f6f3e5c66c852e)
[Alpha Go](https://texonom.com/alpha-go-ad312555a0f1409e9d84cdcd1abe63a9) ์์ ์ฌ์ฉ
### Open Source
[mctx](https://github.com/deepmind/mctx)
|
5d1b1e2b77d9456f888e049be398c0c5
|
AI Chess
|
AI Games
|
Nov 5, 2019
|
Alan Jo
|
Seong-lae Cho
|
Jul 4, 2022
|
[Chess](https://texonom.com/chess-54e64edfb6ae4c48a88f199f9ae163d2)
|
[Alpha zero](https://texonom.com/alpha-zero-be1f223295d74c7c81e09ffed13a52f5)
## Stockfish
- One of the strongest chess engines in the world (without deep learning)
- Used complex heuristics and prunings
- Utilize all known human chess strategies
> # Micro chess
>
> [MChess by pytorch](https://texonom.com/mchess-by-pytorch-b2f5c51427bb49299ae455c347de19b7)
>
> โ Each player has one of chess piece excepting for Queen
> 
> โ The movement of knight is changed and all other pieces have same movements with the original chess (Due to neutralize Stockfish)
> โ Castling is available
> โ Pawn can move 2 step for the first time
> โ Tie rule exist
> โ Pawn promotion is possible
> โ Max turn: 30
> โ Winning condition: check-mate the opponent King
> โ If no check-mate until max turn, then calculate the score of remaining pieces
>
> 
> Tie rule: checkmate unavailable due to lack of pieces, No movement of any pawn or no capture of any piece for 50 turns, repeating same movement for 5 times
> 
|
ab986bfce29c48d9b54f19a7c848d250
|
AI Go
|
AI Games
|
Jun 13, 2021
|
Alan Jo
|
Alan Jo
|
Feb 21, 2023
|
### Adversial Attack
> [Man beats machine at Go in human victory over AI](https://arstechnica.com/information-technology/2023/02/man-beats-machine-at-go-in-human-victory-over-ai)
> [New Go-playing trick defeats world-class Go AI-but loses to human amateurs](https://arstechnica.com/information-technology/2022/11/new-go-playing-trick-defeats-world-class-go-ai-but-loses-to-human-amateurs/)
|
cbbe92b307434aabb3817fc9f5e02827
|
|
MineRL
|
AI Games
|
Jun 27, 2022
|
Alan Jo
|
Alan Jo
|
Jun 27, 2022
|
[Minecraft](https://texonom.com/minecraft-e017165733e8411dbaa7bc3ca6ab4161)
|
> [OpenAI's latest AI builds a Diamond Axe in Minecraft - why it matters](https://mixed-news.com/en/openai-ai-builds-diamond-pickaxe-in-minecraft-why-it-matters)
|
0edfbd70d62b4bca9c726e1dee047634
|
Alpha zero
|
AI Chess
| null | null | null | null | null |
> [Reimagining Chess with AlphaZero](https://cacm.acm.org/magazines/2022/2/258230-reimagining-chess-with-alphazero/fulltext)
|
be1f223295d74c7c81e09ffed13a52f5
|
MChess by pytorch
|
AI Chess
| null | null | null | null | null |
# run_game.py
board value
|
b2f5c51427bb49299ae455c347de19b7
|
AI Chip Design
|
AI Invents
|
Jun 14, 2021
|
Alan Jo
|
Alan Jo
|
Feb 24, 2022
|
[ALU](https://texonom.com/alu-9db54e5e725e42f2afce50277b912e82)
|
> [Q&A: Here's How AI Will Change Chip Design](https://spectrum.ieee.org/ai-chip-design-matlab)
> [A graph placement methodology for fast chip design](https://www.nature.com/articles/s41586-021-03544-w.epdf?sharing_token=7Cz6mcX5GzRSAgq5xQnOjdRgN0jAjWel9jnR3ZoTv0PW0K0NmVrRsFPaMa9Y5We97spjdO-aPpvZYXPHhKbfpfPljZaIm3b-kyQ3gKElVBjZIxn_5lBKsnqIIUn2YkCI3IFe5puGE49yIrhVbJrW9eUbKmMo7FS9KDgM4hs9TFHDVDitjKVuxCsQAfIhtA_MzSEnjVQ80qKUD2LxH9DNNzxihJ2e2_5AzKMg6cnzAWA%3D&tracking_referrer=www.theverge.com)
|
cd1ec22efe234e6aa57144c2641ffb74
|
AI DNA
|
AI Invents
|
Dec 16, 2020
|
Alan Jo
|
Alan Jo
|
Mar 3, 2022
|
[Alpha Fold](https://texonom.com/alpha-fold-2927d5595d1b471b864ad49f464e0513)
|
### AI [Protein](https://texonom.com/protein-783aaf7f6d23476982e04be296aa7e4a)
> [AI-designed protein awakens silenced genes, one by one](https://phys.org/news/2022-03-ai-designed-protein-awakens-silenced-genes.html)
> [๋ฅ๋ง์ธ๋ ์ํํด๋2 ์ผํฌ ์ดํ ํ๊ณ ๋นํ ์๋... "๊ณผ์ ๋ณด๋๋ค"](http://www.aitimes.com/news/articleView.html?idxno=134586)
|
0248b79f27644fbe8d772b15180bfeef
|
AI Inventor
|
AI Invents
|
Aug 18, 2021
|
Alan Jo
|
Alan Jo
|
Sep 6, 2022
|
[DABUS](https://texonom.com/dabus-1b4b50dfd5a74371bf0896b410015e81)
> [AI is my Co-Author...](https://storiesby.ai/p/ai-is-my-co-writer)
|
68eee9e8fd3e4ac7a6bcb32c3fe1c57a
|
|
AI Math
|
AI Invents
|
Aug 18, 2021
|
Alan Jo
|
Alan Jo
|
Jul 19, 2023
|
[Symbolic Regression](https://texonom.com/symbolic-regression-3cf8077c78ac49aba35372fdcd029e79)
[LeanDojo](https://texonom.com/leandojo-999e111b1e944fcab7989ce7ebb6c397)
[PRM800K](https://texonom.com/prm800k-ec20f44f9eaa445f9d04f6d30ffe8d9f)
[Minerva](https://texonom.com/minerva-7d618af4dbf84339b4847975b744237d)
|
96df1de989704237886b3bfd8ae646ff
|
|
DABUS
|
AI Inventor
| null | null | null | null | null |
first official ai inventor
ํธ์ฃผ ์ฐ๋ฐฉ๋ฒ์
> [Artificial Inventors - The Artificial Inventor Project](https://artificialinventor.com/dabus/)
|
1b4b50dfd5a74371bf0896b410015e81
|
LeanDojo
|
AI Math
| null | null | null | null | null |
[LeanDojo](https://github.com/lean-dojo/LeanDojo)
> [LeanDojo: Theorem Proving with Retrieval-Augmented Language Models](https://leandojo.org/)
|
999e111b1e944fcab7989ce7ebb6c397
|
Minerva
|
AI Math
| null | null | null | null | null |
7d618af4dbf84339b4847975b744237d
|
|
PRM800K
|
AI Math
| null | null | null | null | null |
### Open ai Letโs Verify Step by Step
**Improving mathematical reasoning with process supervision**
[RLHF](https://texonom.com/rlhf-4b184f9c9e8b4c7a8861fb6374e91aa6)
> [Improving mathematical reasoning with process supervision](https://openai.com/research/improving-mathematical-reasoning-with-process-supervision)
|
ec20f44f9eaa445f9d04f6d30ffe8d9f
|
Symbolic Regression
|
AI Math
| null | null | null | null | null |
### ์์ด์ ๋ํ ์์์ ์ฐพ์๋ด๋ ์ผ
[Sequence Theory](https://texonom.com/sequence-theory-708a6d7cc6e142f4b77c20c4f5179840)
[Navier-Stokes equation](https://texonom.com/navier-stokes-equation-e78ecd0f6c114c57b873537447779ac3)
