microsoft/DeepSpeed
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Language: Python
Stars trend:
11 Apr 2023

 5pm ▎ 2

 6pm ▏ 1

 7pm ▎ 2

 8pm ▋ 5

 9pm █▋ 13

10pm ▌ 4

11pm █▊ 14

12 Apr 2023

12am █▊ 14

 1am ██▍ 19

 2am ██▊ 22

 3am ████▉ 39

 4am ██▌ 20

#python
#billionparameters, #compression, #dataparallelism, #deeplearning, #gpu, #inference, #machinelearning, #mixtureofexperts, #modelparallelism, #pipelineparallelism, #pytorch, #trillionparameters, #zero

dvmazur/mixtral-offloading
Run Mixtral-8x7B models in Colab or consumer desktops
Language:Python
Total stars: 521
Stars trend:
1 Jan 2024
6am ▍ +3
7am +0
8am ▎ +2
9am ▍ +3
10am +0
11am ▉ +7
12pm ███ +24
1pm ██▋ +21
2pm ███▏ +25
3pm ██▋ +21
4pm ███▍ +27
5pm ████ +32
#python
#colabnotebook, #deeplearning, #googlecolab, #languagemodel, #llm, #mixtureofexperts, #offloading, #pytorch, #quantization

hiyouga/LLaMA-Factory
Unify Efficient Fine-tuning of 100+ LLMs
Language:Python
Total stars: 12204
Stars trend:
28 Feb 2024
2am ▋ +5
3am ▋ +5
4am +0
5am +0
6am ▉ +7
7am █ +8
8am █▎ +10
9am ▉ +7
10am ▎ +2
11am ▍ +3
12pm ▋ +5
1pm ▍ +3
#python
#agent, #baichuan, #chatglm, #finetuning, #generativeai, #gpt, #instructiontuning, #languagemodel, #largelanguagemodels, #llama, #llm, #lora, #mistral, #mixtureofexperts, #peft, #qlora, #quantization, #qwen, #rlhf, #transformers

codelion/optillm
Optimizing inference proxy for LLMs
Language:Python
Total stars: 512
Stars trend:
17 Sep 2024
4am ▊ +6
5am ▎ +2
6am ▊ +6
7am █▏ +9
8am ▋ +5
9am ▎ +2
10am ▉ +7
11am ▊ +6
12pm ▏ +1
1pm ▋ +5
2pm █▌ +12
3pm ██ +16
#python
#agent, #agenticai, #agenticframework, #agenticworkflow, #agents, #apigateway, #genai, #largelanguagemodels, #llm, #llminference, #llmapi, #mixtureofexperts, #moa, #openai, #openaiapi, #optimization, #proxyserver

rhymes-ai/Aria
Codebase for Aria - an Open Multimodal Native MoE
Language:Jupyter Notebook
Total stars: 88
Stars trend:
10 Oct 2024
2am ▎ +2
3am ▎ +2
4am ▍ +3
5am ▏ +1
6am ▏ +1
7am ▉ +7
8am ▊ +6
9am █▎ +10
10am ▊ +6
11am ██ +16
12pm █▎ +10
1pm █▋ +13
#jupyternotebook
#mixtureofexperts, #multimodal, #visionandlanguage