gpt-neox
FASTER
Our great sponsors
| gpt-neox | FASTER | |
|---|---|---|
| 52 | 8 | |
| 6,569 | 6,199 | |
| 2.2% | 3.7% | |
| 8.9 | 7.2 | |
| 4 days ago | 11 days ago | |
| Python | C# | |
| Apache License 2.0 | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
gpt-neox
- FLaNK Stack 26 February 2024
- GPT-Neox
- GPT-NeoX
-
Best open source LLM model for commercial use
Gpt neox 20B can be used commerically.
- Do not register domains with the word "gpt" in it!
-
Read this post if you have general questions
GPT-Neo: GPT-Neo is a free and open-source language model developed by EleutherAI. It is a powerful model that can be used for a variety of tasks, including text generation, and question-answering. here is the GitHub
-
What's the current state of actually free and open source LLMs?
Doesn't gpt-neox 20b require like 40gb+ of VRAM? From their github repo the slim weights are 39GB and I think one of the devs has previously mentioned aiming for 48GB for inference.
-
Any real competitor to GPT-3 which is open source and downloadable?
3.) EleutherAI's GPT-Neo and GPT-NeoX: EleutherAI is an independent research organization that aims to promote open research in artificial intelligence. They have released GPT-Neo, an open-source language model based on the GPT architecture, and are developing GPT-NeoX, a highly-scalable GPT-like model. You can find more information on their GitHub repositories: GPT-Neo: https://github.com/EleutherAI/gpt-neo GPT-NeoX: https://github.com/EleutherAI/gpt-neox
-
Whatever happened to quantum computing?
It's really not that complicated. https://github.com/EleutherAI/gpt-neox
-
Fuck Luka Inc. WIP AI Freedom!
this forces me to create my own ai using gpt-neox .. i was being lazy, completely satisfied with my little ai baby .. i'll keep the account but not renew pro .. fuck luka, inc ... want it done right, do it yourself .. the python isn't that difficult and i have a spare linux vps .. my new work in progress .. freedom #!
FASTER
-
A MySQL compatible database engine written in pure Go
You would be surprised by performance of modern .NET :)
Writing no-alloc is oftentimes done by reducing complexity and not doing "stupid" tricks that actually work against JIT and CoreLib features.
For databases specifically, .NET is actually positioned very well with its low-level features (intrisics incl. SIMD, FFI, struct generics though not entirely low-level) and high-throughput GC.
Interesting example of this applied in practice is Garnet[0]/FASTER[1]. Keep in mind that its codebase still consist of un-idiomatic C# and you can do way better by further simplification, but it already does the job well enough.
[0] https://github.com/microsoft/garnet
[1] https://github.com/microsoft/FASTER
- FLaNK Stack 26 February 2024
- Fast persistent recoverable log and key-value store
- GitHub - microsoft/FASTER: Fast persistent recoverable log and key-value store + cache, in C# and C++.
-
FoundationDB: A Distributed Unbundled Transactional Key Value Store
A vaguely similar project that might be of interest is: https://github.com/microsoft/FASTER
It's also an "unbundled" low-level component that one could use as the foundation for a database engine or whatever. According to Microsoft, FASTER is not just "fast", but significantly faster than even some basic in-memory data structures that ship in the .NET standard library!
The downside is that it doesn't (yet) support some more advanced features like multi-server distributed mode.
However, that relative simplicity may be preferred in some scenarios...
-
Event Sourcing
Last time i looked into it there weren't that many i could find. There is https://github.com/tikv/tikv which uses rocksdb with raft. and there is faster https://github.com/microsoft/FASTER/ .
-
Experiences with Concurrent Hash Map Libraries
you could use fasterkv https://github.com/microsoft/FASTER
-
Faster A fast concurrent persistent key-value store and log, in C# and C++
FTA, https://github.com/Microsoft/FASTER/wiki/Performance-of-FAST...
What are some alternatives?
fairseq - Facebook AI Research Sequence-to-Sequence Toolkit written in Python.
libcuckoo - A high-performance, concurrent hash table
gpt-neo - An implementation of model parallel GPT-2 and GPT-3-style models using the mesh-tensorflow library.
parallel-hashmap - A family of header-only, very fast and memory-friendly hashmap and btree containers.
DeepSpeed - DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
foundationdb - FoundationDB - the open source, distributed, transactional key-value store
YaLM-100B - Pretrained language model with 100B parameters
plumber - A swiss army knife CLI tool for interacting with Kafka, RabbitMQ and other messaging systems.
open-ai - OpenAI PHP SDK : Most downloaded, forked, contributed, huge community supported, and used PHP (Laravel , Symfony, Yii, Cake PHP or any PHP framework) SDK for OpenAI GPT-3 and DALL-E. It also supports chatGPT-like streaming. (ChatGPT AI is supported)
tikv - Distributed transactional key-value database, originally created to complement TiDB
lm-evaluation-harness - A framework for few-shot evaluation of language models.
Folly - An open-source C++ library developed and used at Facebook.