[Regression](https://texonom.com/regression-d43885ba8cd044f4942cb591bdf1a612)
> [[PDF] An Approach to Symbolic Regression Using Feyn | Semantic Scholar](https://www.semanticscholar.org/paper/An-Approach-to-Symbolic-Regression-Using-Feyn-Brolos-Machado/447e9035c43ca947178174d10df21483137d396d)
|
3cf8077c78ac49aba35372fdcd029e79
|
**Hierarchical task network**
|
AI Planning Notion
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
[Dynamic Programming](https://texonom.com/dynamic-programming-810a9aa590434f02b9fe4d869fb9c371)
|
### HNT
์์
๊ฐ์ ์ข
์์ฑ์ ๊ณ์ธต์ ์ผ๋ก ๊ตฌ์กฐํ๋ ๋คํธ์ํฌ์ ํํ๋ก ์ ๊ณต
> [Hierarchical task network - Wikipedia](https://en.wikipedia.org/wiki/Hierarchical_task_network)
|
8919507c00324a68813c5af9c6134197
|
Extract Information invoice documents
|
Extract Information Target
|
Mar 28, 2021
|
Alan Jo
|
Alan Jo
|
Mar 28, 2021
| null |
### Extract Information invoice documents Nueral Network
|Title|
|:-:|
|[InvoiceNet](https://texonom.com/invoicenet-ab3ce1d060744cbe87f3bb1ea179ee67)|
|
a7e75461aa5c4c4696c660e52a2fb2db
|
InvoiceNet
|
Extract Information invoice documents Nueral Network
|
Mar 28, 2021
|
Alan Jo
|
Alan Jo
|
Mar 28, 2021
| null |
> [naiveHobo/InvoiceNet](https://github.com/naiveHobo/InvoiceNet)
|
ab3ce1d060744cbe87f3bb1ea179ee67
|
Gato
|
Multimodal AI Usages
|
Jul 15, 2023
|
Alan Jo
|
Alan Jo
|
Jul 15, 2023
|
### ****A Generalist Agent****
> [A Generalist Agent](https://www.deepmind.com/blog/a-generalist-agent)
|
c620dbf7164a4fcd9ec03cabcd4444cd
|
|
Gemini Google
|
Multimodal AI Usages
|
May 10, 2023
|
Alan Jo
|
Alan Jo
|
Jun 15, 2023
|
> [Gemini: Google's Latest AI Challenging GPT-4](https://www.youtube.com/watch?v=gwEuvrI4fx4)
|
2f38dbcd04f94e70a99548876aa9550f
|
|
ImageBind
|
Multimodal AI Usages
|
May 10, 2023
|
Alan Jo
|
Alan Jo
|
May 10, 2023
|
[ImageBind](https://github.com/facebookresearch/ImageBind) [IMU](https://texonom.com/imu-b40895cc42ef46d9b5d8039e7acfd067) [Depth Data](https://texonom.com/depth-data-3f1b6800acae4165ae80373d627d851c)
|
### ImageBind Usages
|Title|
|:-:|

> [ImageBind: Holistic AI learning across six modalities](https://ai.facebook.com/blog/imagebind-six-modalities-binding-ai/)
> [ImageBind by Meta AI](https://imagebind.metademolab.com/)
> [Meta's open-source ImageBind AI aims to mimic human perception | Engadget](https://www.engadget.com/metas-open-source-imagebind-ai-aims-to-mimic-human-perception-181500560.html)
|
e15b50fde2f149d3b5531ccf3ac7b718
|
Vide LLaMa
|
Multimodal AI Usages
|
Jun 15, 2023
|
Alan Jo
|
Alan Jo
|
Jun 15, 2023
|
[Video-LLaMA](https://github.com/DAMO-NLP-SG/Video-LLaMA)
> [Video LLaMA - a Hugging Face Space by DAMO-NLP-SG](https://huggingface.co/spaces/DAMO-NLP-SG/Video-LLaMA)
|
c60f632be89c4e29a7fce4ecdbc5c59e
|
|
AMR
|
NLP Notion
|
Jun 6, 2023
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
[AST](https://texonom.com/ast-6da505f4909a41809a74637101a4733c)
|
### Abstract Meaning Representation
์ ์ฒด ๋ฌธ์ฅ์ผ๋ก ๊ตฌ์ฑ๋ ๋ฃจํธ, ๋ ์ด๋ธ, ๋ฐฉํฅ์ฑ, ๋น์ํ ๊ทธ๋ํ
์๋ฏธ๊ฐ ์ ์ฌํ ๋ฌธ์ฅ์ ๋จ์ด๊ฐ ๋์ผํ์ง ์๋๋ผ๋ ๋์ผํ AMR์ ํ ๋นํด์ผ ํ๋ค
### AMR Tools
|Title|
|:-:|
|[AMR Parser](https://texonom.com/amr-parser-c773435d77864caba5ed303ecfb91161)|
|
b5578921e5b045228e7829e8190080c5
|
Factual Inconsistency
|
NLP Notion
|
Mar 4, 2023
|
Alan Jo
|
Alan Jo
|
Mar 6, 2023
|
๊ฐ์ ๊ฑฐ๋ฅผ ๋ฌผ์ด๋ด๋ ์ฌ์ค์ด ๋งค๋ฒ ๋ฐ๋
[Extractive Summarization](https://texonom.com/extractive-summarization-3a4cf7407c114d89a9b330d7c16eda9e) ๋ฉด ๊ฑฐ์ ์์๊ธฐ๋๋ฐ [Abstractive Summarization](https://texonom.com/abstractive-summarization-499de8f986314d6bba102d08bb90be0e) ๋ฉด ๋ง์ด ์๊ธธ์๋
|
2506a0346c544cd8b9bd5e5e4acfdee8
|
|
Information Extraction
|
NLP Notion
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
Jun 20, 2023
|
### Information Extraction Tools
|Title|
|:-:|
|[AllenAI OpenIE](https://texonom.com/allenai-openie-0aa33b9a7646435f9ee4fb0c0c1f73ec)|
|[Stanford OpenIE](https://texonom.com/stanford-openie-91d65013748045838a8ca85055780228)|
### Information Extraction Notion
|Title|
|:-:|
|[NER](https://texonom.com/ner-67dab0e286d642659d6c1f08e04ca8b8)|
|[Latent Dirichlet Allocation](https://texonom.com/latent-dirichlet-allocation-4b1b2c9bf02c4b25a57b08c9d682a133)|
|
5f202d57685c4a429d617ef37179fec3
|
|
Language Model
|
NLP Notion
|
May 26, 2021
|
Alan Jo
|
Alan Jo
|
May 29, 2023
|
[prize](https://github.com/inverse-scaling/prize) [Transformer Model](https://texonom.com/transformer-model-f3e8053cc5b447a2bc9c6b5d0874dafc)
|
### Language Usages
|Title|
|:-:|
|[LLM](https://texonom.com/llm-b9157a9d105d450ba86f2570619551b4)|
|[AI Hallucination](https://texonom.com/ai-hallucination-bd24d3510f0748ec9a6347311e1357bc)|
|[Perplexity](https://texonom.com/perplexity-2db74a7c527e4796931aa9e5a4fedaeb)|
|[Inter-AI Communication](https://texonom.com/inter-ai-communication-ce6f0e6ea4654be3b3bddf80189231d8)|
|[AI Detector](https://texonom.com/ai-detector-fe2dba46160f4d9b9df0958de9dc1ffa)|
|[Statistical Language Model](https://texonom.com/statistical-language-model-920a63d849bd4073a0ab5d841ccb2b32)|
|[NNLM](https://texonom.com/nnlm-862dfdb4f5964e738c86bc10ef788ec7)|
|[Text Generation](https://texonom.com/text-generation-cb0e475216c043fdbb4f309d324c9d19)|
### Models History
> [Five years of GPT progress](https://finbarr.ca/five-years-of-gpt-progress)
> [AI Revolution - Transformers and Large Language Models (LLMs)](http://blog.eladgil.com/2022/08/ai-revolution-transformers-and-large.html)
> [03. ์ธ์ด ๋ชจ๋ธ(Language Model)](https://wikidocs.net/21695)
|
577d47779c5f4da98e3fb0df753b9e81
|
NLP History
|
NLP Notion
|
Oct 12, 2021
|
Alan Jo
|
Alan Jo
|
Mar 6, 2023
|

