Initial commit with adapted deliverables from Clarin: http://hdl.handle.net/20.500.12537/301
4f09c24
| { | |
| "dataset_reader": { | |
| "type": "conllu", | |
| "features": [ | |
| "token", | |
| "char" | |
| ], | |
| "lemma_indexers": { | |
| "char": { | |
| "type": "characters_const_padding", | |
| "character_tokenizer": { | |
| "end_tokens": [ | |
| "__END__" | |
| ], | |
| "start_tokens": [ | |
| "__START__" | |
| ] | |
| }, | |
| "min_padding_length": 32, | |
| "namespace": "lemma_characters" | |
| } | |
| }, | |
| "targets": [ | |
| "deprel", | |
| "feats", | |
| "head", | |
| "lemma", | |
| "upostag", | |
| "xpostag" | |
| ], | |
| "token_indexers": { | |
| "char": { | |
| "type": "characters_const_padding", | |
| "character_tokenizer": { | |
| "end_tokens": [ | |
| "__END__" | |
| ], | |
| "start_tokens": [ | |
| "__START__" | |
| ] | |
| }, | |
| "min_padding_length": 32 | |
| }, | |
| "feats": { | |
| "type": "feats_indexer" | |
| }, | |
| "lemma": { | |
| "type": "characters_const_padding", | |
| "character_tokenizer": { | |
| "end_tokens": [ | |
| "__END__" | |
| ], | |
| "start_tokens": [ | |
| "__START__" | |
| ] | |
| }, | |
| "min_padding_length": 32 | |
| }, | |
| "token": { | |
| "type": "pretrained_transformer_mismatched_fixed", | |
| "model_name": "transformer_models/electra-base-igc-is/", | |
| "tokenizer_kwargs": {} | |
| }, | |
| "upostag": { | |
| "type": "single_id", | |
| "feature_name": "pos_", | |
| "namespace": "upostag" | |
| }, | |
| "xpostag": { | |
| "type": "single_id", | |
| "feature_name": "tag_", | |
| "namespace": "xpostag" | |
| } | |
| }, | |
| "use_sem": false | |
| }, | |
| "model": { | |
| "type": "semantic_multitask", | |
| "dependency_relation": { | |
| "type": "combo_dependency_parsing_from_vocab", | |
| "dependency_projection_layer": { | |
| "activation": "tanh", | |
| "dropout_rate": 0.25, | |
| "in_features": 1024, | |
| "out_features": 128 | |
| }, | |
| "head_predictor": { | |
| "cycle_loss_n": 0, | |
| "dependency_projection_layer": { | |
| "activation": "tanh", | |
| "in_features": 1024, | |
| "out_features": 512 | |
| }, | |
| "head_projection_layer": { | |
| "activation": "tanh", | |
| "in_features": 1024, | |
| "out_features": 512 | |
| } | |
| }, | |
| "head_projection_layer": { | |
| "activation": "tanh", | |
| "dropout_rate": 0.25, | |
| "in_features": 1024, | |
| "out_features": 128 | |
| }, | |
| "vocab_namespace": "deprel_labels" | |
| }, | |
| "lemmatizer": { | |
| "type": "combo_lemma_predictor_from_vocab", | |
| "activations": [ | |
| "relu", | |
| "relu", | |
| "relu", | |
| "linear" | |
| ], | |
| "char_vocab_namespace": "token_characters", | |
| "dilation": [ | |
| 1, | |
| 2, | |
| 4, | |
| 1 | |
| ], | |
| "embedding_dim": 256, | |
| "filters": [ | |
| 256, | |
| 256, | |
| 256 | |
| ], | |
| "input_projection_layer": { | |
| "activation": "tanh", | |
| "dropout_rate": 0.25, | |
| "in_features": 1024, | |
| "out_features": 32 | |
| }, | |
| "kernel_size": [ | |
| 3, | |
| 3, | |
| 3, | |
| 1 | |
| ], | |
| "lemma_vocab_namespace": "lemma_characters", | |
| "padding": [ | |
| 1, | |
| 2, | |
| 4, | |
| 0 | |
| ], | |
| "stride": [ | |
| 1, | |
| 1, | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "loss_weights": { | |
| "deprel": 0.8, | |
| "feats": 0.2, | |
| "head": 0.2, | |
| "lemma": 0.05, | |
| "semrel": 0.05, | |
| "upostag": 0.05, | |
| "xpostag": 0.05 | |
| }, | |
| "morphological_feat": { | |
| "type": "combo_morpho_from_vocab", | |
| "activations": [ | |
