Giters

clayrisser / awesome-ai-models

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Awesome-ai-models

a curated list of awesome ai models

Environments

  • Server environment
  • Local Models

Server environment:

  • Localai
  • Transformers

Local environment Models:

  • Transformers.js
  • api
  • core-ml
  • windows-ml
  • ml-kit

Server Environments:

  • Localai

NLP (Natural Language Processing)

[✓] baichuan-vicuna-7b.ggmlv3.q4_0

Baichuan-7B is an open-source large-scale pre-trained model developed by Baichuan Intelligent Technology. Based on the Transformer architecture, it is a model with 7 billion parameters trained on approximately 1.2 trillion tokens. It supports both Chinese and English, with a context window length of 4096.

STEPS:

Download the model-> wget https://huggingface.co/TheBloke/baichuan-vicuna-7B-GGML/resolve/main/baichuan-vicuna-7b.ggmlv3.q4_0.bin -O models/baichuan-vicuna-7b.ggmlv3.q4_0

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{ "model": "baichuan-vicuna-7b.ggmlv3.q4_0", "messages": [{"role": "user", "content": "i need a coffee for"}], "temperature": 0.9 }'

  • [✓] rest api

[✓] ggml-gpt4all-j

A commercially licensable model based on GPT-J and trained by Nomic AI on the v0 GPT4All dataset.

The prompt below is a question to answer, a task to complete, or a conversation to respond to; decide which and write an appropriate response.

STEPS:

Download the model-> wget https://gpt4all.io/models/ggml-gpt4all-j.bin -O models/ggml-gpt4all-j.bin

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{ "model": "ggml-gpt4all-j", "messages": [{"role": "user", "content": "How are you?"}], "temperature": 0.9 }'

  • [✓] rest api

[] ggml-mpt-7b-base

MPT-7B is a decoder-style transformer pretrained from scratch on 1T tokens of English text and code. This model was trained by MosaicML.

  • this model generates number of possible answers for a given question.

STEPS:

Download the model-> wget https://gpt4all.io/models/ggml-mpt-7b-base.bin -O models/ggml-mpt-7b-base.bin

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{ "model": "ggml-mpt-7b-base", "messages": [{"role": "user", "content": "How are you?"}], "temperature": 0.9 }'

  • [✓] rest api

[✓] ggml-mpt-7b-instruct

This model generates the better instructions that describes a task. Write a response that appropriately completes the request.

STEPS:

Download the model-> wget https://gpt4all.io/models/ggml-mpt-7b-instruct.bin -O models/ggml-mpt-7b-instruct.bin

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{ "model": "ggml-mpt-7b-instruct", "messages": [{"role": "user", "content": "teach me how to use linux terminal"}], "temperature": 0.9 }'

response:

  • need a tutorial that will teach me the basics of using it.

  • [✓] rest api

[✓] open-llama-7B-open-instruct.ggmlv3.q4_0

Below is an instruction that describes a task. Write a response that appropriately completes the request

Instruction: {{.Input}}
 Response: I'm new to linux and iptables. I want to know how to use them to block specific ip addresses from connecting to my server.

STEPS:

Download the model-> wget https://huggingface.co/TheBloke/open-llama-7b-open-instruct-GGML/resolve/main/open-llama-7B-open-instruct.ggmlv3.q4_0.bin -O models/open-llama-7B-open-instruct.ggmlv3.q4_0
request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{"model": "open-llama-7B-open-instruct.ggmlv3.q4_0", "messages": [{"role": "user", "content": "teach me how to use linux terminal"}], "temperature": 0.9 }'

-[✓] rest api

[✓] open-llama-7b-q4_0

This model fills the sentence with the most appropriate word.

STEPS:

Download the model-> wget https://huggingface.co/SlyEcho/open_llama_7b_ggml/resolve/main/open-llama-7b-q4_0.bin -O models/open-llama-7b-q4_0

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{"model": "open-llama-7b-q4_0", "messages": [{"role": "user", "content": "i had a dream to control my"}], "temperature": 0.9 }'

  • [✓] rest api

[✓] open-llama-3b-q4_0

This model fills the sentence with the most appropriate word.