> [Google 'LaMDA' is the next breakthough in AI understanding - 9to5Google](https://9to5google.com/2021/05/18/google-lamda-natural-language/?utm_source=tldrnewsletter)
> [Microsoft and Nvidia team up to train one of the world's largest language models](https://venturebeat.com/2021/10/11/microsoft-and-nvidia-team-up-to-train-one-of-the-worlds-largest-language-models/?utm_source=tldrnewsletter)
|
d2869d4b281e4bf4b6c359784885307f
|
|
STS
|
NLP Notion
|
Mar 10, 2023
|
Alan Jo
|
Alan Jo
|
Apr 30, 2023
|
[Text embedding](https://texonom.com/text-embedding-f592251a59af4e5b8a7a70c255546ef7) [Vector Similarity](https://texonom.com/vector-similarity-846675a408e74235acc836302110ab07)
|
## Semantic Textual Similarity
### Semantic Similarities
|Title|
|:-:|
### Interpretable Semantic Textual Similarity
> [Toward Interpretable Semantic Textual Similarity via Optimal Transport-based Contrastive Sentence Learning](https://aclanthology.org/2022.acl-long.412/)
|
f04e643343d74118a2c6822a54f0caf4
|
Text Analysis
|
NLP Notion
|
May 8, 2023
|
Alan Jo
|
Alan Jo
|
Jun 20, 2023
|
[Text Preprocessing](https://texonom.com/text-preprocessing-6d38abe33c1b4039813bf186daa38a2c)
|
1. Morphological Analysis
2. Syntax Analysis
3. Semantic Analysis
4. Pragmatic Analysis
> [[NLP] ์์ฐ์ด ๋ถ์์ 4๋จ๊ณ ๋ฐ ๊ธฐ๊ณํ์ต์์ ์ ์ฉ ๋ฐฉ์](https://marketingscribbler.tistory.com/4#4._ํ์ฉ_๋ถ์_(Pragmatic_Analysis))
|
bf74204abecf46119d5f2db64d55d5de
|
Text **embedding**
|
NLP Notion
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Jul 30, 2023
|
[Text Tokenizer](https://texonom.com/text-tokenizer-2bbc41eaa76c4674a4f4b9127fbe5da1) [Latent Space](https://texonom.com/latent-space-d67b6bdef18b4058bfbc3d25f87ec087) [Text Encoding](https://texonom.com/text-encoding-ab7377cfc5c648059de4860510ad9134) [Recommend System](https://texonom.com/recommend-system-e94b364527554e65bfd9d3dc63fb1af7)
|
### Word string โ Tokenizer โ Token โ Embedding layer โ Dense vector
ํ
์คํธ ๋ถ์์ ์ํด ๋จ์ด๋ฅผ ํํํ๋ ๋ฐ ์ฌ์ฉ๋๋ ์ฉ์ด
๋ฒกํฐ ๊ณต๊ฐ์์ ๋ ๊ฐ๊น์ด ๋จ์ด๊ฐ ์์๋๋๋ก ๋จ์ด์ ์๋ฏธ๋ฅผ ์ธ์ฝ๋ฉํ๋ ์ค์ ๋ฒกํฐ์ ํํ
์-ํซ ๋ฒกํฐ๋ ๋จ์ด ๋ฒกํฐ ๊ฐ ์ ์๋ฏธํ ์ ์ฌ๋๋ฅผ ๊ณ์ฐํ ์ ์๋ค๋ ๋จ์
### Text **embedding Notion**
|Title|
|:-:|
|[Positional Encoding](https://texonom.com/positional-encoding-19f7bc463c34450987b69ee4b610a08b)|
|[Token Embeddings](https://texonom.com/token-embeddings-deca26ce26df47f190b4fdabfa8ac973)|
|[Measures of Contextuality](https://texonom.com/measures-of-contextuality-b3a0443e9489424f82ce4bc19dd1014c)|
### Static **embedding Methods**
|Title|
|:-:|
|[Glove Embedding](https://texonom.com/glove-embedding-7356854e1b0e4aff8142c2fadd24070c)|
|[Bag of words](https://texonom.com/bag-of-words-83c7dd03d30d4ed99c8130ba319fe484)|
|[TF-IDF](https://texonom.com/tf-idf-e564d1cb86074a7fa2f46c38cab552b2)|
|[Word2Vec](https://texonom.com/word2vec-b3b6fa1a01324dd2b10109bdeb192711)|
### Contextual **embedding Methods**
|Title|
|:-:|
|[Sentence Transformer](https://texonom.com/sentence-transformer-5b7338c829c748d49960e592ef4ba8db)|
|[ELMo](https://texonom.com/elmo-6e0418bc10274d29b2cb7a1aae8993a4)|
> [Contextual Embedding - How Contextual are Contextualized Word Representations?](https://velog.io/@stapers/Contextual-Embedding-How-Contextual-are-Contextualized-Word-Representations)
> [Embedding vs Encoding](https://beausty23.tistory.com/223)
> [[๋ฅ๋ฌ๋] ์ธ์ด๋ชจ๋ธ, RNN, GRU, LSTM, Attention, Transformer, GPT, BERT ๊ฐ๋
์ ๋ฆฌ](https://velog.io/@rsj9987/๋ฅ๋ฌ๋-์ฉ์ด์ ๋ฆฌ)
|
f592251a59af4e5b8a7a70c255546ef7
|
Text Encoding
|
NLP Notion
|
Mar 26, 2023
|
Alan Jo
|
Alan Jo
|
Jul 6, 2023
|
[Text embedding](https://texonom.com/text-embedding-f592251a59af4e5b8a7a70c255546ef7)
|
### usually for sentence
๋จ์ด์ ์๋ฏธ๋ฅผ ๋ณด์กดํ์ง ์๋ ๊ฒ์ ์๋ฏธํ๊ฑฐ๋ Embedding ๋ ๋จ์ด ๋ฒกํฐ๋ค์ ์กฐํฉํ๋ ๊ฒ์ ์๋ฏธ
[One Hot encoding](https://texonom.com/one-hot-encoding-10bd34f1bd34477ab4d6c946c750cf43)
> [Embedding vs Encoding](https://beausty23.tistory.com/223)
|
ab7377cfc5c648059de4860510ad9134
|
Text Tokenizer
|
NLP Notion
|
Mar 26, 2023
|
Alan Jo
|
Alan Jo
|
Jul 30, 2023
|
[Text embedding](https://texonom.com/text-embedding-f592251a59af4e5b8a7a70c255546ef7)
|
### Text Tokenizer Notion
|Title|
|:-:|
|[Subword Tokenizer](https://texonom.com/subword-tokenizer-9366550159e74e338b914fb2647186fa)|
|[Special Token](https://texonom.com/special-token-cc4caa9ee9d84a1aadf59a545b5d4e1c)|
### Text Tokenizers
|Title|
|:-:|
|[](https://texonom.com/a2a9fb1c6c69477f8c12b6a086e24ded)|
|[tiktoken](https://texonom.com/tiktoken-df0cc79b510742018c269875a4918468)|
|[SubwordTextEncoder](https://texonom.com/subwordtextencoder-2eea9bbc74a94d6cb627b45957b64252)|
|[Huggingface Tokenizer](https://texonom.com/huggingface-tokenizer-15d35355cc9946b1ac0a0e74c1fa1474)|
> [Understanding GPT tokenizers](https://simonwillison.net/2023/Jun/8/gpt-tokenizers/)
|
2bbc41eaa76c4674a4f4b9127fbe5da1
|
AMR Parser
|
AMR Tools
|
Jun 6, 2023
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
### AMR Parsers
|Title|
|:-:|
|[Transition AMR Parser](https://texonom.com/transition-amr-parser-31152bb2e7bd4428896ada216d04a1cb)|
|
c773435d77864caba5ed303ecfb91161
|
|
Transition AMR Parser
|
AMR Parsers
|
Jun 6, 2023
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
[transition-amr-parser](https://github.com/IBM/transition-amr-parser)
|
31152bb2e7bd4428896ada216d04a1cb
|
|
Latent Dirichlet Allocation
|
Information Extraction Notion
|
Jul 13, 2021
|
Alan Jo
|
Alan Jo
|
Jun 20, 2023
|
### ์ ์ฌ ๋๋ฆฌํด๋ ํ ๋น
์ฃผ์ด์ง ๋ฌธ์์ ๋ํ์ฌ ๊ฐ ๋ฌธ์์ ์ด๋ค ์ฃผ์ ๋ค์ด ์กด์ฌํ๋์ง๋ฅผ ์์ ํ๋ ๋ํ ํ๋ฅ ์ ํ ํฝ ๋ชจ๋ธ ๊ธฐ๋ฒ ์ค ํ๋
|
4b1b2c9bf02c4b25a57b08c9d682a133
|
|
NER
|
Information Extraction Notion
|
May 8, 2023
|
Alan Jo
|
Alan Jo
|
Jul 4, 2023
|
### Named Entity Recognition
๋น์ ํ ํ
์คํธ์ ๊ฐ์ฒด๋ช
์ธ๊ธ์ ๋ฏธ๋ฆฌ ์ ์๋ ๋ถ๋ฅ๋ก ์์น์ํค๊ณ ๋ถ๋ฅ
