Update

app.py

@@ -13,22 +13,35 @@ from transformers import AutoTokenizer, AutoModelWithLMHead
 st.set_page_config(
     page_title="Romanian Text Generator",
     page_icon="🇷🇴",
+    layout="wide"
 )
 
-tokenizer = AutoTokenizer.from_pretrained("dumitrescustefan/gpt-neo-romanian-780m")
+model_list = ["dumitrescustefan/gpt-neo-romanian-780m"]
+st.sidebar.header("Select Model")
+model_checkpoint = st.sidebar.radio("", model_list)
+text_element = st.text_input('Text:', 'Acesta este un exemplu,')
 
 
-@st.cache
-def load_model(model_name):
-    model = AutoModelWithLMHead.from_pretrained("dumitrescustefan/gpt-neo-romanian-780m")
-    return model
 
-model = load_model("dumitrescustefan/gpt-neo-romanian-780m")
 
 
-def infer(input_ids, max_length, temperature, top_k, top_p):
+st.sidebar.header("Select type of PERSON detection")
+max_length = st.sidebar.slider("Max Length", value=20, min_value=10, max_value=200)
+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=15, step=1, value=0)
+top_p = st.sidebar.slider("Top-p", min_value=0.0, max_value=1.0, step=0.05, value=0.9)
+
+
+@st.cache(allow_output_mutation=True)
+def setModel(model_name):
+    model = AutoModelWithLMHead.from_pretrained(model_name)
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    return model, tokenizer
+
+def infer(model, tokenizer, text, input_ids, max_length, temperature, top_k, top_p):
+    encoded_prompt = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt")
     output_sequences = model.generate(
-        input_ids=input_ids,
+        input_ids=encoded_prompt.input_ids,
         max_length=max_length,
         temperature=temperature,
         top_k=top_k,
@@ -40,26 +53,7 @@ def infer(input_ids, max_length, temperature, top_k, top_p):
     return output_sequences
 
 
-default_value = "See how a modern neural network auto-completes your text 🤗 This site, built by the Hugging Face team, lets you write a whole document directly from your browser, and you can trigger the Transformer anywhere using the Tab key. Its like having a smart machine that completes your thoughts 😀 Get started by typing a custom snippet, check out the repository, or try one of the examples. Have fun!"
-
-# prompts
-st.title("Write")
-st.write(
-    "The almighty king of text generation, GPT-2 comes in four available sizes, only three of which have been publicly made available. Feared for its fake news generation capabilities, it currently stands as the most syntactically coherent model. A direct successor to the original GPT, it reinforces the already established pre-training/fine-tuning killer duo. From the paper: Language Models are Unsupervised Multitask Learners by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.")
-
-sent = st.text_area("Text", default_value, height=275)
-max_length = st.sidebar.slider("Max Length", min_value=10, max_value=30)
-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)
-
-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
-
-output_sequences = infer(input_ids, max_length, temperature, top_k, top_p)
+output_sequences = infer(model, tokenizer, text_element, input_ids, max_length, temperature, top_k, top_p)
 
 for generated_sequence_idx, generated_sequence in enumerate(output_sequences):
     print(f"=== GENERATED SEQUENCE {generated_sequence_idx + 1} ===")
@@ -79,5 +73,5 @@ for generated_sequence_idx, generated_sequence in enumerate(output_sequences):
     generated_sequences.append(total_sequence)
     print(total_sequence)
 
-st.write(generated_sequences[-1])
+st.write(generated_sequences[-1], text_element)