Init

README.md

@@ -1,3 +1,74 @@
 ---
 license: apache-2.0
 ---
+
+## How to use the discriminator in `transformers` on a custom dataset
+(Heavily based on: https://github.com/huggingface/notebooks/blob/master/examples/text_classification-tf.ipynb)
+
+```python
+import math
+
+import tensorflow as tf
+from datasets import Dataset, ClassLabel, Features, Value
+from transformers import TFAutoModelForSequenceClassification, AutoTokenizer, DataCollatorWithPadding, create_optimizer
+
+# This example shows how this model can be used:
+#  you should finetune the model of your specific corpus if commands, bogger than this
+dict_train = {
+    "idx": ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15"],
+    "sentence": ["e", "get pen", "drop book", "x paper", "i", "south", "get paper", "drop pen", "x book", "inventory",
+                 "n", "get book", "drop paper", "examine Pen", "inv", "w"],
+    "label": ["v01835496", "v01214265", "v01977701", "v02131279", "v02472495", "v01835496", "v01214265", "v01977701",
+              "v02131279", "v02472495", "v01835496", "v01214265", "v01977701", "v02131279", "v02472495", "v01835496"]
+}
+
+num_labels = len(set(dict_train["label"]))
+features = Features({'idx': Value('uint32'), 'sentence': Value('string'),
+                     'label': ClassLabel(names=list(set(dict_train["label"])))})
+
+raw_train_dataset = Dataset.from_dict(dict_train, features=features)
+
+discriminator = TFAutoModelForSequenceClassification.from_pretrained("Aureliano/distilbert-base-uncased-if", num_labels=num_labels)
+tokenizer = AutoTokenizer.from_pretrained("Aureliano/distilbert-base-uncased-if")
+
+tokenize_function = lambda example: tokenizer(example["sentence"], truncation=True)
+
+pre_tokenizer_columns = set(raw_train_dataset.features)
+train_dataset = raw_train_dataset.map(tokenize_function, batched=True)
+tokenizer_columns = list(set(train_dataset.features) - pre_tokenizer_columns)
+
+data_collator = DataCollatorWithPadding(tokenizer=tokenizer, return_tensors="tf")
+
+batch_size = 16
+tf_train_dataset = train_dataset.to_tf_dataset(
+    columns=tokenizer_columns,
+    label_cols=["labels"],
+    shuffle=True,
+    batch_size=batch_size,
+    collate_fn=data_collator
+)
+
+loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
+num_epochs = 100
+batches_per_epoch = math.ceil(len(train_dataset) / batch_size)
+total_train_steps = int(batches_per_epoch * num_epochs)
+
+optimizer, schedule = create_optimizer(
+    init_lr=1e-5, num_warmup_steps=1, num_train_steps=total_train_steps
+)
+
+discriminator.compile(optimizer=optimizer, loss=loss)
+discriminator.fit(
+    tf_train_dataset,
+    epochs=num_epochs
+)
+
+text = "get lamp"
+encoded_input = tokenizer(text, return_tensors='tf')
+output = discriminator(encoded_input)
+prediction = tf.nn.softmax(output["logits"][0], -1)
+label = dict_train["label"][tf.math.argmax(prediction)]
+print(text, ":", label)
+# ideally [v01214265 -> take.v.04 -> "get into one's hands, take physically"], but probably only with a better dataset
+
+```

config.json

@@ -0,0 +1,197 @@
+{
+  "_name_or_path": "wn_full_classifier-trainer",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertModel"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2",
+    "3": "LABEL_3",
+    "4": "LABEL_4",
+    "5": "LABEL_5",
+    "6": "LABEL_6",
+    "7": "LABEL_7",
+    "8": "LABEL_8",
+    "9": "LABEL_9",
+    "10": "LABEL_10",
+    "11": "LABEL_11",
+    "12": "LABEL_12",
+    "13": "LABEL_13",
+    "14": "LABEL_14",
+    "15": "LABEL_15",
+    "16": "LABEL_16",
+    "17": "LABEL_17",
+    "18": "LABEL_18",
+    "19": "LABEL_19",
+    "20": "LABEL_20",
+    "21": "LABEL_21",
+    "22": "LABEL_22",
+    "23": "LABEL_23",
+    "24": "LABEL_24",
+    "25": "LABEL_25",
+    "26": "LABEL_26",
+    "27": "LABEL_27",
+    "28": "LABEL_28",
+    "29": "LABEL_29",
+    "30": "LABEL_30",
+    "31": "LABEL_31",
+    "32": "LABEL_32",
+    "33": "LABEL_33",
+    "34": "LABEL_34",
+    "35": "LABEL_35",
+    "36": "LABEL_36",
+    "37": "LABEL_37",
+    "38": "LABEL_38",
+    "39": "LABEL_39",
+    "40": "LABEL_40",
+    "41": "LABEL_41",
+    "42": "LABEL_42",
+    "43": "LABEL_43",
+    "44": "LABEL_44",
+    "45": "LABEL_45",
+    "46": "LABEL_46",
+    "47": "LABEL_47",
+    "48": "LABEL_48",
+    "49": "LABEL_49",
+    "50": "LABEL_50",
+    "51": "LABEL_51",
+    "52": "LABEL_52",
+    "53": "LABEL_53",
+    "54": "LABEL_54",
+    "55": "LABEL_55",
+    "56": "LABEL_56",
+    "57": "LABEL_57",
+    "58": "LABEL_58",
+    "59": "LABEL_59",
+    "60": "LABEL_60",
+    "61": "LABEL_61",
+    "62": "LABEL_62",
+    "63": "LABEL_63",
+    "64": "LABEL_64",
+    "65": "LABEL_65",
+    "66": "LABEL_66",
+    "67": "LABEL_67",
+    "68": "LABEL_68",
+    "69": "LABEL_69",
+    "70": "LABEL_70",
+    "71": "LABEL_71",
+    "72": "LABEL_72",
+    "73": "LABEL_73",
+    "74": "LABEL_74",
+    "75": "LABEL_75",
+    "76": "LABEL_76",
+    "77": "LABEL_77",
+    "78": "LABEL_78",
+    "79": "LABEL_79",
+    "80": "LABEL_80",
+    "81": "LABEL_81",
+    "82": "LABEL_82",
+    "83": "LABEL_83"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_10": 10,
+    "LABEL_11": 11,
+    "LABEL_12": 12,
+    "LABEL_13": 13,
+    "LABEL_14": 14,
+    "LABEL_15": 15,
+    "LABEL_16": 16,
+    "LABEL_17": 17,
+    "LABEL_18": 18,
+    "LABEL_19": 19,
+    "LABEL_2": 2,
+    "LABEL_20": 20,
+    "LABEL_21": 21,
+    "LABEL_22": 22,
+    "LABEL_23": 23,
+    "LABEL_24": 24,
+    "LABEL_25": 25,
+    "LABEL_26": 26,
+    "LABEL_27": 27,
+    "LABEL_28": 28,
+    "LABEL_29": 29,
+    "LABEL_3": 3,
+    "LABEL_30": 30,
+    "LABEL_31": 31,
+    "LABEL_32": 32,
+    "LABEL_33": 33,
+    "LABEL_34": 34,
+    "LABEL_35": 35,
+    "LABEL_36": 36,
+    "LABEL_37": 37,
+    "LABEL_38": 38,
+    "LABEL_39": 39,
+    "LABEL_4": 4,
+    "LABEL_40": 40,
+    "LABEL_41": 41,
+    "LABEL_42": 42,
+    "LABEL_43": 43,
+    "LABEL_44": 44,
+    "LABEL_45": 45,
+    "LABEL_46": 46,
+    "LABEL_47": 47,
+    "LABEL_48": 48,
+    "LABEL_49": 49,
+    "LABEL_5": 5,
+    "LABEL_50": 50,
+    "LABEL_51": 51,
+    "LABEL_52": 52,
+    "LABEL_53": 53,
+    "LABEL_54": 54,
+    "LABEL_55": 55,
+    "LABEL_56": 56,
+    "LABEL_57": 57,
+    "LABEL_58": 58,
+    "LABEL_59": 59,
+    "LABEL_6": 6,
+    "LABEL_60": 60,
+    "LABEL_61": 61,
+    "LABEL_62": 62,
+    "LABEL_63": 63,
+    "LABEL_64": 64,
+    "LABEL_65": 65,
+    "LABEL_66": 66,
+    "LABEL_67": 67,
+    "LABEL_68": 68,
+    "LABEL_69": 69,
+    "LABEL_7": 7,
+    "LABEL_70": 70,
+    "LABEL_71": 71,
+    "LABEL_72": 72,
+    "LABEL_73": 73,
+    "LABEL_74": 74,
+    "LABEL_75": 75,
+    "LABEL_76": 76,
+    "LABEL_77": 77,
+    "LABEL_78": 78,
+    "LABEL_79": 79,
+    "LABEL_8": 8,
+    "LABEL_80": 80,
+    "LABEL_81": 81,
+    "LABEL_82": 82,
+    "LABEL_83": 83,
+    "LABEL_9": 9
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "pad_token_id": 0,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.17.0",
+  "vocab_size": 30522
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12eaf2d9db88163a36fbbc5e45331658bc35d26421cb061f05c473c7708a30d4
+size 265487161

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d789b0b93a1504985b62a35e46433ea14548042e4bdb2e9710619198d6a7b62e
+size 265571752

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "distilbert-base-uncased", "tokenizer_class": "DistilBertTokenizer"}