๊ฐ์ฒด๋ช
์ธ์
[SpanMarkerNER](https://github.com/tomaarsen/SpanMarkerNER)
> [12-03 ๊ฐ์ฒด๋ช
์ธ์(Named Entity Recognition)](https://wikidocs.net/30682)
|
67dab0e286d642659d6c1f08e04ca8b8
|
|
AllenAI OpenIE
|
Information Extraction Tools
|
Apr 8, 2023
|
Alan Jo
|
Alan Jo
|
May 8, 2023
|
### AllenAI OpenIE Versions
|Title|
|:-:|
|[AllenAI OpenIE 5](https://texonom.com/allenai-openie-5-0b296be12ed64e9f9f94e2567bd798e8)|
|[AllenAI OpenIE 4](https://texonom.com/allenai-openie-4-2fdc95a876714580a571d2fd168e6f7d)|
> [Docker](https://hub.docker.com/search?q=openie)
|
0aa33b9a7646435f9ee4fb0c0c1f73ec
|
|
Stanford OpenIE
|
Information Extraction Tools
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
May 17, 2023
|
### Do not use this
### OpenIE Notion
|Title|
|:-:|
|[RnnOIE](https://texonom.com/rnnoie-5720bbb60ea347eeb0eab6796c2bf853)|
|[SpanOIE](https://texonom.com/spanoie-6598041ca45f4263beb8a7d442877599)|
|[BIO Tagging](https://texonom.com/bio-tagging-cb44c4d6f940486e9a26392790889682)|
|[Seq2seq OIE](https://texonom.com/seq2seq-oie-7b156f968be9465aa5727d367a8c928c)|
### Triple Extraction
> [Stanford-OpenIE๋ฅผ ํ์ฉํ text -> triple (2)](https://velog.io/@whanhee97/Stanford-OpenIE๋ฅผ-ํ์ฉํ-text-triple-2)
### Structure
> [Untitled](http://dsba.korea.ac.kr/seminar/?mod=document&uid=255)
### Python wrapper
> [GitHub - philipperemy/stanford-openie-python: Stanford Open Information Extraction made simple!](https://github.com/philipperemy/stanford-openie-python)
|
91d65013748045838a8ca85055780228
|
|
AllenAI OpenIE 4
|
AllenAI OpenIE Versions
|
Apr 8, 2023
|
Alan Jo
|
Alan Jo
|
Apr 8, 2023
|
[openie-standalone](https://github.com/allenai/openie-standalone)
|
> [Docker](https://hub.docker.com/layers/schmmd/openie-standalone/latest/images/sha256-7aa57a7b4a45adb2c5a0b423771a96ff15db54ca23f0131fb54a44f3bd15235a?context=explore)
|
2fdc95a876714580a571d2fd168e6f7d
|
AllenAI OpenIE 5
|
AllenAI OpenIE Versions
|
Apr 8, 2023
|
Alan Jo
|
Alan Jo
|
May 18, 2023
|
[OpenIE-standalone](https://github.com/dair-iitd/OpenIE-standalone) [python-wrapper-OpenIE5](https://github.com/vaibhavad/python-wrapper-OpenIE5)
|
```typecd /openie/openie-standalone && \ java -Xmx10g -XX:+UseConcMarkSweepGC -jar \ openie-assembly-5.0-SNAPSHOT.jar --httpPort 8000```
```typeUsage: openie [options] input-file ouput-file --parser-server <value> Parser server --srl-server <value> SRL server --encoding <value> Character encoding --format <value> Output format simple | column --httpPort <value> Http Port on which to run OpenIE as a server -u | --usage show cli usage** --ignore-errors** ignore errors --include-unknown-arg2 includes arg2 [UNKNOWN] extractions from relnoun -b | --binary binary output -s | --split Split paragraphs into sentences input-file input file ouput-file output file```
### Working
> [Docker](https://hub.docker.com/r/chengchingwen/openie)
### Official not working
> [Docker](https://hub.docker.com/r/allenai/openie-demo)
|
0b296be12ed64e9f9f94e2567bd798e8
|
BIO Tagging
|
OpenIE Notion
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
Mar 5, 2023
|
### Beginning, Inside, Outside
|
cb44c4d6f940486e9a26392790889682
|
|
RnnOIE
|
OpenIE Notion
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
Mar 5, 2023
|
[BiLSTM](https://texonom.com/bilstm-cc8242f116e643abb71167e859561991)
|
5720bbb60ea347eeb0eab6796c2bf853
|
|
Seq2seq OIE
|
OpenIE Notion
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
Mar 5, 2023
|
[Seq2Seq](https://texonom.com/seq2seq-01a9854dffa6417c87d92c11a607250c)
|
7b156f968be9465aa5727d367a8c928c
|
|
SpanOIE
|
OpenIE Notion
|
Mar 5, 2023
|
Alan Jo
|
Alan Jo
|
Mar 5, 2023
|
6598041ca45f4263beb8a7d442877599
|
||
AI Detector
|
Language Usages
|
Apr 15, 2023
|
Alan Jo
|
Alan Jo
|
Apr 15, 2023
|
### AI Detectors
|Title|
|:-:|
|[Zero GPT](https://texonom.com/zero-gpt-80ff609fdaac431f9f0fecbc97d33ee6)|
|
fe2dba46160f4d9b9df0958de9dc1ffa
|
|
AI Hallucination
|
Language Usages
|
Mar 20, 2023
|
Alan Jo
|
Alan Jo
|
Apr 30, 2023
|
### ๊ฑฐ์ง๋ง ๋ฌธ๋งฅ๊ณผ ๋ค๋ฅธ ๋ฐ์ดํฐ๋ก ํ์
ํด์ ํ์ตํ๋ ๋ชจ๋ธ ํ์
[Contrastive Learning](https://texonom.com/contrastive-learning-73afe4f0d2554be392d775a0a05fe328)
|
bd24d3510f0748ec9a6347311e1357bc
|
|
Inter-AI Communication
|
Language Usages
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Apr 12, 2023
|
### Not have to be a Natural Language, Vector or Programming Language
### Inter-AI Communication Usages
|Title|
|:-:|
|[Inter-AI Protocol](https://texonom.com/inter-ai-protocol-2c48f2304a974ec5bb76a4addf4c79e5)|
|[Multi-AI Controller](https://texonom.com/multi-ai-controller-fcd2de8fdf0a4853a5bce77b2a8dc4e7)|
> [Surprising things happen when you put 25 AI agents together in an RPG town](https://arstechnica.com/information-technology/2023/04/surprising-things-happen-when-you-put-25-ai-agents-together-in-an-rpg-town/)
|
ce6f0e6ea4654be3b3bddf80189231d8
|
|
LLM
|
Language Usages
|
Feb 21, 2023
|
Alan Jo
|
Alan Jo
|
Aug 7, 2023
|
[Transformer Model](https://texonom.com/transformer-model-f3e8053cc5b447a2bc9c6b5d0874dafc) [open-llms](https://github.com/eugeneyan/open-llms) [Search Engine](https://texonom.com/search-engine-f29353bb427241a88e03dd02f3f52af0)
|
## Engine for AI Product using [Prompt Engineering](https://texonom.com/prompt-engineering-eb0deb4baf844bebb873c19a0e307e7e)
### LLM Usages
|Title|
|:-:|
|[Chat AI](https://texonom.com/chat-ai-54e11ff265864d949816a2f86a1e1b40)|
|[Learn LLM](https://texonom.com/learn-llm-fe951103d9394d0b8102a5b7719f4a21)|
|[AI Question answering](https://texonom.com/ai-question-answering-05d026befa7441949b7bff695e24268c)|
|[LLM Development](https://texonom.com/llm-development-04d6c7eb5be24cc28be8bc2175f1a3f0)|
