Darayut commited on
Commit
6c79074
Β·
verified Β·
1 Parent(s): 36e7d31

Label the spec table columns

Browse files

The empty header row rendered as a blank strip.

Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -49,7 +49,7 @@ The code is on [GitHub](https://github.com/netra-ai-lab/Romanized-Khmer-to-Khmer
49
 
50
  *AkaraAlpha turning `brodae` into αž”αŸ’αžšαžŠαŸ‚.* The encoder reads upwards: each letter becomes a 32-dimensional embedding, a bidirectional GRU reads the letters in both directions, and a projection brings each state to 64 dimensions, the keys and values for attention. The decoder reads downwards. At step 4, with αž” αŸ’ រ written, its GRU output is the query for additive attention, and a softmax layer over 81 ids gives ដ with probability 0.997. The chosen character joins the input and the loop repeats until the end marker.
51
 
52
- | | |
53
  | --- | --- |
54
  | Input | Romanized word, up to 25 characters: 26 Latin letters |
55
  | Output | Khmer word, up to 24 characters: 77 Khmer characters |
 
49
 
50
  *AkaraAlpha turning `brodae` into αž”αŸ’αžšαžŠαŸ‚.* The encoder reads upwards: each letter becomes a 32-dimensional embedding, a bidirectional GRU reads the letters in both directions, and a projection brings each state to 64 dimensions, the keys and values for attention. The decoder reads downwards. At step 4, with αž” αŸ’ រ written, its GRU output is the query for additive attention, and a softmax layer over 81 ids gives ដ with probability 0.997. The chosen character joins the input and the loop repeats until the end marker.
51
 
52
+ | Part | Setting |
53
  | --- | --- |
54
  | Input | Romanized word, up to 25 characters: 26 Latin letters |
55
  | Output | Khmer word, up to 24 characters: 77 Khmer characters |