TavernAI
RWKV-LM
TavernAI | RWKV-LM | |
---|---|---|
17 | 84 | |
76 | 11,704 | |
- | - | |
10.0 | 8.8 | |
about 1 year ago | 16 days ago | |
JavaScript | Python | |
- | Apache License 2.0 |
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.
TavernAI
-
Trying to download SillyTavern on Termux: Username for 'https://github.com":
I keep getting this last code after I input "git clone https://github.com/SillyLossy/TavernAI" what am I supposed to do..?
- How to get and use the Poe API/p-b cookie for SillyTavern through Kiwi Browser, Another 10 Step Android Guide
-
Dark future of AI generated girls
It's already the case with games like Silly Tavern https://github.com/SillyLossy/TavernAI and AI girls that are created on https://www.characterhub.org for AI personalities. Basically 4chan people use open source AI models like meta's Llama.cpp to create horny personalities to interact with, and use local models with SD to create big boobed waifus.
- 玩ai角色扮演导了一下午,一步步道德绑架纯洁少女和你做爱
-
SillyTavern Android, 10 Quick Step Guide.
1) Download Termux scroll down and look for download apk, install it. 2) Type in or copy and paste and hit enter apt update Enter, Press y to anything that comes up 3) Type in or copy and paste and hit enter apt update Enter, Press y to anything that comes up 4) Type in or copy and paste and hit enter pkg install git 5) Type in or copy and paste and hit enter git clone https://github.com/SillyLossy/TavernAI 6) Optional to get the most updated stuff add the dev branch if you just want it basic just skip this step Type in or copy and paste and hit enter git clone -b dev https://github.com/SillyLossy/TavernAI 7) Type in or copy and paste and hit enter cd SillyTavern The caps and lowercase HAS to specifically be this way. If you get green /SillyTavern you did it correctly. 8) Type in or copy and paste and hit enter pkg install nodejs 9) Type in or copy and paste and hit enter npm install 10) Type in or copy and paste and hit enter node server.js It should open your browser. You have correctly installed SillyTavern onto Termux. Under the plug you can click this link https://platform.openai.com/account/api-keys sign in and copy paste the link/api key
-
Comparing models: GPT4xAlpaca, Vicuna, and OASST
But if you're really serious about chatting, the best experience is definitely with TavernAI. It's just a frontend so you still run the AI using oobabooga's textgen or one of the *cpp engines, but because it's entirely focused on chatting, its chat capabilities are much more advanced.
-
KoboldCpp - Combining all the various ggml.cpp CPU LLM inference projects with a WebUI and API (formerly llamacpp-for-kobold)
Have you tried to talk to both at the same time? With TavernAI group chats are actually possible. The current version isn't compatible with koboldcpp, but the dev version has a fix, and I'm just getting started playing around with it.
-
What.
That looks like Cohee's TavernAI fork! https://github.com/SillyLossy/TavernAI
- Creating characters for TavenAI
- Jesus fucking christ what happened here?
RWKV-LM
-
Do LLMs need a context window?
https://github.com/BlinkDL/RWKV-LM#rwkv-discord-httpsdiscord... lists a number of implementations of various versions of RWKV.
https://github.com/BlinkDL/RWKV-LM#rwkv-parallelizable-rnn-w... :
> RWKV: Parallelizable RNN with Transformer-level LLM Performance (pronounced as "RwaKuv", from 4 major params: R W K V)
> RWKV is an RNN with Transformer-level LLM performance, which can also be directly trained like a GPT transformer (parallelizable). And it's 100% attention-free. You only need the hidden state at position t to compute the state at position t+1. You can use the "GPT" mode to quickly compute the hidden state for the "RNN" mode.
> So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding (using the final hidden state).
> "Our latest version is RWKV-6,*
- People who've used RWKV, whats your wishlist for it?
- Paving the way to efficient architectures: StripedHyena-7B
-
Understanding Deep Learning
That is not true. There are RNNs with transformer/LLM-like performance. See https://github.com/BlinkDL/RWKV-LM.
-
Q-Transformer: Scalable Reinforcement Learning via Autoregressive Q-Functions
This is what RWKV (https://github.com/BlinkDL/RWKV-LM) was made for, and what it will be good at.