|[MLLM](https://texonom.com/mllm-98172e9446c04bcc9cf52b2bc5d0bd17)|
|[LLM Chunking](https://texonom.com/llm-chunking-ae00551922854c1198744f19cc9e6ac4)|
|[LLMA](https://texonom.com/llma-0f7a9f91fb014d2180e50bcd66abb36b)|
|[LLM Optimization](https://texonom.com/llm-optimization-fd44b16e13b5410b84f17554198b1b24)|
|[Document AI](https://texonom.com/document-ai-073d20eac42148de9adcf9e25aafea22)|
|[LLM Performance](https://texonom.com/llm-performance-64fb9ce675344d0390d6155ed6efb0fe)|
|[In-context learning](https://texonom.com/in-context-learning-2f55f51a87a04785be78dc65ef752ca4)|
|[LLM API](https://texonom.com/llm-api-21350df762de48469ee6285c9dc171eb)|
|[LLM Context](https://texonom.com/llm-context-252aa3db0c164002aba92482142bcc2d)|
|[Uncensored LLM](https://texonom.com/uncensored-llm-61d50af6dc494ed1960c8c632394c8a5)|
|[Prompt Engineering](https://texonom.com/prompt-engineering-eb0deb4baf844bebb873c19a0e307e7e)|
|[Multilingual LLM](https://texonom.com/multilingual-llm-7467ae0b84964657a5e0bc80bd845313)|
|[LLM History](https://texonom.com/llm-history-0f189de18e6945e4812199276e5ce118)|
### Language Models
|Title|
|:-:|
|[BERT](https://texonom.com/bert-e282abe8a34543988a0e71f6c8701ad2)|
|[GPT](https://texonom.com/gpt-d07bb69aa3794c44b3cdc1347132f75b)|
|[LLaMA](https://texonom.com/llama-f2b6721202d44d469add84d8a366809c)|
|[YaLM 100B](https://texonom.com/yalm-100b-e3af4398a1af4b3a8453d7eb98a59739)|
|[PaLM](https://texonom.com/palm-a44fb7aaf8234a828d7b0f87d1fa243f)|
|[BLOOM AI](https://texonom.com/bloom-ai-178dc03bae7847a5b17f45bad30a307e)|
|[KOSMOS-1](https://texonom.com/kosmos-1-eef4f69c8dfb4fab8cc3ee3019ad2e6f)|
|[Dolly](https://texonom.com/dolly-c5164b3964a6425283042fa6878e9526)|
|[Stable LM](https://texonom.com/stable-lm-2274fed51886447abc7a9e8c1550f9b8)|
|[MPT](https://texonom.com/mpt-bf2abacce4b14fe99929c1fb19ea2c71)|
|[Falcon](https://texonom.com/falcon-38be0309cd56451ebb0f5da43bfe3508)|
|[Qwen](https://texonom.com/qwen-98e9c8006e764682ab87af31c84d6bbf)|

> [Catching up on the weird world of LLMs](https://simonwillison.net/2023/Aug/3/weird-world-of-llms)
### Emergent
> [Emergent Abilities of Large Language Models](https://www.assemblyai.com/blog/emergent-abilities-of-large-language-models)
|
b9157a9d105d450ba86f2570619551b4
|
NNLM
|
Language Usages
|
Jul 6, 2023
|
Alan Jo
|
Alan Jo
|
Jul 6, 2023
|
### NNLM Notion
|Title|
|:-:|
|[NNLM Projection Layer](https://texonom.com/nnlm-projection-layer-a1ffa22b14d941eeb7cd8576da9a84a3)|
|
862dfdb4f5964e738c86bc10ef788ec7
|
|
Perplexity
|
Language Usages
|
Mar 28, 2023
|
Alan Jo
|
Alan Jo
|
Jul 15, 2023
|
## PPL
$$PPL(W) = P(w_1, ..., w_N)^{\frac{1}{N}}$$
Time step๋ง๋ค perplexity๋งํผ์ ๋จ์ด๋ฅผ ๊ฐ์ง๊ณ ์ด๋ค ๊ฒ์ด ์ ๋ต์ธ์ง ๊ณ ๋ฏผ
PPL์ ๊ฐ์ด ๋ฎ๋ค๋ ๊ฒ์ **๋ฐ์ดํฐ ์์์ ๋์ ์ ํ๋๋ฅผ ๋ณด์ธ๋ค๋ ๊ฒ**์ด์ง, ์ฌ๋์ด ์ง์ ๋๋ผ๊ธฐ์ ์ข์ ์ธ์ด ๋ชจ๋ธ์ด๋ผ๋ ๊ฒ์ ๋ฐ๋์ ์๋ฏธํ์ง ์๋๋ค
### Perplexity Notion
|Title|
|:-:|
> [03-05 ํํ๋ ์ํฐ(Perplexity, PPL)](https://wikidocs.net/21697)
|
2db74a7c527e4796931aa9e5a4fedaeb
|
|
Statistical Language Model
|
Language Usages
|
May 9, 2023
|
Alan Jo
|
Alan Jo
|
May 9, 2023
|
### Statistical Language Models
|Title|
|:-:|
|[N gram Language Model](https://texonom.com/n-gram-language-model-e600ddf01e2645e1b1d76604454bc3c9)|
|
920a63d849bd4073a0ab5d841ccb2b32
|
|
Text Generation
|
Language Usages
|
Jul 17, 2023
|
Alan Jo
|
Alan Jo
|
Jul 17, 2023
|
์ฐ๋ฆฌ๋ ๋ํ๋ฅผ ํ๋ฉด์ ๋ฐํ ๋งฅ๋ฝ์ ๊ธฐ์ค์ผ๋ก ๋ค์ ๋จ์ด๋ฅผ ๋ด๋ฑ๋๋ค
๊ทธ ๊ณผ์ ์ ์์ํ ๋๋ ์๊ณ ์์ํ์ง ์์ ๋๋ ์๋ ์์ ๊ณผ ๋ฌด์์ ์์ด ์ด๋๊ฐ
[Ilya Sutskever](https://texonom.com/ilya-sutskever-f38707abee824d21b2ec9824a1706342) ๋ SEO์ ๋ฌธ๊ฐ์์ Search input์์ ๋ค์ ๋จ์ด ์์ธก์ ์ค์์ฑ์ ๊นจ๋ซ๋๋ค
์ด ๊ณผ์ ์ ์ธ๊ฐ ์์์ [AI Function Imitation](https://texonom.com/ai-function-imitation-0461a205664f4339958496692f8cc8bb) ์ด๋ค [Back Propagation](https://texonom.com/back-propagation-18f4493692ad43449d4271f1bb293781) ๋ [Neuroplasticity](https://texonom.com/neuroplasticity-63de0a67105e4c6ab8f6958b281c4e19) ์ ๊ธฐ๋ฅ๋ชจ๋ฐฉ์ด๋ค
### Text Generation Notion
|Title|
|:-:|
|[Text Generation Temperature](https://texonom.com/text-generation-temperature-756bf9858409478788c37b204d7f2bcb)|
|
cb0e475216c043fdbb4f309d324c9d19
|
|
Zero GPT
|
AI Detectors
|
Apr 15, 2023
|
Alan Jo
|
Alan Jo
|
Apr 15, 2023
|
80ff609fdaac431f9f0fecbc97d33ee6
|
||
Inter-AI Protocol
|
Inter-AI Communication Usages
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Apr 12, 2023
|
### Inter-AI Protocols
|Title|
|:-:|
|[Template Page](https://texonom.com/template-page-c191ecc68b5e49f1946e5704866dd4cd)|
|
2c48f2304a974ec5bb76a4addf4c79e5
|
|
Multi-AI Controller
|
Inter-AI Communication Usages
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Apr 12, 2023
|
### Multi-AI Controllers
|Title|
|:-:|
|[HuggingGPT](https://texonom.com/hugginggpt-4116dc9574604a468ccaff76aa768b4b)|
|[Mixture-of-Experts](https://texonom.com/mixture-of-experts-eb0cadda3237429dbc3b61d14fab81d1)|
|
fcd2de8fdf0a4853a5bce77b2a8dc4e7
|
|
Template Page
|
Inter-AI Protocols
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Apr 12, 2023
|
c191ecc68b5e49f1946e5704866dd4cd
|
||
HuggingGPT
|
Multi-AI Controllers
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Jun 1, 2023
|
[JARVIS](https://github.com/microsoft/JARVIS) [ChatGPT](https://texonom.com/chatgpt-dbe93533b058404c82bbf0173f2a004d) [HuggingFace](https://texonom.com/huggingface-eb25c513b432477e9da51ca19bb06833) [ControlNet](https://texonom.com/controlnet-75a904f4a34c4cc293cbab417da42569)
|
## Microsoft Jarvis
1. Task Planning
2. Model Selection
3. Task Execution
4. Response Generation
### HuggingGPT Usages
|Title|
|:-:|
|[HuggingGPT Inference Mode](https://texonom.com/hugginggpt-inference-mode-cdf17a48677a468b94f0933b744dc3ad)|
