torchscale
bertviz
| torchscale | bertviz | |
|---|---|---|
| 2 | 15 | |
| 2,927 | 6,398 | |
| 1.6% | - | |
| 7.2 | 3.9 | |
| 25 days ago | 9 months ago | |
| Python | Python | |
| MIT License | 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.
torchscale
-
Retentive Network: A Successor to Transformer Implemented in PyTorch
A retnet commit has now appeared in Microsoft's torchscale repo:
https://github.com/microsoft/torchscale/commit/bf65397b26469...
- [R] TorchScale: Transformers at Scale - Microsoft 2022 Shuming Ma et al - Improves modeling generality and capability, as well as training stability and efficiency.
bertviz
-
StreamingLLM: tiny tweak to KV LRU improves long conversations
This seems only to work cause large GPTs have redundant, undercomplex attentions. See this issue in BertViz about attention in Llama: https://github.com/jessevig/bertviz/issues/128
-
[D] Is there a tool that indicates which parts of the input prompt impact the LLM's output the most?
https://github.com/jessevig/bertviz this could be helpful .. I was playing around with it a while ago to see how the attention weights are distributed across prompts
-
Show HN: Fully client-side GPT2 prediction visualizer
It would be interesting to have attention visualized as well, similar to how it's done in BertViz:
https://github.com/jessevig/bertviz
-
How to visualise LLMs ?
link for lazy: https://github.com/jessevig/bertviz
-
Ask HN: Can someone ELI5 Transformers and the “Attention is all we need” paper
The Illustrated Transfomer ( https://jalammar.github.io/illustrated-transformer/ ) and Visualizing attention ( https://towardsdatascience.com/deconstructing-bert-part-2-vi... ), are both really good resources. For a more ELI5 approach this non-technical explainer ( https://www.parand.com/a-non-technical-explanation-of-chatgp... ) covers it at a high level.
- Perplexity.ai Prompt Leakage
-
[Discussion] is attention an explanation?
You can get some information this way, but not everything you would want to know. You can try it yourself with BertViz.
-
using bert for relation extraction
2) BERT learns a lot in its embeddings: the BERTOLOGY paper (https://arxiv.org/abs/2002.12327) provides a great in-depth look at some of the broader linguistic traits that BERT learns. Different layers often learn different patterns, so the embeddings aren't really interpretable, but you can use something like bertviz (https://github.com/jessevig/bertviz) to explore attention weights across layers for predetermined examples
-
Maintaining context vs. overloading your Replika
I messed up a few things and mixed a couple others, anyways this site has a lot of decent information about it. https://towardsdatascience.com/deconstructing-bert-part-2-visualizing-the-inner-workings-of-attention-60a16d86b5c1
-
[D] code to visualize attention heads
Big fan of BertViz for this, widely used in research for this very purpose: https://github.com/jessevig/bertviz
What are some alternatives?
towhee - Towhee is a framework that is dedicated to making neural data processing pipelines simple and fast.
ecco - Explain, analyze, and visualize NLP language models. Ecco creates interactive visualizations directly in Jupyter notebooks explaining the behavior of Transformer-based language models (like GPT2, BERT, RoBERTA, T5, and T0).
fairscale - PyTorch extensions for high performance and large scale training.
FARM - :house_with_garden: Fast & easy transfer learning for NLP. Harvesting language models for the industry. Focus on Question Answering.
extreme-bert - ExtremeBERT is a toolkit that accelerates the pretraining of customized language models on customized datasets, described in the paper “ExtremeBERT: A Toolkit for Accelerating Pretraining of Customized BERT”.
BERT-pytorch - Google AI 2018 BERT pytorch implementation
xformers - Hackable and optimized Transformers building blocks, supporting a composable construction.
transformers - 🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
glami-1m - The largest multilingual image-text classification dataset. It contains fashion products.
transformer-pytorch - Transformer: PyTorch Implementation of "Attention Is All You Need"
DeBERTa - The implementation of DeBERTa