init

.gitattributes

@@ -515,3 +515,46 @@ data/tweet_sentiment_small/test_1.jsonl filter=lfs diff=lfs merge=lfs -text
 data/tweet_sentiment_small/test_2.jsonl filter=lfs diff=lfs merge=lfs -text
 data/tweet_sentiment_small/test_3.jsonl filter=lfs diff=lfs merge=lfs -text
 data/tweet_sentiment_small/test_4.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed0/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed0/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed2/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed1/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed2/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/test_1.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed1/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed0/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed2/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed2/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed0/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed2/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed0/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed0/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed1/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed2/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed0/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed2/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/test_4.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed0/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed2/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed0/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed1/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed0/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed1/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/test_2.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed1/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test0_seed1/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed1/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed0/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed1/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed2/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/test_3.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed1/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed0/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed1/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed2/train.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test1_seed2/validation.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test2_seed2/test.jsonl filter=lfs diff=lfs merge=lfs -text
+data/tweet_hate_test3_seed1/train.jsonl filter=lfs diff=lfs merge=lfs -text

data/tweet_hate/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43b9d4a9db852d8da4e299981096a83557cbbe28dde9184e781e89ea86008fb7
+size 1068405

data/tweet_hate/test_1.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed6597a73c19371997a316d75cd864944c3c407cca334e7d791d8392510cf7
+size 267275

data/tweet_hate/test_2.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fc150221e8da1c8375a4c02040a0a116847df2784e435927406236775e04d8a
+size 269030

data/tweet_hate/test_3.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3222984613f3776a1b7238551b150623f1949e6ecf427810d0a0c452f0380b4
+size 263958

data/tweet_hate/test_4.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b7a3ef3b33df6cf9f852548cd1e5f17feb2f06583cbcebe2e4393065feed1a
+size 268139

data/tweet_hate/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c18e35554c1a640ac1e8a9a715288153146267d7f6cd8f5b73c6e9ddaa7ed7e
+size 815849

data/tweet_hate/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3834b8a9257daf7a0c3452ca30e9743727d227cd073497a9effaa21941730b36
+size 203960

data/tweet_hate_test0_seed0/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed6597a73c19371997a316d75cd864944c3c407cca334e7d791d8392510cf7
+size 267275

data/tweet_hate_test0_seed0/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:174544a9026d4119a2e7022617eb9c5d5e2ca0960e49ea4f211015a7a547f30d
+size 840880

data/tweet_hate_test0_seed0/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a399256c2a55aea7f884b704ad8a7aab2720318f6e536f717d6e117466fa0538
+size 210897

data/tweet_hate_test0_seed1/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed6597a73c19371997a316d75cd864944c3c407cca334e7d791d8392510cf7
+size 267275

data/tweet_hate_test0_seed1/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2484567154aa9abbe66e77f936af53879b71136b2a4dcacd8343e77aeb3b8061
+size 840660

data/tweet_hate_test0_seed1/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f9bfaa47f0d1e4fab93a2af581257e2e9df56347f67315981288ce77172b9a0
+size 209114

data/tweet_hate_test0_seed2/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed6597a73c19371997a316d75cd864944c3c407cca334e7d791d8392510cf7
+size 267275

data/tweet_hate_test0_seed2/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd7f623f6204d08e70cfa4ada9d15aa53448c22ab0999f2dc661b9772d5158b1
+size 841027

data/tweet_hate_test0_seed2/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:351806667abfb3801f417e8b2f932d137302853fa4146fa2df59388782a82c74
+size 211196

data/tweet_hate_test1_seed0/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fc150221e8da1c8375a4c02040a0a116847df2784e435927406236775e04d8a
+size 269030

data/tweet_hate_test1_seed0/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6017139e0b986dd0ef2457dc0bdfa6f4ab1f22b3969fa65170331d97e9c5539
+size 839166

data/tweet_hate_test1_seed0/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f00239f7d2efb1b426cde73b151ef81b4b6c0275275a3e85d19c817d5395259
+size 209502

data/tweet_hate_test1_seed1/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fc150221e8da1c8375a4c02040a0a116847df2784e435927406236775e04d8a
+size 269030

data/tweet_hate_test1_seed1/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac2f0fa53e488333c4bac8aecbf61ffea86159075a109d920d52129e6ac70bea
+size 838927

data/tweet_hate_test1_seed1/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f455382ee8ba826d0ba941a3a25836f6694d1ec194f07962b17e1727c1dde228
+size 209959

data/tweet_hate_test1_seed2/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fc150221e8da1c8375a4c02040a0a116847df2784e435927406236775e04d8a
+size 269030