> [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in HuggingFace](https://arxiv.org/abs/2303.17580v2)
> [HuggingGPT - ์ฌ๋ฌ ์ธ๊ณต ์ง๋ฅ ๋ชจ๋ธ์ ์ฐ๊ฒฐํ๋ ๋ชจ๋ธ](https://tilnote.io/pages/642ce7f346b81350e02b47b4)
> [HuggingGPT - a Hugging Face Space by microsoft](https://huggingface.co/spaces/microsoft/HuggingGPT)
|
4116dc9574604a468ccaff76aa768b4b
|
**Mixture-of-Experts**
|
Multi-AI Controllers
|
Apr 12, 2023
|
Alan Jo
|
Alan Jo
|
Jun 29, 2023
|
### MoE
> [Mixture-of-Experts with Expert Choice Routing](https://ai.googleblog.com/2022/11/mixture-of-experts-with-expert-choice.html)
> [Outrageously Large Neural Networks: The Sparsely-Gated...](https://arxiv.org/abs/1701.06538)
|
eb0cadda3237429dbc3b61d14fab81d1
|
|
HuggingGPT Inference Mode
|
HuggingGPT Usages
|
May 31, 2023
|
Alan Jo
|
Alan Jo
|
Jun 1, 2023
|
cdf17a48677a468b94f0933b744dc3ad
|
||
BERT
|
Language Models
|
Aug 17, 2020
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
[bert](https://github.com/google-research/bert) [AutoEncoder](https://texonom.com/autoencoder-7de48b307d1347ac8da05f17f314caff) [BART](https://texonom.com/bart-48bd2258eee74af3b2d929b03bb9553b)
|
## Masked Language Model
### **Bidirectional Encoder Representations from Transformers**
2018
> [BERT: Pre-training of Deep Bidirectional Transformers for Language...](https://arxiv.org/abs/1810.04805)
> [๊ธฐ๋ณธ ์ ๊ณต BERT ์๊ณ ๋ฆฌ์ฆ ์์ํ๊ธฐ ย |ย AI Platform Training ย |ย Google Cloud](https://cloud.google.com/ai-platform/training/docs/algorithms/bert-start?hl=ko)
|
e282abe8a34543988a0e71f6c8701ad2
|
BLOOM AI
|
Language Models
|
Mar 21, 2023
|
Alan Jo
|
Alan Jo
|
May 31, 2023
|
[bloomz.cpp](https://github.com/NouamaneTazi/bloomz.cpp)
|
### BLOOM AI Usages
|Title|
|:-:|
> [BLOOM](https://bigscience.huggingface.co/blog/bloom)
> [BLOOM](https://huggingface.co/docs/transformers/model_doc/bloom)
b
|
178dc03bae7847a5b17f45bad30a307e
|
Dolly
|
Language Models
|
Apr 15, 2023
|
Alan Jo
|
Alan Jo
|
Apr 15, 2023
|
[dolly](https://github.com/databrickslabs/dolly)
|
> [Databricks releases Dolly 2.0, the first open, instruction-following LLM for commercial use](https://venturebeat.com/ai/databricks-releases-dolly-2-0-the-first-open-instruction-following-llm-for-commercial-use/)
|
c5164b3964a6425283042fa6878e9526
|
Falcon
|
Language Models
|
Jun 2, 2023
|
Alan Jo
|
Alan Jo
|
Aug 5, 2023
|
## `model_type: `****`RefinedWeb`****
[HuggingFace](https://texonom.com/huggingface-eb25c513b432477e9da51ca19bb06833) leaderboard 1st (2023-5)
์๋์๋ฏธ๋ฆฌํธ ์๋ถ๋ค๋น์ TII(Technology Innovation Institute)์์ ๊ฐ๋ฐ
[Falcon Openassistant](https://texonom.com/falcon-openassistant-77666e440c964fdab7bcca371dab911d)
### Implementations
> [TheBloke/WizardLM-Uncensored-Falcon-40B-3bit-GPTQ ยท Hugging Face](https://huggingface.co/TheBloke/WizardLM-Uncensored-Falcon-40B-3bit-GPTQ)
> [TheBloke/samantha-falcon-7B-GPTQ ยท Hugging Face](https://huggingface.co/TheBloke/samantha-falcon-7B-GPTQ)
> [Open Source AI Has a New Champion](https://analyticsindiamag.com/open-source-ai-has-a-new-champion/)
> [The RefinedWeb Dataset for Falcon LLM: Outperforming Curated...](https://arxiv.org/abs/2306.01116)
|
38be0309cd56451ebb0f5da43bfe3508
|
|
GPT
|
Language Models
|
Aug 17, 2020
|
Alan Jo
|
Alan Jo
|
Jul 16, 2023
|
[OpenAI](https://texonom.com/openai-501e879eb1ca4ef8b388ba5b11e92e23) [Decoder Model](https://texonom.com/decoder-model-36e78a40265c473d90197089aebfa83b)
|
### Generative Pre-trained Transformer
Transformer์ ๋์ฝ๋ ๋ธ๋ญ์ 12๊ฐ ์์์ฌ๋ ค ๋ง๋ ๋ชจ๋ธ
### GPT Implementations
|Title|
|:-:|
|[GPT 1](https://texonom.com/gpt-1-d5b4cee52c894a31950a6dff5aec0195)|
|[GPT 2](https://texonom.com/gpt-2-6998797e5634436b9cdfece14aa1f3db)|
|[GPT 3](https://texonom.com/gpt-3-a69d4bc827004e0690b36132632f46b0)|
|[GPT 4](https://texonom.com/gpt-4-0d74adcc08b44c23b6f8b0b17e54cc3e)|
|[GPT F](https://texonom.com/gpt-f-5fb785ea644a43e3ada98450f138655d)|
|[nanoGPT](https://texonom.com/nanogpt-b1755b382e244584915cc7e837dd0773)|
|[minGPT](https://texonom.com/mingpt-dc6fe2127806478dbcf08d495898d94a)|
|[gpt-neo](https://texonom.com/gpt-neo-c8fd4e28563c498c80a8101b3472bc4b)|
|[GPT J](https://texonom.com/gpt-j-3b4f0a39a5a343398ee1e73f93ed0e4b)|
### History
> [Five years of GPT progress](https://finbarr.ca/five-years-of-gpt-progress)
### GPT in simple explanation
> [GPT in 60 Lines of NumPy | Jay Mody](https://jaykmody.com/blog/gpt-from-scratch/)
> [๋ฏธ๋ผํด ์ดํค๋ MiraKle Ahead](https://mirakle.mk.co.kr/list.php?sc=51800017)
> [์ธ๊ณต์ง๋ฅ GPT-f๊ฐ ์์ฑํ ์ํ ์ฆ๋ช
, ๊ณต์ ์ํ์ปค๋ฎค๋ํฐ์์ ์ฑํ - AIํ์์ค](https://easybranches.com/feed/kr/science/1640356)
|
d07bb69aa3794c44b3cdc1347132f75b
|
KOSMOS-1
|
Language Models
|
Apr 9, 2023
|
Alan Jo
|
Alan Jo
|
Apr 9, 2023
|
[Multimodal AI](https://texonom.com/multimodal-ai-381b5ebe1218438689de97941a3d9829) [MLLM](https://texonom.com/mllm-98172e9446c04bcc9cf52b2bc5d0bd17) [Language Is Not All You Need](https://texonom.com/language-is-not-all-you-need-0a6f8ec4af2f44bfacf6d4da5dd5da02)
|
[Zero shot learning](https://texonom.com/zero-shot-learning-8c92d9386f6648f5b877cb593ec2747b) ์์ ๋ฐ์ด๋จ
Masked Contrastive Predictive Coding (MCPC)๋ฅผ ์ฌ์ฉํด ์
๋ ฅ ๋ฐ์ดํฐ์ ์ผ๋ถ๋ฅผ ๋ง์คํนํ๊ณ , ๋ง์คํน๋ ๋ถ๋ถ์ ์์ธกํ๋ ์์
์ ์ํ [Contrastive Learning](https://texonom.com/contrastive-learning-73afe4f0d2554be392d775a0a05fe328)
### KOSMOS-1 Usages
|Title|
|:-:|
> [Untitled](https://arxiv.org/pdf/2302.14045.pdf)
|
eef4f69c8dfb4fab8cc3ee3019ad2e6f
|
LLaMA
|
Language Models
|
Feb 25, 2023
|
Alan Jo
|
Alan Jo
|
Aug 10, 2023
|
[llama](https://github.com/facebookresearch/llama)
|
I thought it would be a few more years before I could run a GPT-3 class model on hardware that I owned. I was wrong: that future is here already.
