File size: 2,581 Bytes
9f3c7b7
65eef23
93798d6
89e27db
93798d6
 
89e27db
 
 
 
93798d6
 
 
 
89e27db
 
 
93798d6
 
 
 
 
 
 
 
 
 
 
 
65eef23
9f3c7b7
93798d6
 
89e27db
 
 
 
93798d6
 
 
 
 
 
 
 
db7fef9
 
 
 
 
 
93798d6
 
 
 
 
 
 
 
 
 
db7fef9
93798d6
db7fef9
93798d6
9f3c7b7
93798d6
9f3c7b7
db7fef9
 
 
 
93798d6
 
 
 
db7fef9
93798d6
42f5f04
93798d6
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
from turtle import onclick
import streamlit as st
import pandas as pd
import time

EDIT_ALGS = [
    "MEND: Model editor networks using gradient decomposition",
    "SERAC: Semi-parametric editing with a retrieval-augmented counterfactual model",
    "ENN: Editable neural networks",
    "KE: KnowledgeEditor",
    "Fine-tuning",
    "Lookup Cache"
]

st.title("Language Model Editing")
st.write("Choose an editing algorithm, apply some edits, and sample from the model to see how its behavior changes. You can sample the model at any time to see the \"before and after\" of the edits you apply.")
st.markdown("***")

# https://discuss.streamlit.io/t/simple-example-of-persistence-and-waiting-for-input/2111
@st.cache(allow_output_mutation=True)
def Edits():
    return pd.DataFrame([], columns=["Edit input", "Edit label"])

@st.cache(allow_output_mutation=True)
def ModelOutput():
    return pd.DataFrame([], columns=["Input", "Generation", "Edits applied"])

edits = Edits()
model_outputs = ModelOutput()

def reset():
    edits.drop(edits.index, inplace=True)
    model_outputs.drop(edits.index, inplace=True)

    selected_alg = st.session_state.alg_selector
    selected_alg_idx = EDIT_ALGS.index(selected_alg)

    ############# Need to reset the model here (and maybe show progress spinner?)

def apply_edit():
    edits.loc[len(edits)] = [str(edit_input), str(edit_label)]

def sample_model():
    model_outputs.loc[len(model_outputs)] = [str(test_input), "blah blah blah", len(edits)]

col1, col2 = st.columns([5,1])
with col1:
    alg_selector = st.selectbox("Editing algorithm:", EDIT_ALGS, key="alg_selector", on_change=reset)
with col2:
    st.text("ㅤ")
    st.button("Clear edits", on_click=reset)

st.write("Edits applied so far:")
st.table(edits)

st.markdown("***")

col1, col2, col3 = st.columns([3, 2, 1])
with col1:
    edit_input = st.text_input("Edit input:", placeholder="e.g., 'What is the tallest mountain on Earth?'")
with col2:
    edit_label = st.text_input("Edit target:", placeholder="e.g., 'Denali'")
with col3:
    st.text("ㅤ")
    edit_button = st.button("Apply edit", on_click=apply_edit)

st.markdown("***")

if len(edits) == 0:
    title = "Input to sample from *unedited* model:"
else:
    title = f"Input to sample from *edited* model:"
col1, col2 = st.columns([5, 1])
with col1:
    test_input = st.text_input(title, placeholder="e.g., 'What is the earth's tallest mountain?'")
with col2:
    st.text("ㅤ")
    generate_button = st.button("Generate", on_click=sample_model)

st.write("Model generation history:")
st.table(model_outputs)