welcome bert

backend/app.py

@@ -1,9 +1,8 @@
-from fastapi import FastAPI, HTTPException
-from fastapi import Body
+from fastapi import FastAPI, HTTPException, Body
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.staticfiles import StaticFiles
 from fastapi.responses import FileResponse
-from transformers import GPT2Tokenizer, GPT2Model, pipeline
+from transformers import BertTokenizer, BertModel, pipeline
 import torch as t
 import logging
 
@@ -21,14 +20,14 @@ app.add_middleware(
 )
 
 # Mount static files (frontend) so that visiting "/" serves index.html
-# Note: The directory path "../frontend" works because when running in Docker,
+# The directory path "../frontend" works because when running in Docker,
 # our working directory is set to /app, and the frontend folder is at /app/frontend.
 app.mount("/static", StaticFiles(directory="frontend", html=True), name="static")
 
-# Load tokenizer and GPT2 model
-tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
+# Load tokenizer and BERT model
+tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
 try:
-    model = GPT2Model.from_pretrained('gpt2', output_attentions=True)
+    model = BertModel.from_pretrained('bert-base-uncased', output_attentions=True)
 except Exception as e:
     logger.error(f"Model loading failed: {e}")
     raise
@@ -37,8 +36,8 @@ except Exception as e:
 async def process_text(text: str = Body(..., embed=True)):
     """
     Process the input text:
-      - Tokenizes the text
-      - Runs the GPT2 model to obtain attentions
+      - Tokenizes the text using BERT's tokenizer
+      - Runs the BERT model to obtain attentions (bidirectional)
       - Returns the tokens and attention values (rounded to 2 decimals)
     """
     try:
@@ -53,9 +52,10 @@ async def process_text(text: str = Body(..., embed=True)):
 
         decimals = 2
         # Convert attention tensors to lists with rounded decimals
-        attn_series = t.round(t.tensor([
-            layer_attention.tolist() for layer_attention in attentions
-        ], dtype=t.double).squeeze(), decimals=decimals).detach().cpu().tolist()
+        attn_series = t.round(
+            t.tensor([layer_attention.tolist() for layer_attention in attentions], dtype=t.double)
+            .squeeze(), decimals=decimals
+        ).detach().cpu().tolist()
 
         return {
             "tokens": tokenizer.convert_ids_to_tokens(inputs["input_ids"][0]),
@@ -65,26 +65,17 @@ async def process_text(text: str = Body(..., embed=True)):
         logger.error(f"Error processing text: {e}")
         raise HTTPException(status_code=500, detail=str(e))
 
-# Initialize the text generation pipeline
-# This function will be able to generate text
-# given an input.
-pipe = pipeline("text2text-generation",
-model="google/flan-t5-small")
+# Initialize the text generation pipeline (unchanged)
+pipe = pipeline("text2text-generation", model="google/flan-t5-small")
 
-# Define a function to handle the GET request at `/generate`
-# The generate() function is defined as a FastAPI route that takes a
-# string parameter called text. The function generates text based on the # input using the pipeline() object, and returns a JSON response
-# containing the generated text under the key "output"
 @app.get("/generate")
 def generate(text: str):
     """
     Using the text2text-generation pipeline from `transformers`, generate text
-    from the given input text. The model used is `google/flan-t5-small`, which
-    can be found [here](<https://huggingface.co/google/flan-t5-small>).
+    from the given input text. The model used is `google/flan-t5-small`.
     """
     # Use the pipeline to generate text from the given input text
     output = pipe(text)
-
     # Return the generated text in a JSON response
     return {"output": output[0]["generated_text"]}