| "tanh", | |
| "linear" | |
| ], | |
| "dropout": [ | |
| 0.25, | |
| 0 | |
| ], | |
| "hidden_dims": [ | |
| 128 | |
| ], | |
| "input_dim": 1024, | |
| "num_layers": 2, | |
| "vocab_namespace": "feats_labels" | |
| }, | |
| "regularizer": { | |
| "regexes": [ | |
| [ | |
| ".*conv1d.*", | |
| { | |
| "alpha": 1e-06, | |
| "type": "l2" | |
| } | |
| ], | |
| [ | |
| ".*forward.*", | |
| { | |
| "alpha": 1e-06, | |
| "type": "l2" | |
| } | |
| ], | |
| [ | |
| ".*backward.*", | |
| { | |
| "alpha": 1e-06, | |
| "type": "l2" | |
| } | |
| ], | |
| [ | |
| ".*char_embed.*", | |
| { | |
| "alpha": 1e-05, | |
| "type": "l2" | |
| } | |
| ] | |
| ] | |
| }, | |
| "seq_encoder": { | |
| "type": "combo_encoder", | |
| "layer_dropout_probability": 0.33, | |
| "stacked_bilstm": { | |
| "hidden_size": 512, | |
| "input_size": 164, | |
| "layer_dropout_probability": 0.33, | |
| "num_layers": 2, | |
| "recurrent_dropout_probability": 0.33 | |
| } | |
| }, | |
| "text_field_embedder": { | |
| "type": "basic", | |
| "token_embedders": { | |
| "char": { | |
| "type": "char_embeddings_from_config", | |
| "dilated_cnn_encoder": { | |
| "activations": [ | |
| "relu", | |
| "relu", | |
| "linear" | |
| ], | |
| "dilation": [ | |
| 1, | |
| 2, | |
| 4 | |
| ], | |
| "filters": [ | |
| 512, | |
| 256, | |
| 64 | |
| ], | |
| "input_dim": 64, | |
| "kernel_size": [ | |
| 3, | |
| 3, | |
| 3 | |
| ], | |
| "padding": [ | |
| 1, | |
| 2, | |
| 4 | |
| ], | |
| "stride": [ | |
| 1, | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "embedding_dim": 64 | |
| }, | |
| "token": { | |
| "type": "transformers_word_embeddings", | |
| "model_name": "transformer_models/electra-base-igc-is/", | |
| "projection_dim": 100 | |
| } | |
| } | |
| }, | |
| "upos_tagger": { | |
| "activations": [ | |
| "tanh", | |
| "linear" | |
| ], | |
| "dropout": [ | |
| 0.25, | |
| 0 | |
| ], | |
| "hidden_dims": [ | |
| 64 | |
| ], | |
| "input_dim": 1024, | |
| "num_layers": 2, | |
| "vocab_namespace": "upostag_labels" | |
| }, | |
| "xpos_tagger": { | |
| "activations": [ | |
| "tanh", | |
| "linear" | |
| ], | |
| "dropout": [ | |
| 0.25, | |
| 0 | |
| ], | |
| "hidden_dims": [ | |
| 128 | |
| ], | |
| "input_dim": 1024, | |
| "num_layers": 2, | |
| "vocab_namespace": "xpostag_labels" | |
| } | |
| }, | |
| "train_data_path": "is_combined-ud-train.conllu", | |
| "validation_data_path": "is_combined-ud-dev.conllu", | |
| "trainer": { | |
| "type": "gradient_descent_validate_n", | |
| "checkpointer": { | |
| "type": "finishing_only_checkpointer" | |
| }, | |
| "cuda_device": 0, | |
| "epoch_callbacks": [ | |
| { | |
| "type": "transfer_patience" | |
| } | |
| ], | |
| "grad_clipping": 5, | |
| "learning_rate_scheduler": { | |
| "type": "combo_scheduler" | |
| }, | |
| "num_epochs": 400, | |
| "optimizer": { | |
| "type": "adam", | |
| "betas": [ | |
| 0.9, | |
| 0.9 | |
| ], | |
| "lr": 0.002 | |
| }, | |
| "patience": 1, | |
| "validation_metric": "+EM" | |
| }, | |
| "vocabulary": { | |
| "type": "from_instances_extended", | |
| "non_padded_namespaces": [ | |
| "head_labels" | |
| ], | |
| "only_include_pretrained_words": true, | |
| "oov_token": "_", | |
| "padding_token": "__PAD__" | |
| }, | |
| "type": "default", | |
| "data_loader": { | |
| "batch_sampler": { | |
| "type": "token_count", | |
| "word_batch_size": 2500 | |
| } | |
| }, | |
| "datasets_for_vocab_creation": [ | |
| "train" | |
| ], | |
| "numpy_seed": 8787, | |
| "pytorch_seed": 8787, | |
| "random_seed": 8787 | |
| } | |