![]() |
Sample training small model - Printable Version +- Python Forum (https://python-forum.io) +-- Forum: Python Coding (https://python-forum.io/forum-7.html) +--- Forum: Data Science (https://python-forum.io/forum-44.html) +--- Thread: Sample training small model (/thread-39644.html) |
Sample training small model - AndrzejB - Mar-21-2023 I am just starting deep learning. I now have only laptop without dedicated GPU. I search code and dataset for deep learning model in reasonable time interval. One classic problem is handwritten letters ,but I prefer not image but other data, especially NLP. I wanna run code on my laptop and in future compare time when I buy computer with GPU. RE: Sample training small model - Larz60+ - Mar-21-2023 Depending on what exactly you plan to do (you are quite vague), a laptop may not have enough memory or processing power to be effective. The NLP corpora can require a large amount of memory and/or disk space, But you might be ok with a smaller project. Please elaborate on exactly what your goal is, what you have for a laptop, memory, disk space, etc. you might want to take a look at the NLTK package here RE: Sample training small model - AndrzejB - Mar-21-2023 Laptop has 1 TB whole disk (about 700 GB free), 20 GiB RAM, processor I3 8th generation, main problem is GPU build with CPU. If no NLP, is any small maybe toy problem? RE: Sample training small model - jefsummers - Mar-22-2023 i3 may be your bottleneck. Consider using Google Colab as a stopgap. You can use GPU or TPU hardware acceleration on their dime. And they have a lot of dimes. Did an IBM course through Coursera where one assignment kept crashing on the IBM platform (you got a Watson account with the course). Moved the code to Colab and it ran just fine. |