askai
nanoGPT
Our great sponsors
askai | nanoGPT | |
---|---|---|
2 | 69 | |
92 | 31,452 | |
- | - | |
10.0 | 5.4 | |
about 1 year ago | 26 days ago | |
Python | Python | |
MIT License | 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.
askai
-
Ask HN: How are you using GPT to be productive?
A while back, I wrote a simple CLI-wrapper around OpenAI's API that I'm using daily (https://github.com/maxvfischer/askai). I use it as an addition to Stackoverflow to ask quick programming-related question straight in the terminal.
- Show HN: I built a simple CLI helper integrating with GPT-3
nanoGPT
-
Show HN: Predictive Text Using Only 13KB of JavaScript. No LLM
Nice work! I built something similar years ago and I did compile the probabilities based on a corpus of text (public domain books) in an attempt to produce writing in the style of various authors. The results were actually quite similar to the output of nanoGPT[0]. It was very unoptimized and everything was kept in memory. I also knew nothing about embeddings at the time and only a little about NLP techniques that would certainly have helped. Using a graph database would have probably been better than the datastructure I came up with at the time. You should look into stuff like Datalog, Tries[1], and N-Triples[2] for more inspiration.
You're idea of splitting the probabilities based on whether you're starting the sentence or finishing it is interesting but you might be able to benefit from an approach that creates a "window" of text you can use for lookup, using an LCS[3] algorithm could do that. There's probably a lot of optimization you could do based on the probabilities of different sequences, I think this was the fundamental thing I was exploring in my project.
Seeing this has inspired me further to consider working on that project again at some point.
[0] https://github.com/karpathy/nanoGPT
[1] https://en.wikipedia.org/wiki/Trie
[2] https://en.wikipedia.org/wiki/N-Triples
[3] https://en.wikipedia.org/wiki/Longest_common_subsequence
-
LLMs Learn to Be "Generative"
where x1 denotes the 1st token, x2 denotes the 2nd token and so on, respectively.
I understand the conditional terms p(x_n|...) where we use cross-entropy to calculate their losses. However, I'm unsure about the probability of the very first token p(x1). How is it calculated? Is it in some configurations of the training process, or in the model architecture, or in the loss function?
IMHO, if the model doesn't learn p(x1) properly, the entire formula for Bayes' rule cannot be completed, and we can't refer to LLMs as "truly generative". Am I missing something here?
I asked the same question on nanoGPT repo: https://github.com/karpathy/nanoGPT/issues/432, but I haven't found the answer I'm looking for yet. Could someone please enlighten me.
-
A simulation of me: fine-tuning an LLM on 240k text messages
This repo, albeit "old" in regards to how much progress there's been in LLMs, has great simple tutorials right there eg. fine-tuning GPT2 with Shakespeare: https://github.com/karpathy/nanoGPT
-
Ask HN: Is it feasible to train my own LLM?
For training from scratch, maybe a small model like https://github.com/karpathy/nanoGPT or tinyllama. Perhaps with quantization.
-
Writing a C compiler in 500 lines of Python
It does remind me of a project [1] Andrej Karpathy did, writing a neural network and training code in ~600 lines (although networks have easier logic to code than a compiler).
-
[D] Can GPT "understand"?
But I'm still not convinced that it can't in theory. Maybe the training set or transformer size I'm using is too small. I'm using nanoGPT implementation (https://github.com/karpathy/nanoGPT) with layers 24, heads 12, and embeddings per head 32. I'm using character-based vocab: every digit is a separate token, +, = and EOL.
-
Transformer Attention is off by one
https://github.com/karpathy/nanoGPT/blob/f08abb45bd2285627d1...
At training time, probabilities for the next token are computed for each position, so if we feed in a sequence of n tokens, we basically get n training examples, one for each position, but at inference time, we only compute the next token since we’ve already output the preceding ones.
-
Sarah Silverman Sues ChatGPT Creator for Copyright Infringement
And there are a bunch of other efforts at making training more efficient. Here's a cool model by Karpathy (OpenAI/used to head up Tesla's efforts): https://github.com/karpathy/nanoGPT
-
Douglas Hofstadter changes his mind on Deep Learning and AI risk
Just being a part of any auto-regressive system does not contradict his statement.
Go look at the GPT training code, here is the exact line: https://github.com/karpathy/nanoGPT/blob/master/train.py#L12...
The model is only trained to predict the next token. The training regime is purely next-token prediction. There is no loopiness whatsoever here, strange or ordinary.
Just because you take that feedforward neural network and wrap it in a loop to feed it its own output does not change the architecture of the neural net itself. The neural network was trained in one direction and runs in one direction. Hofstadter is surprised that such an architecture yields something that looks like intelligence.
He specifically used the correct term "feedforward" to constrast with recurrent neural networks, which GPT is not: https://en.wikipedia.org/wiki/Feedforward_neural_network
-
NTK-Aware Scaled RoPE allows LLaMA models to have extended (8k+) context size without any fine-tuning and minimal perplexity degradation.
Does anyone have or know of an example implementation in plain pytorch, not huggingface transformers. Like something you could plug into https://github.com/karpathy/nanoGPT ?
What are some alternatives?
KoboldAI-Client
minGPT - A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training
text-generation-webui - A Gradio web UI for Large Language Models. Supports transformers, GPTQ, AWQ, EXL2, llama.cpp (GGUF), Llama models.
RWKV-LM - RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). 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.
llama.cpp - LLM inference in C/C++
PaLM-rlhf-pytorch - Implementation of RLHF (Reinforcement Learning with Human Feedback) on top of the PaLM architecture. Basically ChatGPT but with PaLM
469 - The bonelord language in Tibia
ChatGPT - 🔮 ChatGPT Desktop Application (Mac, Windows and Linux)
chatgpt-api - Node.js client for the official ChatGPT API. 🔥
nn-zero-to-hero - Neural Networks: Zero to Hero
ChatGPT-ProBot - 🤖️ A GPTChat based GitHub robot. dialogue/CR/etc..
gpt_index - LlamaIndex (GPT Index) is a project that provides a central interface to connect your LLM's with external data. [Moved to: https://github.com/jerryjliu/llama_index]