data/tweet_hate_test1_seed2/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1d5c7efa5db8ab2ecd6131c6958f4f0a41cdb5806808d45de34dc753c697723
+size 841009

data/tweet_hate_test1_seed2/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43ad0ede79283aae0c0c58116e45503618ac11a475e39729e9d19c88aee243f0
+size 210602

data/tweet_hate_test2_seed0/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3222984613f3776a1b7238551b150623f1949e6ecf427810d0a0c452f0380b4
+size 263958

data/tweet_hate_test2_seed0/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c3413462ad64727c505eb8cb6572671aa5b05ea2140a0e227c5d564a22d4bbe
+size 844785

data/tweet_hate_test2_seed0/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aed7f2e55776d1e59a870387a868575168016683c7b8c1d3cc1669786b822c2
+size 211090

data/tweet_hate_test2_seed1/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3222984613f3776a1b7238551b150623f1949e6ecf427810d0a0c452f0380b4
+size 263958

data/tweet_hate_test2_seed1/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1761a32da30f2a6cfa17995bcbc5f5f131fb94ae85b020d455c31236825c1610
+size 838534

data/tweet_hate_test2_seed1/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df48ca1357eacb231bcc2e3bb77509e3b41c7f7e513eb4c301b50728f40b068b
+size 212723

data/tweet_hate_test2_seed2/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3222984613f3776a1b7238551b150623f1949e6ecf427810d0a0c452f0380b4
+size 263958

data/tweet_hate_test2_seed2/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a2f2207f3bcb451d067743f16866693524bfc2bb16bdbaab813c5ce106f3b40
+size 841369

data/tweet_hate_test2_seed2/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a93b57db080ef7cb71be8606b07f5cf9b5431f16b5a0ae52d4387af13b330c2c
+size 211008

data/tweet_hate_test3_seed0/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b7a3ef3b33df6cf9f852548cd1e5f17feb2f06583cbcebe2e4393065feed1a
+size 268139

data/tweet_hate_test3_seed0/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:216a53acac01a66030236282738da43adb48ac26935eee415fd3a1e678936316
+size 841157

data/tweet_hate_test3_seed0/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bc31bfcbcc0f0f3d6212c98a22ff142909ee58fe736b63eda1e08c72872a15d
+size 213049

data/tweet_hate_test3_seed1/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b7a3ef3b33df6cf9f852548cd1e5f17feb2f06583cbcebe2e4393065feed1a
+size 268139

data/tweet_hate_test3_seed1/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c47d6e191f32f757bcc6f58ba89e1ff9adb3d24eeaf192a596c5c86e5f1f2dd3
+size 840179

data/tweet_hate_test3_seed1/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:864db3e45cd644e2f552488dc06a76ad973533ba04eddaf9cb5d87cab1ab5c07
+size 210892

data/tweet_hate_test3_seed2/test.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b7a3ef3b33df6cf9f852548cd1e5f17feb2f06583cbcebe2e4393065feed1a
+size 268139

data/tweet_hate_test3_seed2/train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd43253ba78ce975eceb84a130b13b39d6192794257941a3b90f78741ba680d5
+size 842557

data/tweet_hate_test3_seed2/validation.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78fddd1e3e29b93c91889b8f6e5b48077e593ca9c6bf1ff2c267c9215f7f6369
+size 211761

