Spaces:
Runtime error
Runtime error
Adding finetuned model as a demo
Browse files
app.py
CHANGED
@@ -1,28 +1,59 @@
|
|
1 |
import streamlit as st
|
2 |
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
|
|
|
3 |
|
4 |
-
st.
|
5 |
-
st.markdown('A simple demo using Sinhala-gpt2 model trained during hf-flax week')
|
6 |
|
7 |
-
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
10 |
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
tokenizer = AutoTokenizer.from_pretrained('flax-community/Sinhala-gpt2')
|
15 |
-
st.success('Model loaded!!')
|
16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
st.markdown('____________')
|
28 |
st.markdown('by Keshan with Flax Community')
|
|
|
1 |
import streamlit as st
|
2 |
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
|
3 |
+
# from huggingface_hub import snapshot_download
|
4 |
|
5 |
+
page = st.sidebar.selectbox("Model ", ["Pretrained GPT2", "Finetuned on News data"])
|
|
|
6 |
|
7 |
+
def load_model(model_name):
|
8 |
+
with st.spinner('Waiting for the model to load.....'):
|
9 |
+
# snapshot_download('flax-community/Sinhala-gpt2')
|
10 |
+
model = AutoModelForCausalLM.from_pretrained(model_name)
|
11 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
12 |
+
st.success('Model loaded!!')
|
13 |
+
return model, tokenizer
|
14 |
|
15 |
+
seed = st.sidebar.text_input('Starting text', 'ආයුබෝවන්')
|
16 |
+
seq_num = st.sidebar.number_input('Number of sentences to generate ', 1, 20, 5)
|
17 |
+
max_len = st.sidebar.number_input('Length of the sentence ', 5, 300, 100)
|
|
|
|
|
18 |
|
19 |
+
if page == "Finetuned on News data":
|
20 |
+
|
21 |
+
st.title('Sinhala Text generation with Finetuned GPT2')
|
22 |
+
st.markdown('This model has been finetuned Sinhala-gpt2 model with 6000 news articles(~12MB)')
|
23 |
+
|
24 |
+
# seed = st.text_input('Starting text', 'ආයුබෝවන්')
|
25 |
+
# seq_num = st.number_input('Number of sentences to generate ', 1, 20, 5)
|
26 |
+
# max_len = st.number_input('Length of the sentence ', 5, 300, 100)
|
27 |
|
28 |
+
gen_news = st.button('Generate')
|
29 |
+
model, tokenizer = load_model('keshan/sinhala-gpt2-newswire')
|
30 |
+
|
31 |
+
|
32 |
+
if gen_news:
|
33 |
+
try:
|
34 |
+
with st.spinner('Generating...'):
|
35 |
+
generator = pipeline('text-generation', model=model, tokenizer=tokenizer)
|
36 |
+
seqs = generator(seed, max_length=max_len, num_return_sequences=seq_num)
|
37 |
+
st.write(seqs)
|
38 |
+
except Exception as e:
|
39 |
+
st.exception(f'Exception: {e}')
|
40 |
+
else:
|
41 |
+
st.title('Sinhala Text generation with GPT2')
|
42 |
+
st.markdown('A simple demo using Sinhala-gpt2 model trained during hf-flax week')
|
43 |
+
|
44 |
+
gen_gpt2 = st.button('Generate')
|
45 |
+
model, tokenizer = load_model('flax-community/Sinhala-gpt2')
|
46 |
+
|
47 |
+
|
48 |
+
if gen_gpt2:
|
49 |
+
try:
|
50 |
+
with st.spinner('Generating...'):
|
51 |
+
generator = pipeline('text-generation', model=model, tokenizer=tokenizer)
|
52 |
+
seqs = generator(seed, max_length=max_len, num_return_sequences=seq_num)
|
53 |
+
st.write(seqs)
|
54 |
+
except Exception as e:
|
55 |
+
st.exception(f'Exception: {e}')
|
56 |
+
|
57 |
|
58 |
st.markdown('____________')
|
59 |
st.markdown('by Keshan with Flax Community')
|