diff --git a/research/nlp/senta/src/config.py b/research/nlp/senta/src/config.py
index f8bc44d092506f15de17c61ff496fdb2fd5d005a..743485ad13020f32bd35dc8ed231ece4b6232884 100644
--- a/research/nlp/senta/src/config.py
+++ b/research/nlp/senta/src/config.py
@@ -72,12 +72,12 @@ sstcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -90,7 +90,7 @@ sstcfg = {
}
],
"config": {
- "data_path": "../data/en/finetune/SST-2/train",
+ "data_path": "data/en/finetune/SST-2/train",
"shuffle": True,
"batch_size": 1,
"epoch": 10,
@@ -122,12 +122,12 @@ sstcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -153,7 +153,7 @@ sstcfg = {
}
],
"config": {
- "data_path": "../data/en/finetune/SST-2/test",
+ "data_path": "data/en/finetune/SST-2/test",
"shuffle": False,
"batch_size": 1,
"epoch": 1,
@@ -197,12 +197,12 @@ sstcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -216,7 +216,7 @@ sstcfg = {
}
],
"config": {
- "data_path": "../data/en/finetune/SST-2/dev",
+ "data_path": "data/en/finetune/SST-2/dev",
"shuffle": False,
"batch_size": 1,
"epoch": 1,
@@ -267,12 +267,12 @@ semcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -292,12 +292,12 @@ semcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -322,15 +322,15 @@ semcfg = {
}
],
"config": {
- "data_path": "../data/en/finetune/absa_laptops/train",
+ "data_path": "data/en/finetune/absa_laptops/train",
"shuffle": True,
"batch_size": 1,
"epoch": 10,
"sampling_rate": 1.0,
"extra_params": {
- "vocab_path": "../roberta_en.vocab.txt",
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json",
+ "vocab_path": "roberta_en.vocab.txt",
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json",
"label_map_config": "",
"max_seq_len": 512,
"do_lower_case": True,
@@ -369,12 +369,12 @@ semcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -395,12 +395,12 @@ semcfg = {
"split_char": " ",
"unk_token": "[UNK]",
"params": {
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json"
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json"
}
},
"need_convert": True,
- "vocab_path": "../roberta_en.vocab.txt",
+ "vocab_path": "roberta_en.vocab.txt",
"max_seq_len": 512,
"truncation_type": 0,
"padding_id": 1,
@@ -426,15 +426,15 @@ semcfg = {
}
],
"config": {
- "data_path": "../data/en/finetune/absa_laptops/test",
+ "data_path": "data/en/finetune/absa_laptops/test",
"shuffle": False,
"batch_size": 1,
"epoch": 1,
"sampling_rate": 1.0,
"extra_params": {
- "vocab_path": "../roberta_en.vocab.txt",
- "bpe_vocab_file": "../roberta_en.vocab.bpe",
- "bpe_json_file": "../roberta_en.encoder.json",
+ "vocab_path": "roberta_en.vocab.txt",
+ "bpe_vocab_file": "roberta_en.vocab.bpe",
+ "bpe_json_file": "roberta_en.encoder.json",
"label_map_config": "",
"max_seq_len": 512,
"do_lower_case": True,
diff --git a/research/nlp/senta/src/make_dataset.py b/research/nlp/senta/src/make_dataset.py
index b9b7f9821b0787ced3e5f8a7f7b004675df4b3fb..b9c0826cb4afe8829f8f646ed49bc4f041579f5d 100644
--- a/research/nlp/senta/src/make_dataset.py
+++ b/research/nlp/senta/src/make_dataset.py
@@ -109,13 +109,13 @@ if __name__ == "__main__":
train_wrapper = dataset_reader.train_reader.data_generator()
make_dataset(
wrapper=train_wrapper,
- output_path='../data/',
+ output_path='data/',
task_name=args.job,
mode="train")
dev_wrapper = dataset_reader.dev_reader.data_generator()
make_dataset(
wrapper=dev_wrapper,
- output_path='../data/',
+ output_path='data/',
task_name=args.job,
mode="dev")
if args.job == "Sem-L":
@@ -127,12 +127,12 @@ if __name__ == "__main__":
train_wrapper = dataset_reader.train_reader.data_generator()
make_dataset(
wrapper=train_wrapper,
- output_path='../data/',
+ output_path='data/',
task_name=args.job,
mode="train")
dev_wrapper = dataset_reader.test_reader.data_generator()
make_dataset(
wrapper=dev_wrapper,
- output_path='../data/',
+ output_path='data/',
task_name=args.job,
mode="dev")
diff --git a/research/nlp/senta/src/utils/params.py b/research/nlp/senta/src/utils/params.py
index 2cea4a45f852844f5e483e739f1b1af8cf4d7faa..1d087e794b5e7b38be1076dfdbabaced639415e4 100644
--- a/research/nlp/senta/src/utils/params.py
+++ b/research/nlp/senta/src/utils/params.py
@@ -25,7 +25,7 @@ def replace_none(params):
try:
value = chr(int(value, base=16))
print("ord(value): ", ord(value))
- except IOError:
+ except ValueError:
pass
params[key] = value
return params