### LLaMA Notion
|Title|
|:-:|
|[LLaMa History](https://texonom.com/llama-history-55dacd67da554bc5bde17eca29e150ad)|
|[LLaMA Implementation](https://texonom.com/llama-implementation-35366933da844f94abe3e1e1d0fb0089)|
|[LLaMA2](https://texonom.com/llama2-9e8d55f66932447680be7fc4bfa4e050)|
### LLaMA D**escendents**
|Title|
|:-:|
|[Stanford Alpaca](https://texonom.com/stanford-alpaca-30e03e4e4dd94ad8957374e1e24caaa4)|
|[Vicuna](https://texonom.com/vicuna-ea1568170bf84148a8d06414a84eaff4)|
|[](https://texonom.com/b57e49f633254fcab08ffcbf4746cdac)|
|[ ](https://texonom.com/fa6139dc98ef4a6c81a62c8ab1106a34)|
|[Orca](https://texonom.com/orca-b0f8e611a23f47dfaad46f488a7be675)|
|[OpenLLaMa](https://texonom.com/openllama-800c80f8f5b94ce9aaa185388a49dee4)|
|[GOod at Arithmetic Tasks](https://texonom.com/good-at-arithmetic-tasks-6288a09f4d7b4713a8b47d6249167fe5)|
> [Llama from scratch (or how to implement a paper without crying)](https://blog.briankitano.com/llama-from-scratch/)
### Even can run on raspberry Pi 4 lol and Pixel 6
> [Large language models are having their Stable Diffusion moment](https://simonwillison.net/2023/Mar/11/llama/?utm_source=tldrnewsletter)
> [Introducing LLaMA: A foundational, 65-billion-parameter language model](https://ai.facebook.com/blog/large-language-model-llama-meta-ai/)
> [Meta unveils a new large language model that can run on a single GPU [Updated]](https://arstechnica.com/information-technology/2023/02/chatgpt-on-your-pc-meta-unveils-new-ai-model-that-can-run-on-a-single-gpu)
|
f2b6721202d44d469add84d8a366809c
|
MPT
|
Language Models
|
May 22, 2023
|
Alan Jo
|
Alan Jo
|
Jul 9, 2023
|
[MosaicML](https://texonom.com/mosaicml-d531ff0e6a51448c84296ab20f675d67) [LLM Context](https://texonom.com/llm-context-252aa3db0c164002aba92482142bcc2d)
|
**MosaicML releases language model with 65,000 tokens context**
### MPT Usages
|Title|
|:-:|
> [MPT-7B: The best open-source LLM available for commercial use](https://the-decoder.com/mpt-7b-the-best-open-source-llm-available-for-commercial-use/)
|
bf2abacce4b14fe99929c1fb19ea2c71
|
PaLM
|
Language Models
|
Apr 7, 2022
|
Alan Jo
|
Alan Jo
|
Jun 26, 2023
|
[PaLM-rlhf-pytorch](https://github.com/lucidrains/PaLM-rlhf-pytorch) [Google AI](https://texonom.com/google-ai-8a4f2d711eac466aaa97b12647f8eeaa)
|
5400์ต parameter
### PaLM Versions
|Title|
|:-:|
|[PaLM 1](https://texonom.com/palm-1-9eb81dc927734b05bd0813a33f12a146)|
|[PaLM 2](https://texonom.com/palm-2-c74baec55e28433093feb97b5ca572d1)|
### PaLM API
> [PaLM API & MakerSuite: an approachable way to start prototyping and building generative AI applications](https://developers.googleblog.com/2023/03/announcing-palm-api-and-makersuite.html)
> [Google opens up its AI language model PaLM to challenge OpenAI and GPT-3](https://www.theverge.com/2023/3/14/23639313/google-ai-language-model-palm-api-challenge-openai)
> [You're Temporarily Blocked](https://m.facebook.com/story.php?story_fbid=pfbid02a99wk7JkBNAuWwgzQG1xVCr3iiLFB4ttSfychoAqcSpq2W3diDeNGGkCLPffrVLrl&id=100001466253230)
|
a44fb7aaf8234a828d7b0f87d1fa243f
|
Qwen
|
Language Models
|
Aug 7, 2023
|
Alan Jo
|
Alan Jo
|
Aug 7, 2023
|
### Qwen Usages
|Title|
|:-:|
|[ReAct Prompting](https://texonom.com/react-prompting-7edbefb5341b45d2a456b5df5b7c94f4)|
[Qwen-7B](https://github.com/QwenLM/Qwen-7B)
> [Qwen](https://github.com/QwenLM)
> [Qwen/Qwen-7B-Chat ยท Hugging Face](https://huggingface.co/Qwen/Qwen-7B-Chat)
|
98e9c8006e764682ab87af31c84d6bbf
|
|
Stable LM
|
Language Models
|
Jan 15, 2023
|
Alan Jo
|
Alan Jo
|
Apr 24, 2023
|
[https://github.com/stability-AI/stableLM/](https://github.com/stability-AI/stableLM/)
> [Stability AI announces new open-source large language model](https://www.theverge.com/2023/4/19/23689883/stability-ai-open-source-large-language-model-stablelm)
|
2274fed51886447abc7a9e8c1550f9b8
|
|
****YaLM 100B****
|
Language Models
|
Jun 24, 2022
|
Alan Jo
|
Alan Jo
|
Mar 5, 2023
|
[YaLM-100B](https://github.com/yandex/YaLM-100B)
|
> [Yandex Publishes YaLM 100B. It's the Largest GPT-Like Neural Network in Open Source](https://medium.com/yandex/yandex-publishes-yalm-100b-its-the-largest-gpt-like-neural-network-in-open-source-d1df53d0e9a6)
|
e3af4398a1af4b3a8453d7eb98a59739
|
Falcon Openassistant
|
Falcon
| null | null | null | null | null |
[FalconLite](https://texonom.com/falconlite-10423207bef04bc18cbb8f2fe54f2f76)
> [OpenAssistant (OpenAssistant)](https://huggingface.co/OpenAssistant)
|
77666e440c964fdab7bcca371dab911d
|
FalconLite
|
Falcon Openassistant
| null | null | null | null | null |
[Amazon](https://texonom.com/amazon-f1722d7da08e423eb63f149c707397f4) [AutoGPTQ](https://texonom.com/autogptq-8a1d898788434aa2bc00fb43fd34411d) [Dynamic NTK](https://texonom.com/dynamic-ntk-662e5354f1f74e5b8d2ee3f9f274ef84)
> [Untitled](https://github.com/awslabs/extending-the-context-length-of-open-source-llms/tree/main/script)
> [amazon/FalconLite ยท Hugging Face](https://huggingface.co/amazon/FalconLite)
|
10423207bef04bc18cbb8f2fe54f2f76
|
GPT 1
|
GPT Implementations
|
Aug 23, 2020
|
Alan Jo
|
Alan Jo
|
Jun 11, 2023
|
12๊ฐ์ Transformer Decoder layers
๋ชจ๋ธ์ ํฌ๊ธฐ๊ฐ Pretrained Model์ ์ฑ๋ฅ์ ์ํฅ์ ์ค๋ค๋ผ๋ ์์ฌ์
> [Instruction Tuning์ด๋?](https://velog.io/@nellcome/Instruction-Tuning์ด๋)
|
d5b4cee52c894a31950a6dff5aec0195
|
|
GPT 2
|
GPT Implementations
|
Sep 14, 2020
|
Alan Jo
|
Alan Jo
|
Jun 11, 2023
|
[Zero shot learning](https://texonom.com/zero-shot-learning-8c92d9386f6648f5b877cb593ec2747b) [Fewย shot learning](https://texonom.com/fewshot-learning-f2a97adcbc774f4abd677a6db974329d)
|
Reddit Data 40GB
Task์ ๋ฐ๋ฅธ Fine tuning ์์ด ๊ธฐ์กด Task์ SOTA ๋ชจ๋ธ๋ค์ ๋์ด์ฌ
์ ํ์ต๋ LLM ๋ชจ๋ธ ํ๋๋ก ๋ชจ๋ Task๋ฅผ ํ ์ ์์์ง๋ ๋ชจ๋ฅธ๋ค๋ ์ํฉํธ
1.5B
> [Algpt2 Part 2](https://bkkaggle.github.io/blog/algpt2/2020/07/17/ALGPT2-part-2.html#replicating-gpt-2)
|
6998797e5634436b9cdfece14aa1f3db
|
GPT 3
|
GPT Implementations
|
Sep 14, 2020
|
Alan Jo
|
Alan Jo
|
Jun 11, 2023
|
[In-context learning](https://texonom.com/in-context-learning-2f55f51a87a04785be78dc65ef752ca4) [RLHF](https://texonom.com/rlhf-4b184f9c9e8b4c7a8861fb6374e91aa6)
|
175B, 600GB data
[Text DaVinci 003](https://texonom.com/text-davinci-003-fba990de22c745a9ab9298657a7e4b94)

### Why gpt fails on complex task
> [openai-cookbook/techniques_to_improve_reliability.md at main ยท openai/openai-cookbook](https://github.com/openai/openai-cookbook/blob/main/techniques_to_improve_reliability.md)
### natural language to code
> [Microsoft uses GPT-3 to let you code in natural language](https://techcrunch.com/2021/05/25/microsoft-uses-gpt-3-to-let-you-code-in-natural-language/?utm_source=tldrnewsletter)
> [A robot wrote this entire article. Are you scared yet, human? | GPT-3](https://www.theguardian.com/commentisfree/2020/sep/08/robot-wrote-this-article-gpt-3)
> [OpenAI begins allowing customers to fine-tune GPT-3](https://venturebeat.com/2021/12/14/openai-begins-allowing-customers-to-fine-tune-gpt-3/)
> [https://simonwillison.net/2022/Jul/9/gpt-3-explain-code/](https://simonwillison.net/2022/Jul/9/gpt-3-explain-code/)
> [The GPT-3 Architecture, on a Napkin](https://dugas.ch/artificial_curiosity/GPT_architecture.html)
|
a69d4bc827004e0690b36132632f46b0
|
GPT 4
|
GPT Implementations
|
Aug 23, 2020
|
Alan Jo
|
Alan Jo
|
Jun 26, 2023
|
[Mixture-of-Experts](https://texonom.com/mixture-of-experts-eb0cadda3237429dbc3b61d14fab81d1)
|
10000 [H100](https://texonom.com/h100-572cbc29dbbd4bc598522bc4b3da0d38)