Wow. Pretty darn cool! <3 :'))))
-
Personal GPT: A tiny AI Chatbot that runs fully offline on your iPhone
Thanks for the support! Two weeks ago, I'd have said longer contexts on small on-device LLMs are at least a year away, but developments from last week seem to indicate that it's well within reach. Once the low hanging product features are done, I think it's a worthy problem to spend a couple of weeks or perhaps even months on. Speaking of context lengths, recurrent models like RWKV technically have infinite context lengths, but in practice the context slowly fades away after a few thousands of tokens.
-
"If you see a startup claiming to possess top-secret results leading to human level AI, they're lying or delusional. Don't believe them!" - Yann LeCun, on the conspiracy theories of "X company has reached AGI in secret"
This is the reason there are only a few AI labs, and they show little of the theoretical and scientific understanding you believe is required. Go check their code, there's nothing there. Even the transformer with it's heads and other architectural elements turns out to not do anything and it is less efficient than RNNs. (see https://github.com/BlinkDL/RWKV-LM)
-
The Secret Sauce behind 100K context window in LLMs: all tricks in one place
I've been pondering the same thing, as simply extending the context window in a straightforward manner would lead to a significant increase in computational resources. I've had the opportunity to experiment with Anthropics' 100k model, and it's evident that they're employing some clever techniques to make it work, albeit with some imperfections. One interesting observation is that their prompt guide recommends placing instructions after the reference text when inputting lengthy text bodies. I noticed that the model often disregarded the instructions if placed beforehand. It's clear that the model doesn't allocate the same level of "attention" to all parts of the input across the entire context window.
Moreover, the inability to cache transformers makes the use of large context windows quite costly, as all previous messages must be sent with each call. In this context, the RWKV-LM project on GitHub (https://github.com/BlinkDL/RWKV-LM) might offer a solution. They claim to achieve performance comparable to transformers using an RNN, which could potentially handle a 100-page document and cache it, thereby eliminating the need to process the entire document with each subsequent query. However, I suspect RWKV might fall short in handling complex tasks that require maintaining multiple variables in memory, such as mathematical computations, but it should suffice for many scenarios.
On a related note, I believe Anthropics' Claude is somewhat underappreciated. In some instances, it outperforms GPT4, and I'd rank it somewhere between GPT4 and Bard overall.
-
Meta's plan to offer free commercial AI models puts pressure on Google, OpenAI
> The only reason open-source LLMs have a heartbeat is they’re standing on Meta’s weights.
Not necessarily.
RWKV, for example, is a different architecture that wasn't based on Facebook's weights whatsoever. I don't know where BlinkDL (the author) got the training data, but they seem to have done everything mostly independently otherwise.
https://github.com/BlinkDL/RWKV-LM
disclaimer: I've been doing a lot of work lately on an implementation of CPU inference for this model, so I'm obviously somewhat biased since this is the model I have the most experience in.
-
Eliezer Yudkowsky - open letter on AI
I think the main concern is that, due to the resources put into LLM research for finding new ways to refine and improve them, that work can then be used by projects that do go the extra mile and create things that are more than just LLMs. For example, RWKV is similar to an LLM but will actually change its own model after every processed token, thus letting it remember things longer-term without the use of 'context tokens'.
What are some alternatives?
SillyTavern - LLM Frontend for Power Users.
llama - Inference code for Llama models
SillyTavern - LLM Frontend for Power Users. [Moved to: https://github.com/SillyTavern/SillyTavern]
alpaca-lora - Instruct-tune LLaMA on consumer hardware
koboldcpp - A simple one-file way to run various GGML and GGUF models with KoboldAI's UI
flash-attention - Fast and memory-efficient exact attention
TavernAI-extras - Extensions API for SillyTavern [Moved to: https://github.com/Cohee1207/SillyTavern-extras]
TavernAI - Atmospheric adventure chat for AI language models (KoboldAI, NovelAI, Pygmalion, OpenAI chatgpt, gpt-4)
gpt4all - gpt4all: run open-source LLMs anywhere
Spermack
RWKV-CUDA - The CUDA version of the RWKV language model ( https://github.com/BlinkDL/RWKV-LM )