Updated app with custom DeBERTaV3 model

app.py

@@ -1,63 +1,30 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()
+import torch
+from transformers import AutoTokenizer
+
+from models import DebertaV3ForCustomClassification
+
+tokenizer = AutoTokenizer.from_pretrained(
+    's986103/DebertaV3ForCustomClassification')
+
+model = DebertaV3ForCustomClassification.from_pretrained(
+    's986103/DebertaV3ForCustomClassification')
+model.eval()
+
+
+def classify_text(text):
+    inputs = tokenizer(text, return_tensors="pt",
+                       truncation=True, padding=True)
+    with torch.no_grad():
+        logits = model(**inputs)
+    prediction = torch.argmax(logits, dim=1).item()
+    return prediction
+
+
+iface = gr.Interface(fn=classify_text,
+                     inputs="text",
+                     outputs="label",
+                     description="自動作文評分")
+
+# 啟動 UI
+iface.launch()

models.py

@@ -0,0 +1,37 @@
+import torch
+from torch import nn
+from transformers import DebertaV2Model, DebertaV2PreTrainedModel
+
+
+class DebertaV3ForCustomClassification(DebertaV2PreTrainedModel):
+    def __init__(self, config):
+        super().__init__(config)
+        self.deberta = DebertaV2Model(config)  # 使用 DebertaV2 作为基础模型
+        self.dropout = nn.Dropout(0.1)  # 添加一个 dropout 层
+        self.classifier = nn.Linear(
+            config.hidden_size, config.num_labels)  # fully connected 层
+        self.config = config  # 保存配置
+
+    def forward(self, input_ids, attention_mask=None, token_type_ids=None, labels=None):
+        # 获取 DeBERTaV3 的输出
+        outputs = self.deberta(
+            input_ids, attention_mask=attention_mask, token_type_ids=token_type_ids)
+
+        # 使用 Mean Pooling
+        # [batch_size, seq_len, hidden_size]
+        last_hidden_state = outputs.last_hidden_state
+        # [batch_size, hidden_size]
+        pooled_output = torch.mean(last_hidden_state, dim=1)
+
+        # Dropout and Fully Connected Layer
+        pooled_output = self.dropout(pooled_output)
+        logits = self.classifier(pooled_output)  # [batch_size, num_labels]
+
+        # 如果提供了标签，则计算损失
+        loss = None
+        if labels is not None:
+            loss_fct = nn.CrossEntropyLoss()
+            loss = loss_fct(
+                logits.view(-1, self.config.num_labels), labels.view(-1))
+
+        return (loss, logits) if loss is not None else logits

requirements.txt

@@ -1 +1,4 @@
-huggingface_hub==0.22.2
+huggingface_hub==0.22.2
+gradio
+torch
+transformers