Upload folder using huggingface_hub

added_tokens.json

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

config.json

@@ -1,23 +1,13 @@
-{
-  "architectures": [
-    "RobertaForMaskedLM"
-  ],
-  "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": 0,
-  "eos_token_id": 2,
-  "gradient_checkpointing": false,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
-  "initializer_range": 0.02,
-  "intermediate_size": 3072,
-  "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 258,
-  "model_type": "roberta",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "pad_token_id": 1,
-  "tokenizer_class": "PhobertTokenizer",
-  "type_vocab_size": 1,
-  "vocab_size": 64001
-}
+{
+  "architectures": ["PhoBERTClassifier"],
+  "model_type": "phobert_aspect",
+  "base_model_name": "vinai/phobert-base",
+  "num_labels": 3,
+  "num_aspects": 4,
+  "dropout": 0.3,
+  "dtype": "float32",
+  "auto_map": {
+    "AutoConfig": "modeling_phobert_aspect.PhoBERTAspectConfig",
+    "AutoModel": "modeling_phobert_aspect.PhoBERTClassifier"
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb579e74bb0081c48014e687c582a4e542154b4042fe56775e3267eb681e0f69
+size 540054608

modeling_phobert_aspect.py

@@ -0,0 +1,53 @@
+from transformers import PretrainedConfig, PreTrainedModel, AutoModel
+import torch
+import torch.nn as nn
+
+class PhoBERTAspectConfig(PretrainedConfig):
+    model_type = "phobert_aspect"
+
+    def __init__(
+        self,
+        num_labels=3,
+        num_aspects=4,
+        dropout=0.3,
+        base_model_name="vinai/phobert-base",
+        **kwargs
+    ):
+        super().__init__(**kwargs)
+        self.num_labels = num_labels
+        self.num_aspects = num_aspects
+        self.dropout = dropout
+        self.base_model_name = base_model_name
+
+
+class PhoBERTClassifier(PreTrainedModel):
+    config_class = PhoBERTAspectConfig
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.phobert = AutoModel.from_pretrained(
+            config.base_model_name
+        )
+
+        hidden_size = self.phobert.config.hidden_size
+        self.dropout = nn.Dropout(config.dropout)
+
+        self.classifiers = nn.ModuleList([
+            nn.Linear(hidden_size, config.num_labels)
+            for _ in range(config.num_aspects)
+        ])
+
+        self.post_init()
+
+    def forward(self, input_ids, attention_mask):
+        outputs = self.phobert(
+            input_ids=input_ids,
+            attention_mask=attention_mask
+        )
+
+        pooled = outputs.last_hidden_state[:, 0, :]
+        pooled = self.dropout(pooled)
+
+        logits = [clf(pooled) for clf in self.classifiers]
+        return torch.stack(logits, dim=1)

special_tokens_map.json

@@ -1,6 +1,9 @@
 {
-    "unk_token": "<unk>",
-    "sep_token": "</s>",
-    "pad_token": "<pad>",
-    "cls_token": "<s>"
-}
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -1,11 +1,55 @@
 {
-    "do_lower_case": false,
-    "bos_token": "<s>",
-    "eos_token": "</s>",
-    "unk_token": "<unk>",
-    "pad_token": "<pad>",
-    "sep_token": "</s>",
-    "cls_token": "<s>",
-    "model_max_length": 120,
-    "tokenizer_class": "RobertaTokenizer"
-  }
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}