experiments/analysis_prediction_topic.py

@@ -0,0 +1,79 @@
+import os
+import json
+
+import numpy as np
+import pandas as pd
+from datasets import load_dataset
+
+os.makedirs("experiments/analysis_topic", exist_ok=True)
+root_dir = "experiments/prediction_files"
+id_to_label = {
+    '0': 'arts_&_culture',
+    '1': 'business_&_entrepreneurs',
+    '2': 'celebrity_&_pop_culture',
+    '3': 'diaries_&_daily_life',
+    '4': 'family',
+    '5': 'fashion_&_style',
+    '6': 'film_tv_&_video',
+    '7': 'fitness_&_health',
+    '8': 'food_&_dining',
+    '9': 'gaming',
+    '10': 'learning_&_educational',
+    '11': 'music',
+    '12': 'news_&_social_concern',
+    '13': 'other_hobbies',
+    '14': 'relationships',
+    '15': 'science_&_technology',
+    '16': 'sports',
+    '17': 'travel_&_adventure',
+    '18': 'youth_&_student_life'
+}
+splits = ["test_1", "test_2", "test_3", "test_4"]
+model_list = [
+    "roberta-base",
+    "bertweet-base",
+    "bernice",
+    "roberta-large",
+    "bertweet-large",
+    "twitter-roberta-base-2019-90m",
+    "twitter-roberta-base-dec2020",
+    "twitter-roberta-base-2021-124m",
+    "twitter-roberta-base-2022-154m",
+    "twitter-roberta-large-2022-154m"
+]
+references = {}
+for s in splits:
+    data = load_dataset("tweettemposhift/tweet_temporal_shift", f"topic_temporal", split=s)
+    references[s] = [{id_to_label[str(n)] for n, k in enumerate(i) if k == 1} for i in data['gold_label_list']]
+
+count = {}
+pred_tmp = {}
+for model_m in model_list:
+    flags = []
+    pred_all = []
+    for s in splits:
+        with open(f"{root_dir}/topic-topic_temporal-{model_m}/{s}.jsonl") as f:
+            pred = [set(json.loads(i)["label"]) for i in f.read().split('\n') if len(i)]
+            flags += [len(a.intersection(b)) > 0 for a, b in zip(references[s], pred)]
+            pred_all += pred
+    for seed_s in range(3):
+        flags_rand = []
+        for random_r in range(4):
+            with open(f"{root_dir}/topic-topic_random{random_r}_seed{seed_s}-{model_m}/test_{random_r + 1}.jsonl") as f:
+                pred = [set(json.loads(i)["label"]) for i in f.read().split('\n') if len(i)]
+                label = references[f"test_{random_r + 1}"]
+                flags_rand += [len(a.intersection(b)) > 0 for a, b in zip(label, pred)]
+        tmp_flag = [not x and y for x, y in zip(flags, flags_rand)]
+        count[f"{model_m}_{seed_s}"] = tmp_flag
+        pred_tmp[f"{model_m}_{seed_s}"] = [list(x) if y else [] for x, y in zip(pred_all, tmp_flag)]
+
+df_tmp = pd.DataFrame([[dict(zip(*np.unique(i, return_counts=True))) for i in pd.DataFrame(pred_tmp).sum(1).values]], index=["errors"]).T
+df_tmp["error_count"] = pd.DataFrame(count).sum(1).values
+gold_label = []
+text = []
+for s in splits:
+    gold_label += load_dataset("tweettemposhift/tweet_temporal_shift", "topic_temporal", split=s)['gold_label_list']
+    text += load_dataset("tweettemposhift/tweet_temporal_shift", "topic_temporal", split=s)['text']
+df_tmp["true_label"] = [", ".join([id_to_label[str(n)] for n, k in enumerate(i) if k == 1]) for i in gold_label]
+df_tmp["text"] = text
+df_tmp.sort_values("error_count", ascending=False).to_csv("experiments/analysis/topic_2.csv")

process/tweet_emoji.py

@@ -0,0 +1,81 @@
+# TODO
+import json
+import os
+from random import shuffle, seed
+
+
+with open("data/tweet_sentiment/test.jsonl") as f:
+    test = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/test_1.jsonl") as f:
+    test_1 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/test_2.jsonl") as f:
+    test_2 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/test_3.jsonl") as f:
+    test_3 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/test_4.jsonl") as f:
+    test_4 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/train.jsonl") as f:
+    train = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_sentiment/validation.jsonl") as f:
+    validation = [json.loads(i) for i in f if len(i)]
+
+
+os.makedirs(f"data/tweet_sentiment_small", exist_ok=True)
+with open(f"data/tweet_sentiment_small/test.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in test]))
+
+with open(f"data/tweet_sentiment_small/test_1.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in test_1]))
+with open(f"data/tweet_sentiment_small/test_2.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in test_2]))
+with open(f"data/tweet_sentiment_small/test_3.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in test_3]))
+with open(f"data/tweet_sentiment_small/test_4.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in test_4]))
+
+with open(f"data/tweet_sentiment_small/validation.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in validation]))
+
+# down sample training set
+n_train_p = 2500
+n_train_n = 2500
+seed(123)
+shuffle(train)
+train_p = [i for i in train if i["gold_label_binary"] == 0][:n_train_p]
+train_n = [i for i in train if i["gold_label_binary"] == 1][:n_train_n]
+train = train_p + train_n
+shuffle(train)
+with open(f"data/tweet_sentiment_small/train.jsonl", "w") as f:
+    f.write("\n".join([json.dumps(i) for i in train]))
+
+
+n_train = len(train)
+n_validation = len(validation)
+n_test = int(len(test)/4)
+
+def sampler(dataset_test, r_seed):
+    seed(r_seed)
+    shuffle(dataset_test)
+    shuffle(train)
+    shuffle(validation)
+    test_tr = dataset_test[:int(n_train / 2)]
+    test_vl = dataset_test[int(n_train / 2): int(n_train / 2) + int(n_validation / 2)]
+    new_train = test_tr + train[:n_train - len(test_tr)]
+    new_validation = test_vl + validation[:n_validation - len(test_vl)]
+    return new_train, new_validation
+
+id2test = {n: t for n, t in enumerate([test_1, test_2, test_3, test_4])}
+for n, _test in enumerate([
+        test_4 + test_2 + test_3,
+        test_1 + test_4 + test_3,
+        test_1 + test_2 + test_4,
+        test_1 + test_2 + test_3]):
+    for s in range(3):
+        os.makedirs(f"data/tweet_sentiment_small_test{n}_seed{s}", exist_ok=True)
+        _train, _valid = sampler(_test, s)
+        with open(f"data/tweet_sentiment_small_test{n}_seed{s}/train.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in _train]))
+        with open(f"data/tweet_sentiment_small_test{n}_seed{s}/validation.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in _valid]))
+        with open(f"data/tweet_sentiment_small_test{n}_seed{s}/test.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in id2test[n]]))

