Відгуки: Text classification using reusable embeddings

Eshal F. · Відгук надано близько 17 годин тому

552_YASH SATISH S. · Відгук надано близько 20 годин тому

PRANAV VINAYAK T. · Відгук надано близько 21 години тому

759_Rishima G. · Відгук надано близько 23 годин тому

Varada V. · Відгук надано близько 23 годин тому

Asad A. · Відгук надано близько 24 годин тому

This lab has significant environment and execution issues that prevent successful completion even when instructions are followed correctly. 1. **Broken Python Environment** Multiple students (including myself) encountered missing core dependencies such as `pkg_resources` and `tensorflow.keras.utils`. These are fundamental packages and should not be missing in a pre-configured lab environment. This indicates improper environment setup. 2. **TensorFlow Hub Integration Failure** The lab requires `tensorflow_hub.KerasLayer`, but the environment frequently fails to import it due to dependency issues. This blocks all subsequent tasks (NNLM and Swivel models), making the lab impossible to complete. 3. **Lack of Dependency Setup Instructions** The notebook assumes all packages are pre-installed and functional, but in reality: * `setuptools` may be missing * `tensorflow-hub` may be broken or incompatible There are no troubleshooting or fallback instructions provided. 4. **Execution Order Sensitivity Without Guidance** The notebook heavily depends on strict cell execution order, but does not explicitly warn users. Any kernel restart or skipped cell leads to cascading `NameError` issues, confusing learners. 5. **No Debugging Support** There is no guidance for handling common failures such as: * import errors * environment inconsistencies * TF-Hub download failures --- ### Suggested Improvements * Provide a verified working environment or requirements check at the start * Add a setup cell to install/validate dependencies (`tensorflow`, `tensorflow-hub`, `setuptools`) * Include explicit instructions to restart kernel and run all cells in order * Add troubleshooting section for common errors * Validate TF-Hub imports before proceeding to model building --- Overall, the lab content is valuable, but the current execution environment makes it unreliable and frustrating for students. Addressing these issues would significantly improve the learning experience.

521_SAURABH H. · Відгук надано 1 день тому

Devesh B. · Відгук надано 1 день тому

521_SAURABH H. · Відгук надано 1 день тому

VIJAYRAGHAVAN U. · Відгук надано 1 день тому

worst lab ever

SNEHA Y. · Відгук надано 2 дні тому

931_Keerthana S. · Відгук надано 2 дні тому

shalini m. · Відгук надано 2 дні тому

759_Rishima G. · Відгук надано 2 дні тому

759_Rishima G. · Відгук надано 2 дні тому

Soham M. · Відгук надано 2 дні тому

Jaswant R. · Відгук надано 2 дні тому

Absolutely pathetic. No dependency is being resolved I had to install packages on my own. No import is being resolved

Arnav D. · Відгук надано 2 дні тому

nice

Durga Aditya P. · Відгук надано 2 дні тому

nice

Sujana M. · Відгук надано 2 дні тому

I completed the assessment but when I clicked check my progress its not updating

lakshay a. · Відгук надано 3 дні тому

MANAL U. · Відгук надано 3 дні тому

Segundo Hadson P. · Відгук надано 3 дні тому

Aswin K. · Відгук надано 3 дні тому

619_Aditya G. · Відгук надано 3 дні тому