update app.py

app.py

@@ -10,15 +10,15 @@ def load_model(model_name):
     return model
 
 model = load_model("VietAI/gpt-neo-1.3B-vietnamese-news")
-def infer(input_ids, max_length, temperature, top_k, top_p):
+def infer(input_ids, max_length):
 
     output_sequences = model.generate(
         input_ids=input_ids,
         max_length=max_length,
         do_sample=True,
-        temperature=temperature,
-        top_k=top_k,
-        top_p=top_p,
+        temperature=0.9,
+        top_k=20,
+        #top_p=top_p,
         #num_return_sequences=1
     )
 
@@ -32,19 +32,23 @@ st.write("It might be a bit slow after you change the generated text length. Be
 
 sent = st.text_area("Text", default_value, height = 275)
 max_length = st.sidebar.slider("Max Length", min_value = 10, max_value=500)
-temperature = st.sidebar.slider("Temperature", value = 1.0, min_value = 0.0, max_value=1.0, step=0.05)
-top_k = st.sidebar.slider("Top-k", min_value = 0, max_value=5, value = 0)
-top_p = st.sidebar.slider("Top-p", min_value = 0.0, max_value=1.0, step = 0.05, value = 0.9)
+
+# We don't really need these params. It's a lot slower.
+# temperature = st.sidebar.slider("Temperature", value = 1.0, min_value = 0.0, max_value=1.0, step=0.05)
+# top_k = st.sidebar.slider("Top-k", min_value = 0, max_value=5, value = 0)
+# top_p = st.sidebar.slider("Top-p", min_value = 0.0, max_value=1.0, step = 0.05, value = 0.9)
 
 if st.button("Generate Text"):
-    encoded_prompt = tokenizer.encode(sent, add_special_tokens=False, return_tensors="pt")
-    if encoded_prompt.size()[-1] == 0:
-        input_ids = None
-    else:
-        input_ids = encoded_prompt
-
-    gen_tokens = infer(encoded_prompt, max_length, temperature, top_k, top_p)
-    gen_text = tokenizer.batch_decode(gen_tokens)[0]
-    st.write(gen_text)
+    with st.spinner("Working Hard..."):
+        encoded_prompt = tokenizer.encode(sent, add_special_tokens=False, return_tensors="pt")
+        if encoded_prompt.size()[-1] == 0:
+            input_ids = None
+        else:
+            input_ids = encoded_prompt
+
+        gen_tokens = infer(encoded_prompt, max_length)
+        gen_text = tokenizer.batch_decode(gen_tokens)[0]
+        st.write(gen_text)
+    st.success("Done!")
 
 st.write("For feedback/requests, write to mel.nguyen273@gmail.com.")