Gold IV
1,550 Elo
- #4,413 on the ladder
- −7 today · −18 this week
- Peak 1,568
9 Elo to Gold III
microsoft/LoRA
View on GitHubCode for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"
microsoft/LoRA is #4,413 on the bestofgithub Elo ladder, in Gold IV, with a rating of 1,550 (−18 this week). It has 13,923 stars, 10 gained in the last week.
Stars
13,923
- +0 today
- +10 this week
- – this month
Repository
- Forks
- 936
- Language
- Python
- Last push
- Category
- AI & Agents
Achievements
Maintainer? Show your league in your README
Card: your league's emblem, from Copper to Legendary
[](https://bestofgithub.dev/repo/microsoft/LoRA)Flat: fits beside your other badges
[](https://bestofgithub.dev/repo/microsoft/LoRA)Install
git clone https://github.com/microsoft/LoRA.gitSimilar repositories
EleutherAI/gpt-neo
Open on GitHubPythonAn implementation of model parallel GPT-2 and GPT-3-style models using the mesh-tensorflow library.
- 1,372+2 Elo, change today
- 8.2k stars
- 955 forks
- archived
git clone https://github.com/EleutherAI/gpt-neo.gitBlinkDL/RWKV-LM
Open on GitHubPythonRWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.
- 1,475−6 Elo, change today
- 14.7k stars
- 1k forks
git clone https://github.com/BlinkDL/RWKV-LM.gitEleutherAI/gpt-neox
Open on GitHubPythonAn implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries
- 1,396±0 Elo, change today
- 7.4k stars
- 1.1k forks
git clone https://github.com/EleutherAI/gpt-neox.gitREADME
Loading README…