adding a2 and a3
Browse files- batch_finetune_eu_jav_small_a2.sh +11 -0
- batch_finetune_eu_jav_small_a3.sh +11 -0
- tasks.py +33 -1
batch_finetune_eu_jav_small_a2.sh
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
PROJECT_DIR=${HOME}"/models/eu-jav-categorisation"
|
2 |
+
export PYTHONPATH=${PROJECT_DIR}
|
3 |
+
INITIAL_CHECKPOINT_PATH=\"gs://t5-data/pretrained_models/t5x/mt5_small/checkpoint_1000000\"
|
4 |
+
TRAIN_STEPS=1002000
|
5 |
+
|
6 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a2\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a2_v1\" &&
|
7 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a2\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a2_v2\" &&
|
8 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a2\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a2_v3\" &&
|
9 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a2\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a2_v4\" &&
|
10 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a2\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a2_v5\"
|
11 |
+
|
batch_finetune_eu_jav_small_a3.sh
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
PROJECT_DIR=${HOME}"/models/eu-jav-categorisation"
|
2 |
+
export PYTHONPATH=${PROJECT_DIR}
|
3 |
+
INITIAL_CHECKPOINT_PATH=\"gs://t5-data/pretrained_models/t5x/mt5_small/checkpoint_1000000\"
|
4 |
+
TRAIN_STEPS=1002000
|
5 |
+
|
6 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a3\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a3_v1\" &&
|
7 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a3\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a3_v2\" &&
|
8 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a3\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a3_v3\" &&
|
9 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a3\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a3_v4\" &&
|
10 |
+
python3 ../../t5x/t5x/train.py --gin_search_paths="./" --gin.TRAIN_STEPS=${TRAIN_STEPS} --gin_file="finetune_classification_small.gin" --gin.INITIAL_CHECKPOINT_PATH=${INITIAL_CHECKPOINT_PATH} --gin.MIXTURE_OR_TASK_NAME=\"classify_tweets_a3\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a3_v5\"
|
11 |
+
|
tasks.py
CHANGED
@@ -62,7 +62,7 @@ seqio.TaskRegistry.add(
|
|
62 |
],
|
63 |
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
64 |
output_features=DEFAULT_OUTPUT_FEATURES,
|
65 |
-
|
66 |
seqio.TaskRegistry.add(
|
67 |
"classify_tweets_a1",
|
68 |
source=seqio.TextLineDataSource(
|
@@ -78,5 +78,37 @@ seqio.TaskRegistry.add(
|
|
78 |
],
|
79 |
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
80 |
output_features=DEFAULT_OUTPUT_FEATURES,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
81 |
)
|
82 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
62 |
],
|
63 |
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
64 |
output_features=DEFAULT_OUTPUT_FEATURES,
|
65 |
+
|
66 |
seqio.TaskRegistry.add(
|
67 |
"classify_tweets_a1",
|
68 |
source=seqio.TextLineDataSource(
|
|
|
78 |
],
|
79 |
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
80 |
output_features=DEFAULT_OUTPUT_FEATURES,
|
81 |
+
))
|
82 |
+
seqio.TaskRegistry.add(
|
83 |
+
"classify_tweets_a2",
|
84 |
+
source=seqio.TextLineDataSource(
|
85 |
+
split_to_filepattern=tsv_path,
|
86 |
+
#num_input_examples=num_nq_examples
|
87 |
+
),
|
88 |
+
preprocessors=[
|
89 |
+
functools.partial(
|
90 |
+
t5.data.preprocessors.parse_tsv,
|
91 |
+
field_names=["annotator1","target","annotator3","placeholder","source","id"]),
|
92 |
+
categorise_preprocessor,
|
93 |
+
seqio.preprocessors.tokenize_and_append_eos,
|
94 |
+
],
|
95 |
+
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
96 |
+
output_features=DEFAULT_OUTPUT_FEATURES,
|
97 |
)
|
98 |
|
99 |
+
seqio.TaskRegistry.add(
|
100 |
+
"classify_tweets_a3",
|
101 |
+
source=seqio.TextLineDataSource(
|
102 |
+
split_to_filepattern=tsv_path,
|
103 |
+
#num_input_examples=num_nq_examples
|
104 |
+
),
|
105 |
+
preprocessors=[
|
106 |
+
functools.partial(
|
107 |
+
t5.data.preprocessors.parse_tsv,
|
108 |
+
field_names=["annotator1","annotator2","target","placeholder","source","id"]),
|
109 |
+
categorise_preprocessor,
|
110 |
+
seqio.preprocessors.tokenize_and_append_eos,
|
111 |
+
],
|
112 |
+
metric_fns=[metrics.accuracy,my_metrics.f1_macro],
|
113 |
+
output_features=DEFAULT_OUTPUT_FEATURES,
|
114 |
+
)
|