LongLoRA
stanford_alpaca
LongLoRA | stanford_alpaca | |
---|---|---|
4 | 108 | |
2,508 | 29,026 | |
2.5% | 0.7% | |
8.6 | 2.0 | |
13 days ago | 3 months ago | |
Python | Python | |
Apache License 2.0 | 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.
LongLoRA
-
Ask HN: AI/ML papers to catch up with current state of AI?
LongAlpaca / One of many ways to extend context, and a useful dataset / https://arxiv.org/abs/2309.12307
-
Aurelian: 70B 32K story-writing (and more) [Alpha]
Finally, LongLORA is a method to reduce the number of computations over a large context, and also specifically train the embed and norm layers fully, that is, no quantization or LORA for those. They are small layers and easy to train without too much VRAM cost, but the LongLORA authors noticed they have a big impact on long context performance. I am not using their computation reduction methods, but I am using their suggestion to train embed/norm layers fully.
-
Why train on Yi 4K instead of 200K?
That used to be true, but things like LongLORA and LongQLoRA demonstrate that you can increase the context length of a foundation model.
-
Using Overfitting to Debug My LLM [P]
For reference, I am using the LongLoRA SFT implementation for fine-tuning a CodeLLaMA model on a code generation instruction. I have also attached my evaluation code below:
stanford_alpaca
-
How Open is Generative AI? Part 2
Alpaca is an instruction-oriented LLM derived from LLaMA, enhanced by Stanford researchers with a dataset of 52,000 examples of following instructions, sourced from OpenAI’s InstructGPT through the self-instruct method. The extensive self-instruct dataset, details of data generation, and the model refinement code were publicly disclosed. This model complies with the licensing requirements of its base model. Due to the utilization of InstructGPT for data generation, it also adheres to OpenAI’s usage terms, which prohibit the creation of models competing with OpenAI. This illustrates how dataset restrictions can indirectly affect the resulting fine-tuned model.
- Ask HN: AI/ML papers to catch up with current state of AI?
- OpenAI board in discussions with Sam Altman to return as CEO
- Are there any AI like ChatGPT without content restrictions?
-
Fine-tuning LLMs with LoRA: A Gentle Introduction
In this article, we're going to experiment with LoRA and fine-tune Llama Alpaca using commercial hardware.
-
Creating a new Finetuned model
Most papers I did read showed at least a thousand, even 10000 at several cases, so I assumed that to be the trend in the case of Low rank adapter(PEFT) training.(source: [2305.14314] QLoRA: Efficient Finetuning of Quantized LLMs (arxiv.org) , Stanford CRFM (Alpaca) and the minimum being openchat/openchat · Hugging Face ; There are a lot more examples)
-
Shock tick up for wage growth to 7.3% in blow for Bank of England
I'm not talking about OpenAI ChatGPT I'm talking about things ALPACA, and where did they train these models? Off the existing models for a fraction of a fraction of a fraction of the cost: https://crfm.stanford.edu/2023/03/13/alpaca.html
- Bye bye Bing
-
The idea maze for AI startups (2015)
I think there's a new approach for “How do you get the data?” that wasn't available when this article was written in 2015. The new text and image generative models can now be used to synthesize training datasets.
I was working on an typing autocorrect project and needed a corpus of "text messages". Most of the traditional NLP corpuses like those available through NLTK [0] aren't suitable. But it was easy to script ChatGPT to generate thousands of believable text messages by throwing random topics at it.
Similarly, you can synthesize a training dataset by giving GPT the outputs/labels and asking it to generate a variety of inputs. For sentiment analysis... "Give me 1000 negative movie reviews" and "Now give me 1000 positive movie reviews".
The Alpaca folks used GPT-3 to generate high-quality instruction-following datasets [1] based on a small set of human samples.
Etc.
[0] https://www.nltk.org/nltk_data/
[1] https://crfm.stanford.edu/2023/03/13/alpaca.html
-
Repos and tutorials for a full finetune (not LoRA)
AFAIK, the original alpaca repo was a full finetune. https://github.com/tatsu-lab/stanford_alpaca
What are some alternatives?
relora - Official code for ReLoRA from the paper Stack More Layers Differently: High-Rank Training Through Low-Rank Updates
alpaca-lora - Instruct-tune LLaMA on consumer hardware
Zicklein - Finetuning instruct-LLaMA on german datasets.
ChatGLM-6B - ChatGLM-6B: An Open Bilingual Dialogue Language Model | 开源双语对话语言模型
torch-adapters - Small Library of PyTorch Adaptation modules
Open-Assistant - OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.
discus - A data-centric AI package for ML/AI. Get the best high-quality data for the best results. Discord: https://discord.gg/t6ADqBKrdZ
llama.cpp - LLM inference in C/C++
punica - Serving multiple LoRA finetuned LLM as one
GPTQ-for-LLaMa - 4 bits quantization of LLaMA using GPTQ
RingAttention - Transformers with Arbitrarily Large Context
text-generation-webui - A Gradio web UI for Large Language Models. Supports transformers, GPTQ, AWQ, EXL2, llama.cpp (GGUF), Llama models.