Spaces:
Sleeping
Sleeping
import streamlit as st | |
from utils import get_res | |
st.sidebar.title('Tokenizers demo') | |
#x = st.slider('Select a value') | |
#st.write(x, 'squared is', x * x) | |
#st.sidebar.subheader('Choose the tokenizer', divider='grey') | |
#option = st.sidebar.selectbox( | |
# 'model_name', | |
# ['deepseek-ai/deepseek-coder-1.3b-instruct', | |
# 'bigcode/starcoder']) | |
model_name_A = st.sidebar.text_input('Model Name A', 'deepseek-ai/deepseek-coder-1.3b-instruct') | |
model_name_B = st.sidebar.text_input('Model Name B', 'deepseek-ai/deepseek-coder-1.3b-instruct') | |
model_option = ['deepseek-ai/deepseek-coder-1.3b-instruct', | |
'MediaTek-Research/Breeze-7B-Instruct-64k-v0_1', | |
'microsoft/phi-2'] | |
with st.sidebar.expander("Models that you might want"): | |
for m in model_option: | |
st.write(m) | |
#'Your choice:', model_name | |
st.sidebar.subheader('Write the input sentence', divider='grey') | |
input_data = st.sidebar.text_input('Input Sentence', 'Hello sunshine!!!') | |
col1, col2 = st.columns(2) | |
with col1: | |
st.subheader(model_name_A, divider='grey') | |
res, token_num = get_res(model_name=model_name_A, input_sentence=input_data, single_print=False) | |
st.subheader('Tokenized result') | |
st.markdown(res, unsafe_allow_html=True) | |
st.subheader('Number of tokens') | |
st.markdown(f'<span style="font-size:1.875em">{str(token_num)}</span>', | |
unsafe_allow_html=True) | |
with col2: | |
st.subheader(model_name_B, divider='grey') | |
res, token_num = get_res(model_name=model_name_B, input_sentence=input_data, single_print=False) | |
st.subheader('Tokenized result') | |
st.markdown(res, unsafe_allow_html=True) | |
st.subheader('Number of tokens') | |
st.markdown(f'<span style="font-size:1.875em">{str(token_num)}</span>', | |
unsafe_allow_html=True) | |