STEPS:

Download the model-> wget https://huggingface.co/SlyEcho/open_llama_3b_ggml/resolve/main/open-llama-3b-q4_0.bin -O models/open-llama-3b-q4_0

request:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{"model": "open-llama-3b-q4_0", "messages": [{"role": "user", "content": "i had a dream to control my"}], "temperature": 0.9 }'

  • [✓] rest api

Local Environments:

api

Steps to call the api:

url = https://api-inference.huggingface.co/models/${model_name} (ex. model_name = stabilityai/stable-diffusion-2-1)

I mentioned the models below for each experience with related few working models sir.

fill-mask:

related models:
- bert-base-uncased
- roberta-base
- roberta-large

- NOTE: THis model needs [MASK] token for the input sentence.

ex-request:

Sentiment Analysis

related models:
- nlptown/bert-base-multilingual-uncased-sentiment

Question Answering:

related models:
- alphakavi22772023/bertQA
- timpal0l/mdeberta-v3-base-squad2
- LLukas22/all-MiniLM-L12-v2-qa-en

Sentence Similarity:

related models:
- intfloat/e5-large-v2
- jinaai/jina-embedding-s-en-v1
- LLukas22/all-MiniLM-L12-v2-qa-en

Summarization:

related models:
- facebook/bart-large-cnn
- csebuetnlp/mT5_multilingual_XLSum
- slauw87/bart_summarisation

Text Classification:

related models:
- SamLowe/roberta-base-go_emotions
- ProsusAI/finbert
- nlptown/bert-base-multilingual-uncased-sentiment

Text Generation:

related models:
- gpt2
- HuggingFaceH4/starchat-beta-
- fxmarty/tiny-testing-gpt2-remote-code

Text-to-text:

    related models:
    - humarin/chatgpt_paraphraser_on_T5_base
    - juierror/flan-t5-text2sql-with-schema
    - voidful/context-only-question-generator

Token Classification:

related models:
- dslim/bert-base-NER
- DOOGLAK/01_SR_500v9_NER_Model_3Epochs_AUGMENTED
- DOOGLAK/Paraphrased_500v3_NER_Model_3Epochs_AUGMENTED

Translation:

related models:
- Helsinki-NLP/opus-mt-fi-mtt5-base
- Helsinki-NLP/opus-mt-fr-tvl
- Helsinki-NLP/opus-mt-eu-ru

Zero-Shot Classification:

related models:
- MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli
- Narsil/deberta-large-mnli-zero-cls
- HiTZ/A2T_RoBERTa_SMFA_TACRED-re

Computer Vision

Image Classification:

related models:
- google/vit-base-patch16-224
- microsoft/swinv2-tiny-patch4-window8-256
- victor/autotrain-satellite-image-classification-40975105875

Image Segmentation:

related models:
- mattmdjaga/segformer_b2_clothes
- nvidia/segformer-b3-finetuned-ade-512-512
- zoheb/mit-b5-finetuned-sidewalk-semantic

Object Detection:

related models:
- facebook/detr-resnet-50
- hustvl/yolos-tiny\
- facebook/detr-resnet-50-dc5
- facebook/detr-resnet-101

Audio

Automatic Speech Recognition:

related models:
- openai/whisper-tiny.en
- m3hrdadfi/wav2vec2-large-xlsr-persian-v3
- facebook/s2t-medium-librispeech-asr

Multi-modal

Feature Extraction:

related models:
- facebook/bart-large
- cambridgeltl/SapBERT-from-PubMedBERT-fulltext
- DeepPavlov/rubert-base-cased-sentence

Image-to-Text:

related models:
- nlpconnect/vit-gpt2-image-captioning
- microsoft/trocr-small-printed
- naver-clova-ix/donut-base-finetuned-cord-v1

Zero-Shot Image Classification:

related models:
-  openai/clip-vit-large-patch14\
- laion/CLIP-ViT-H-14-laion2B-s32B-b79K
- laion/CLIP-ViT-B-32-CommonPool.M.basic-s128M-b4K

Text-to-image:

related models:
- stabilityai/stable-diffusion-2-1

ex-request:
- curl https://api-inference.huggingface.co/models/stabilityai/stable-diffusion-2-1 -H "Content-Type: application.json" -d '{"inputs":"make a  gun image  ",""responseType":"blob" }' --output output.jpg

[X] TransformersJS WEB
[✓] Proxy
  • [NOTE] ( --output output.jpg) is used to save the output image.

About