test with a1

batch_finetune_eu_jav_small_a1.sh

@@ -0,0 +1,11 @@
+PROJECT_DIR=${HOME}"/models/eu-jav-categorisation"
+export PYTHONPATH=${PROJECT_DIR}
+INITIAL_CHECKPOINT_PATH=\"gs://t5-data/pretrained_models/t5x/mt5_small/checkpoint_1000000\"
+TRAIN_STEPS=1002000
+
+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_a1\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a1_v1\" &&
+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_a1\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a1_v2\" &&
+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_a1\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a1_v3\" &&
+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_a1\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a1_v4\" &&
+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_a1\" --gin.MODEL_DIR=\"gs://eu-jav-t5x/finetuned/italian_tweets/classify_tweets_small_a1_v5\"
+

tasks.py

@@ -63,4 +63,20 @@ seqio.TaskRegistry.add(
     metric_fns=[metrics.accuracy,my_metrics.f1_macro],
     output_features=DEFAULT_OUTPUT_FEATURES,
 )
+seqio.TaskRegistry.add(
+    "classify_tweets_a1",
+    source=seqio.TextLineDataSource(
+        split_to_filepattern=tsv_path,
+        #num_input_examples=num_nq_examples
+        ),
+    preprocessors=[
+      functools.partial(
+          t5.data.preprocessors.parse_tsv,
+          field_names=["target","annotator2","annotator3","placeholder","source","id"]),
+      categorise_preprocessor,
+      seqio.preprocessors.tokenize_and_append_eos,
+    ],
+    metric_fns=[metrics.accuracy,my_metrics.f1_macro],
+    output_features=DEFAULT_OUTPUT_FEATURES,
+)