turn into module for docs

README.md

@@ -6,7 +6,7 @@ colorTo: indigo
 python_version: 3.9
 sdk: streamlit
 sdk_version: 1.10.0
-app_file: src/main.py
+app_file: src/app.py
 pinned: true
 ---
 

src/{main.py → app.py}

@@ -2,8 +2,8 @@ import pandas as pd
 import streamlit as st
 from streamlit_option_menu import option_menu
 
-from load import load_context
-from subpages import (
+from src.load import load_context
+from src.subpages import (
     DebugPage,
     FindDuplicatesPage,
     HomePage,
@@ -16,10 +16,10 @@ from subpages import (
     RandomSamplesPage,
     RawDataPage,
 )
-from subpages.attention import AttentionPage
-from subpages.hidden_states import HiddenStatesPage
-from subpages.inspect import InspectPage
-from utils import classmap
+from src.subpages.attention import AttentionPage
+from src.subpages.hidden_states import HiddenStatesPage
+from src.subpages.inspect import InspectPage
+from src.utils import classmap
 
 sts = st.sidebar
 st.set_page_config(

src/data.py

@@ -7,7 +7,7 @@ from datasets import Dataset, DatasetDict, load_dataset  # type: ignore
 from torch.nn.functional import cross_entropy
 from transformers import DataCollatorForTokenClassification  # type: ignore
 
-from utils import device, tokenizer_hash_funcs
+from src.utils import device, tokenizer_hash_funcs
 
 
 @st.cache(allow_output_mutation=True)

src/load.py

@@ -4,10 +4,10 @@ import pandas as pd
 import streamlit as st
 from datasets import Dataset  # type: ignore
 
-from data import encode_dataset, get_collator, get_data, get_split_df
-from model import get_encoder, get_model, get_tokenizer
-from subpages import Context
-from utils import align_sample, device, explode_df
+from src.data import encode_dataset, get_collator, get_data, get_split_df
+from src.model import get_encoder, get_model, get_tokenizer
+from src.subpages import Context
+from src.utils import align_sample, device, explode_df
 
 _TOKENIZER_NAME = (
     "xlm-roberta-base",

src/subpages/__init__.py

@@ -1,14 +1,14 @@
-from subpages.attention import AttentionPage
-from subpages.debug import DebugPage
-from subpages.find_duplicates import FindDuplicatesPage
-from subpages.hidden_states import HiddenStatesPage
-from subpages.home import HomePage
-from subpages.inspect import InspectPage
-from subpages.losses import LossesPage
-from subpages.lossy_samples import LossySamplesPage
-from subpages.metrics import MetricsPage
-from subpages.misclassified import MisclassifiedPage
-from subpages.page import Context, Page
-from subpages.probing import ProbingPage
-from subpages.random_samples import RandomSamplesPage
-from subpages.raw_data import RawDataPage
+from src.subpages.attention import AttentionPage
+from src.subpages.debug import DebugPage
+from src.subpages.find_duplicates import FindDuplicatesPage
+from src.subpages.hidden_states import HiddenStatesPage
+from src.subpages.home import HomePage
+from src.subpages.inspect import InspectPage
+from src.subpages.losses import LossesPage
+from src.subpages.lossy_samples import LossySamplesPage
+from src.subpages.metrics import MetricsPage
+from src.subpages.misclassified import MisclassifiedPage
+from src.subpages.page import Context, Page
+from src.subpages.probing import ProbingPage
+from src.subpages.random_samples import RandomSamplesPage
+from src.subpages.raw_data import RawDataPage

src/subpages/attention.py

@@ -2,7 +2,7 @@ import ecco
 import streamlit as st
 from streamlit.components.v1 import html
 
-from subpages.page import Context, Page  # type: ignore
+from src.subpages.page import Context, Page  # type: ignore
 
 SETUP_HTML = """
 <script src="https://requirejs.org/docs/release/2.3.6/minified/require.js"></script>

src/subpages/debug.py

@@ -1,7 +1,7 @@
 import streamlit as st
 from pip._internal.operations import freeze
 