process/tweet_hate.py

@@ -0,0 +1,51 @@
+import json
+import os
+from random import shuffle, seed
+
+
+with open("data/tweet_hate/test.jsonl") as f:
+    test = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/test_1.jsonl") as f:
+    test_1 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/test_2.jsonl") as f:
+    test_2 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/test_3.jsonl") as f:
+    test_3 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/test_4.jsonl") as f:
+    test_4 = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/train.jsonl") as f:
+    train = [json.loads(i) for i in f if len(i)]
+with open("data/tweet_hate/validation.jsonl") as f:
+    validation = [json.loads(i) for i in f if len(i)]
+
+n_train = len(train)
+n_validation = len(validation)
+n_test = int(len(test)/4)
+
+
+def sampler(dataset_test, r_seed):
+    seed(r_seed)
+    shuffle(dataset_test)
+    shuffle(train)
+    shuffle(validation)
+    test_tr = dataset_test[:int(n_train / 2)]
+    test_vl = dataset_test[int(n_train / 2): int(n_train / 2) + int(n_validation / 2)]
+    new_train = test_tr + train[:n_train - len(test_tr)]
+    new_validation = test_vl + validation[:n_validation - len(test_vl)]
+    return new_train, new_validation
+
+id2test = {n: t for n, t in enumerate([test_1, test_2, test_3, test_4])}
+for n, _test in enumerate([
+        test_4 + test_2 + test_3,
+        test_1 + test_4 + test_3,
+        test_1 + test_2 + test_4,
+        test_1 + test_2 + test_3]):
+    for s in range(3):
+        os.makedirs(f"data/tweet_hate_test{n}_seed{s}", exist_ok=True)
+        _train, _valid = sampler(_test, s)
+        with open(f"data/tweet_hate_test{n}_seed{s}/train.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in _train]))
+        with open(f"data/tweet_hate_test{n}_seed{s}/validation.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in _valid]))
+        with open(f"data/tweet_hate_test{n}_seed{s}/test.jsonl", "w") as f:
+            f.write("\n".join([json.dumps(i) for i in id2test[n]]))

process/tweet_nerd.py

@@ -48,20 +48,20 @@ test_1 = test[:n_test]
 test_2 = test[n_test:n_test*2]
 test_3 = test[n_test*2:n_test*3]
 test_4 = test[n_test*3:]
-os.makedirs("data/tweet_nerd_new", exist_ok=True)
-with open("data/tweet_nerd_new/test.jsonl", "w") as f:
+os.makedirs("data/tweet_hate", exist_ok=True)
+with open("data/tweet_hate/test.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in test]))
-with open("data/tweet_nerd_new/test_1.jsonl", "w") as f:
+with open("data/tweet_hate/test_1.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in test_1]))
-with open("data/tweet_nerd_new/test_2.jsonl", "w") as f:
+with open("data/tweet_hate/test_2.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in test_2]))
-with open("data/tweet_nerd_new/test_3.jsonl", "w") as f:
+with open("data/tweet_hate/test_3.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in test_3]))
-with open("data/tweet_nerd_new/test_4.jsonl", "w") as f:
+with open("data/tweet_hate/test_4.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in test_4]))
-with open("data/tweet_nerd_new/train.jsonl", "w") as f:
+with open("data/tweet_hate/train.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in train]))
-with open("data/tweet_nerd_new/validation.jsonl", "w") as f:
+with open("data/tweet_hate/validation.jsonl", "w") as f:
     f.write("\n".join([json.dumps(i) for i in valid]))
 
 

process/tweet_sentiment_small.py

@@ -1,4 +1,3 @@
-# TODO
 import json
 import os
 from random import shuffle, seed