Supernova Roman Restoration V4

High-performance deterministic Romanized Nepali restoration using the Supernova V2 + V3.1 systems with a Cython-optimized longest-match trie.

V4 is designed around one principle:

Use deterministic linguistic evidence instead of a large pretrained generative model when the task can be solved reliably with explicit mappings and rules.


πŸš€ Overview

Supernova Roman Restoration V4 combines:

                    INPUT
                      β”‚
                      β–Ό
             Script / Language Route
                      β”‚
          β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
          β”‚                       β”‚
          β–Ό                       β–Ό
   Romanized Nepali          General Text
          β”‚                       β”‚
          β–Ό                       β–Ό
       V3.1                      V2
          β”‚                       β”‚
          β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
                      β–Ό
              Cython Longest-
                 Match Trie
                      β”‚
                      β–Ό
                   OUTPUT
Downloads last month
16
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support