-from subpages.page import Context, Page
+from src.subpages.page import Context, Page
 
 
 class DebugPage(Page):

src/subpages/faiss.py

@@ -1,8 +1,8 @@
 import streamlit as st
 from datasets import Dataset
 
-from subpages.page import Context, Page  # type: ignore
-from utils import device, explode_df, htmlify_labeled_example, tag_text
+from src.subpages.page import Context, Page  # type: ignore
+from src.utils import device, explode_df, htmlify_labeled_example, tag_text
 
 
 class FaissPage(Page):

src/subpages/find_duplicates.py

@@ -1,7 +1,7 @@
 import streamlit as st
 from sentence_transformers.util import cos_sim
 
-from subpages.page import Context, Page
+from src.subpages.page import Context, Page
 
 
 @st.cache()

src/subpages/hidden_states.py

@@ -3,7 +3,7 @@ import plotly.express as px
 import plotly.graph_objects as go
 import streamlit as st
 
-from subpages.page import Context, Page
+from src.subpages.page import Context, Page
 
 
 @st.cache

src/subpages/home.py

@@ -4,9 +4,9 @@ from typing import Optional
 
 import streamlit as st
 
-from data import get_data
-from subpages.page import Context, Page
-from utils import PROJ, classmap, color_map_color
+from src.data import get_data
+from src.subpages.page import Context, Page
+from src.utils import PROJ, classmap, color_map_color
 
 _SENTENCE_ENCODER_MODEL = (
     "sentence-transformers/all-MiniLM-L6-v2",

src/subpages/inspect.py

@@ -1,7 +1,7 @@
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import aggrid_interactive_table, colorize_classes
+from src.subpages.page import Context, Page
+from src.utils import aggrid_interactive_table, colorize_classes
 
 
 class InspectPage(Page):

src/subpages/losses.py

@@ -1,7 +1,7 @@
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import AgGrid, aggrid_interactive_table
+from src.subpages.page import Context, Page
+from src.utils import AgGrid, aggrid_interactive_table
 
 
 @st.cache

src/subpages/lossy_samples.py

@@ -1,8 +1,13 @@
 import pandas as pd
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import colorize_classes, get_bg_color, get_fg_color, htmlify_labeled_example
+from src.subpages.page import Context, Page
+from src.utils import (
+    colorize_classes,
+    get_bg_color,
+    get_fg_color,
+    htmlify_labeled_example,
+)
 
 
 class LossySamplesPage(Page):

src/subpages/metrics.py

@@ -8,7 +8,7 @@ import streamlit as st
 from seqeval.metrics import classification_report
 from sklearn.metrics import ConfusionMatrixDisplay, confusion_matrix
 
-from subpages.page import Context, Page
+from src.subpages.page import Context, Page
 
 
 def _get_evaluation(df):

src/subpages/misclassified.py

@@ -4,8 +4,8 @@ import pandas as pd
 import streamlit as st
 from sklearn.metrics import confusion_matrix
 
-from subpages.page import Context, Page
-from utils import htmlify_labeled_example
+from src.subpages.page import Context, Page
+from src.utils import htmlify_labeled_example
 
 
 class MisclassifiedPage(Page):

src/subpages/probing.py

@@ -1,7 +1,7 @@
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import device, tag_text
+from src.subpages.page import Context, Page
+from src.utils import device, tag_text
 
 _DEFAULT_SENTENCES = """
 Damit hatte er auf ihr letztes , völlig schiefgelaufenes Geschäftsessen angespielt .

src/subpages/random_samples.py

@@ -1,8 +1,8 @@
 import pandas as pd
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import htmlify_labeled_example
+from src.subpages.page import Context, Page
+from src.utils import htmlify_labeled_example
 
 
 class RandomSamplesPage(Page):

src/subpages/raw_data.py

@@ -1,8 +1,8 @@
 import pandas as pd
 import streamlit as st
 
-from subpages.page import Context, Page
-from utils import aggrid_interactive_table
+from src.subpages.page import Context, Page
+from src.utils import aggrid_interactive_table
 
 
 @st.cache