### 8 Experts? It is meaningful tho
> [GPT-4's Secret Has Been Revealed](https://thealgorithmicbridge.substack.com/p/gpt-4s-secret-has-been-revealed)
> [OpenAIโs GPT-4 exhibits โhuman-level performanceโ on professional benchmarks](https://arstechnica.com/information-technology/2023/03/openai-announces-gpt-4-its-next-generation-ai-language-model)
> [Product Hunt ๐ธ on Twitter](https://twitter.com/ProductHunt/status/1633954325994119170)
> [GPT-4 Rumors From Silicon Valley](https://thealgorithmicbridge.substack.com/p/gpt-4-rumors-from-silicon-valley)
> [12ft |](https://12ft.io/proxy?q=https://towardsdatascience.com/gpt-4-is-coming-soon-heres-what-we-know-about-it-64db058cfd45)
> [The Multi-modal, Multi-model, Multi-everything Future of AGI](https://lspace.swyx.io/p/multimodal-gpt4)
> [GPT-4 Is a Reasoning Engine](https://every.to/chain-of-thought/gpt-4-is-a-reasoning-engine)
|
0d74adcc08b44c23b6f8b0b17e54cc3e
|
GPT F
|
GPT Implementations
|
Sep 14, 2020
|
Alan Jo
|
Alan Jo
|
Apr 19, 2022
|
5fb785ea644a43e3ada98450f138655d
|
||
GPT J
|
GPT Implementations
|
Jun 19, 2023
|
Alan Jo
|
Alan Jo
|
Jun 19, 2023
|
after gpt neo, zero shot
|
3b4f0a39a5a343398ee1e73f93ed0e4b
|
|
gpt-neo
|
GPT Implementations
|
Aug 23, 2020
|
Alan Jo
|
Alan Jo
|
Jul 16, 2023
|
## Open source GPT3
> [MBZUAI/LaMini-Neo-1.3B ยท Hugging Face](https://huggingface.co/MBZUAI/LaMini-Neo-1.3B)
> [GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)
> [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo)
|
c8fd4e28563c498c80a8101b3472bc4b
|
|
minGPT
|
GPT Implementations
|
Sep 14, 2020
|
Alan Jo
|
Alan Jo
|
Jul 16, 2023
|
[Andrej Karpathy](https://texonom.com/andrej-karpathy-e77764924dec46f985c97fb3b87b4717) [nanoGPT](https://texonom.com/nanogpt-b1755b382e244584915cc7e837dd0773) [minGPT](https://github.com/karpathy/minGPT)
|
dc6fe2127806478dbcf08d495898d94a
|
|
nanoGPT
|
GPT Implementations
|
Jan 9, 2023
|
Alan Jo
|
Alan Jo
|
Jul 31, 2023
|
[Andrej Karpathy](https://texonom.com/andrej-karpathy-e77764924dec46f985c97fb3b87b4717) [nanoGPT](https://github.com/karpathy/nanoGPT)
|
[llama2.c](https://github.com/karpathy/llama2.c)
> m1 110token/s
|
b1755b382e244584915cc7e837dd0773
|
Text DaVinci 003
|
GPT 3
| null | null | null | null | null |
### GPT 3.5 structure
> [New GPT-3 model: text-davinci-003](https://www.reddit.com/r/OpenAI/comments/z7f3ki/new_gpt3_model_textdavinci003/)
> [OpenAI's latest GPT-3 model generates better and longer texts](https://the-decoder.com/openais-latest-gpt-3-model-generates-better-and-longer-texts/)
|
fba990de22c745a9ab9298657a7e4b94
|
[WizardLM](https://github.com/nlpxucan/WizardLM)
|
LLaMA Descendents
|
Apr 27, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
b57e49f633254fcab08ffcbf4746cdac
|
||
[FastChat](https://texonom.com/fastchat-4074369b09a744a8818c78ead5381289)
|
LLaMA Descendents
|
May 1, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
fa6139dc98ef4a6c81a62c8ab1106a34
|
||
GOod at Arithmetic Tasks
|
LLaMA Descendents
|
Jun 22, 2023
|
Alan Jo
|
Alan Jo
|
Jun 22, 2023
|
[goat](https://github.com/liutiedong/goat)
|
> [Goat: Fine-tuned LLaMA Outperforms GPT-4 on Arithmetic Tasks](https://arxiv.org/abs/2305.14201)
> [tiedong/goat-lora-7b ยท Hugging Face](https://huggingface.co/tiedong/goat-lora-7b)
|
6288a09f4d7b4713a8b47d6249167fe5
|
OpenLLaMa
|
LLaMA Descendents
|
Jun 20, 2023
|
Alan Jo
|
Alan Jo
|
Jun 29, 2023
|
[open_llama](https://github.com/openlm-research/open_llama)
[Lit GPT](https://texonom.com/lit-gpt-dbb56045388f4daca0ab71ad110785b9)
> [openlm-research/open_llama_13b ยท Hugging Face](https://huggingface.co/openlm-research/open_llama_13b)
|
800c80f8f5b94ce9aaa185388a49dee4
|
|
Orca
|
LLaMA Descendents
|
Jun 20, 2023
|
Alan Jo
|
Alan Jo
|
Jul 8, 2023
|
[Progressive Learning](https://texonom.com/progressive-learning-46a222e83bc64bd49f4c44c3d0519fdb)
|
[Open Orca](https://texonom.com/open-orca-91240599dea34eb1834cc63a645b9080)
> [Orca 13B: the New Open Source Rival for GPT-4 from Microsoft โ Kanaries](https://docs.kanaries.net/tutorials/ChatGPT/orca-13b)
> [Untitled](https://arxiv.org/pdf/2306.02707.pdf)
|
b0f8e611a23f47dfaad46f488a7be675
|
Stanford Alpaca
|
LLaMA Descendents
|
Apr 8, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
โฃ [alpaca.cpp](https://github.com/antimatter15/alpaca.cpp)
|
### Stanford Alpaca D**escendents**
|Title|
|:-:|
|[](https://texonom.com/875d961f46e24c10a2d8ac851a8f1bf6)|
|[Alpaca LoRA](https://texonom.com/alpaca-lora-efe1dade0d994e75abc3dd75aa8b75f2)|
|
30e03e4e4dd94ad8957374e1e24caaa4
|
Vicuna
|
LLaMA Descendents
|
Jun 2, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
### Vicuna D**escendents**
|Title|
|:-:|
|[Wizard Vicuna](https://texonom.com/wizard-vicuna-c555316c38a24097ab53b72a9067dacc)|
|[WizardLM](https://texonom.com/wizardlm-efb6e9e167794674a3f4fe115752d9ac)|
|
ea1568170bf84148a8d06414a84eaff4
|
|
Lit GPT
|
OpenLLaMa
| null | null | null | null | null |
[lit-llama](https://github.com/Lightning-AI/lit-llama)
[lit-gpt](https://github.com/Lightning-AI/lit-gpt)
- [nanoGPT](https://texonom.com/nanogpt-b1755b382e244584915cc7e837dd0773)
- [Flash Attention](https://texonom.com/flash-attention-ca8093deb6c648059aff1580b0ddbf68)
- [GPTQ](https://texonom.com/gptq-87428aee2a774b93906ab2213a1b6dc6) for int4
- [bitsandbytes](https://texonom.com/bitsandbytes-1575b433faaf455ba0d86cf5f7e5190b) for int8, fp4, nf4
- [LoRA](https://texonom.com/lora-fda3706be3674496898ad2e5e00007c9)
not available for
```typegit clone https://github.com/Lightning-AI/lit-gptcd lit-gpt# pip install pytorchpip install --index-url https://download.pytorch.org/whl/nightly/cu118 --pre 'torch>=2.1.0dev'pip install -r requirements.txtpip uninstall -y lightning; pip install -r requirements.txtpip install sentencepiece```
### Download mode and so annoying~
```typepython scripts/download.py --repo_id openlm-research/open_llama_3bpython scripts/convert_hf_checkpoint.py --checkpoint_dir checkpoints/openlm-research/open_llama_3bpython quantize/gptq.py --checkpoint_dir checkpoints/openlm-research/open_llama_3bpython generate/base.py --prompt "Hello, my name is" --checkpoint_dir checkpoints/openlm-research/open_llama_3b --quantize gptq.int4```
### mac
```typeexport PYTORCH_ENABLE_MPS_FALLBACK=1--precision 16-true```
> [Why We're Building Lit-LLaMA - Lightning AI](https://lightning.ai/pages/community/why-were-building-lit-llama/)
|
dbb56045388f4daca0ab71ad110785b9
|
Open Orca
|
Orca
| null | null | null | null | null |
> [Open-Orca/OpenOrca ยท Datasets at Hugging Face](https://huggingface.co/datasets/Open-Orca/OpenOrca)
|
91240599dea34eb1834cc63a645b9080
|
[Alpaca-Turbo](https://github.com/alexanderatallah/Alpaca-Turbo)
|
Stanford Alpaca Descendents
|
Jun 7, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
875d961f46e24c10a2d8ac851a8f1bf6
|
||
Alpaca LoRA
|
Stanford Alpaca Descendents
|
Jun 7, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
[alpaca-lora](https://github.com/kairess/alpaca-lora)
|
### Alpaca LoRA D**escendents**
|Title|
|:-:|
|[SuperCOT](https://texonom.com/supercot-a546e09b488741ed9455b508e0c37a96)|
|
efe1dade0d994e75abc3dd75aa8b75f2
|
SuperCOT
|
Alpaca LoRA Descendents
|
Jun 7, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
### SuperCOT D**escendents**
|Title|
|:-:|
|[Lazarus](https://texonom.com/lazarus-049b3aa5281548efbedea1431aa7b335)|
> [kaiokendev/SuperCOT-LoRA ยท Hugging Face](https://huggingface.co/kaiokendev/SuperCOT-LoRA)
|
a546e09b488741ed9455b508e0c37a96
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.