docs: update readme

README.md

@@ -1,31 +1,37 @@
 
 # Tensorli
 
-Absolute minimalistic implementation of a GPT-like transformer using only numpy.
+Absolute minimalistic implementation of a GPT-like transformer using only `numpy`.
 
 The implementation includes:
 - Automatic differentiation
 - `Tensorli` object (PyTorch like)
 - Simple NN layers: `Linearli`, `Embeddingli`, `MultiheadAttentionli`, `LayerNorm`
 - Optimizers: `Adamli`
 
-everything that is needed to train and execute a GPT-like transformer model.
+All that is "needed" to train and execute a GPT-like transformer model.
 
 ```
-Everything else are optimizations.
+Everything else is optimizations.
 ```
 
 ## Naming
 
-In the region where I grew up we have a word for "little" that is used as a suffix. For example, "little dog" would be "dogli". I thought it would be a nice name for a minimalistic implementation of a neural network library.
+In the region where I grew up, a word for "little" is used as a suffix \[[1](https://de.wikipedia.org/wiki/-li)\]. For example, "little dog" would be "dogli". I thought it would be a nice name for a minimalistic implementation of a neural network library.
 
 ### Caveats
 
-This library works but it is not optimized. It is not meant to be used in production. It is meant to be used as a learning tool.
+This library works, but it is NOT optimized. It is not meant to be used in production or for anything at scale. It is meant to be used as a learning tool.
 
+## Inspiration
+
+This library is heavily inspired by the following projects:
+- [minGPT](https://github.com/karpathy/minGPT)
+- [tinygrad](https://github.com/tinygrad/tinygrad)
+
+I highly recommend checking them out.
 
 ### Plans & Outlook
 
 - [ ] Add Dropout
 - [ ] Add more experimental architectures
-