I think it is awesome! Especially the fact that it's going to be a superset of Python is really highly appreciated. But when it comes to AI workflows, I think we'll still have to see what the speed advantage is since they currently don't support GPUs and CUDA afaik.
1/2
@rasbt
-
Python Superset Language Promise Tempered by GPU Support Concerns
By
–
-
GPT-4 Recent Changes Show Negative Performance Trend
By
–
Actually if you consider the recent changes to GPT-4, closed-source now has a negative slope 😛
-
Checkpoint Support Development in Progress for Lit-GPT
By
–
Checkpoint support is in the works: https://
github.com/Lightning-AI/l
it-gpt/issues/464
… -
Benchmarks comparison with GPT-4 and ChatGPT-3.5
By
–
The benchmarks include GPT-4 and ChatGPT-3.5 (table on the right)
-
Fine-tuning with 16k tokens but capable of handling 100k
By
–
When I understood correctly, they used 16k for finetuning, but it can handle up to 100k even?
-
Why no 70B model for 100k context windows
By
–
Also why no 70B model?
1) Because of the 4x smaller dataset (compared to Llama 2 pretraining)?
2) to make 100k context windows work hardware-wise? -
Gen AI in Moderation: Balance Over Excessive Use
By
–
I think gen AI is fine if used in moderation, the same way CGI is fine in moderation. The problem is when it gets excessively used.
-
CodeLlama-Python 34B finetuned on unnatural instructions
By
–
It's CodeLlama-Python 34B finetuned on 15,000 unnatural instructions. What's an unnatural instruction? Would like to see a few examples 😛
-
CodeLlama 34B Significantly Improves Coding Performance Over Llama 2
By
–
Llama 2 is awesome, however, coding tasks were not its strong suite. (See HumanEval, a coding-related evaluation task from the paper Evaluating Large Language Models Trained on Code.) The new 34B CodeLlama model is twice as good as the original 70B Llama 2 model and closes the
-
AI enthusiast skeptical about AI integration in everyday products
By
–
I really enjoy working on AI, but I don’t really want AI in every product I am using.