FrameSemanticParser

This is the source code for the work done for my Master's thesis. This work is built upon the ELECTRA+AGE architecture propsed in 2022 and the code for the original architecture is at link.

I have expaned the code to work with different language models for vecotrisation and the code currently provides the option to train the model with

• deberta
• GTE
• T5

---

Requirements

Please make sure your environment includes:

python (tested on 3.10)
pytorch (tested on 2.0.1)

Then, run the command:

pip install -r requirements.txt

---

vecotrisation

---

Run

Run the parser using CL: