Initial upload: PhoBERT NER for Vietnamese Smart Home slot filling

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

label_mappings.json

@@ -0,0 +1,47 @@
+{
+  "label2id": {
+    "O": 0,
+    "B-changing value": 1,
+    "I-changing value": 2,
+    "B-device": 3,
+    "I-device": 4,
+    "B-duration": 5,
+    "I-duration": 6,
+    "B-living_space": 7,
+    "I-living_space": 8,
+    "B-target number": 9,
+    "I-target number": 10,
+    "B-time at": 11,
+    "I-time at": 12
+  },
+  "id2label": {
+    "0": "O",
+    "1": "B-changing value",
+    "2": "I-changing value",
+    "3": "B-device",
+    "4": "I-device",
+    "5": "B-duration",
+    "6": "I-duration",
+    "7": "B-living_space",
+    "8": "I-living_space",
+    "9": "B-target number",
+    "10": "I-target number",
+    "11": "B-time at",
+    "12": "I-time at"
+  },
+  "slot_labels": [
+    "O",
+    "B-changing value",
+    "I-changing value",
+    "B-device",
+    "I-device",
+    "B-duration",
+    "I-duration",
+    "B-living_space",
+    "I-living_space",
+    "B-target number",
+    "I-target number",
+    "B-time at",
+    "I-time at"
+  ]
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7434a65d3615865a003237f6feecfb92234c96ed2818e38dd4251e871e27ae5d
+size 537741802

slot_labels.json

@@ -0,0 +1,15 @@
+[
+  "O",
+  "B-changing value",
+  "I-changing value",
+  "B-device",
+  "I-device",
+  "B-duration",
+  "I-duration",
+  "B-living_space",
+  "I-living_space",
+  "B-target number",
+  "I-target number",
+  "B-time at",
+  "I-time at"
+]

special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "model_max_length": 256,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}