diff options
Diffstat (limited to 'training/conf/model')
| -rw-r--r-- | training/conf/model/lit_transformer.yaml | 8 | ||||
| -rw-r--r-- | training/conf/model/mapping/word_piece.yaml | 9 |
2 files changed, 16 insertions, 1 deletions
diff --git a/training/conf/model/lit_transformer.yaml b/training/conf/model/lit_transformer.yaml index 4e04b85..5341d8e 100644 --- a/training/conf/model/lit_transformer.yaml +++ b/training/conf/model/lit_transformer.yaml @@ -1,4 +1,10 @@ +defaults: + - mapping: word_piece + _target_: text_recognizer.models.transformer.TransformerLitModel interval: null monitor: val/loss -ignore_tokens: ["<s>", "<e>", "<p>"] +ignore_tokens: [ <s>, <e>, <p> ] +start_token: <s> +end_token: <e> +pad_token: <p> diff --git a/training/conf/model/mapping/word_piece.yaml b/training/conf/model/mapping/word_piece.yaml new file mode 100644 index 0000000..39e2ba4 --- /dev/null +++ b/training/conf/model/mapping/word_piece.yaml @@ -0,0 +1,9 @@ +_target_: text_recognizer.data.mappings.WordPieceMapping +num_features: 1000 +tokens: iamdb_1kwp_tokens_1000.txt +lexicon: iamdb_1kwp_lex_1000.txt +data_dir: null +use_words: false +prepend_wordsep: false +special_tokens: ["<s>", "<e>", "<p>"] +extra_symbols: ["\n"] |