Duplicate from shangrilar/cat_prediction

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,13 @@
+---
+title: Cat Prediction
+emoji: 🏆
+colorFrom: red
+colorTo: purple
+sdk: gradio
+sdk_version: 3.39.0
+app_file: app.py
+pinned: false
+duplicated_from: shangrilar/cat_prediction
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,139 @@
+import os
+import pickle
+import torch
+import random
+import numpy as np
+import pandas as pd
+import gradio as gr
+from setfit import SetFitModel
+from huggingface_hub import login, hf_hub_download
+hf_token = os.getenv('hf_token')
+login(hf_token)
+
+def prepare_setfit_model(repo_id):
+    model = SetFitModel.from_pretrained(repo_id)
+    id2cat_path = hf_hub_download(repo_id, filename='id2cat.pkl')
+    with open(id2cat_path, "rb") as file:
+        id2cat = pickle.load(file)
+    cat2id_path = hf_hub_download(repo_id, filename='cat2id.pkl')
+    with open(cat2id_path, "rb") as file:
+        cat2id = pickle.load(file)
+    cat_name_path = hf_hub_download(repo_id, filename='cat_name.csv')
+    df_cat = pd.read_csv(cat_name_path)
+    return model, id2cat, cat2id, df_cat
+
+cat1_model, cat1_id2cat, cat1_cat2id, df_cat = prepare_setfit_model(os.getenv('cat1_repo_id'))
+cat2_model, cat2_id2cat, cat2_cat2id, df_cat = prepare_setfit_model(os.getenv('cat2_repo_id'))
+cid_model, cid_id2cat, cid_cat2id, df_cat = prepare_setfit_model(os.getenv('cid_repo_id'))
+
+
+# -
+
+def model_predict(model, sentence):
+    with torch.no_grad():
+        predict_result = model.predict_proba(sentence).cpu().detach().numpy()
+        sorted_ids = np.argsort(predict_result)[::-1]
+        sorted_probs = np.sort(predict_result)[::-1]
+    return sorted_ids, sorted_probs
+
+
+# +
+def run_prediction(sentence, state):
+    sorted_cat1_ids, sorted_cat1_probs = model_predict(cat1_model, sentence)
+    sorted_cat2_ids, sorted_cat2_probs = model_predict(cat2_model, sentence)
+    sorted_cid_ids, sorted_cid_probs = model_predict(cid_model, sentence)
+    sorted_cat1_ids = [cat1_id2cat[item] for item in sorted_cat1_ids]
+    sorted_cat2_ids = [cat2_id2cat[item] for item in sorted_cat2_ids]
+    sorted_cid_ids = [cid_id2cat[item] for item in sorted_cid_ids]
+    
+    cat1_names = ['select'] + list(map(id2catname.get, sorted_cat1_ids))
+    cat2_names = ['select'] + list(map(id2catname.get, sorted_cat2_ids))
+    cid_names = ['select'] + list(map(id2catname.get, sorted_cid_ids))
+    
+    state['cat1_names'] = cat1_names
+    state['sorted_cat1_probs'] = sorted_cat1_probs
+    state['cat2_names'] = cat2_names
+    state['sorted_cat2_probs'] = sorted_cat2_probs
+    state['cid_names'] = cid_names
+    state['sorted_cid_probs'] = sorted_cid_probs
+    return gr.Dropdown.update(
+        choices = cat1_names, value = cat1_names[0], interactive=True
+    ), gr.Dropdown.update(
+        choices = cat2_names, value = cat2_names[0], interactive=True
+    ), gr.Dropdown.update(
+        choices = cid_names, value = cid_names[0], interactive=True
+    ), state
+
+def filter_cat2(cat1_name, state):
+    cat2_names = []
+    cat2_list = parent_cat_map.get(cat1_name)
+    
+    for item in state['cat2_names']:
+        if item in cat2_list and item not in cat2_names:
+            cat2_names.append(item)
+    cat2_names = ['select'] + cat2_names
+    return gr.Dropdown.update(
+        choices=cat2_names, value=cat2_names[0], interactive=True
+    ), state
+
+def filter_cid(cat2_name, state):
+    cid_names = []
+    cid_list = parent_cat_map.get(cat2_name)
+    if cid_list is None:
+        return gr.Dropdown.update(
+        choices=['None'], value='None', interactive=False
+    )
+    for item in state['cid_names']:
+        if item in cid_list and item not in cid_names:
+            cid_names.append(item)
+    cid_names = ['select'] + cid_names
+    return gr.Dropdown.update(
+        choices=cid_names, value=cid_names[0], interactive=True
+    )
+
+# def predict_with_title_and_description(title, description):
+#     temp_list = list(locations.keys())
+#     random.shuffle(temp_list)
+#     countries = ['select'] + temp_list
+#     return gr.Dropdown.update(
+#                 choices=countries, value=countries[0], interactive=True
+#             )
+
+parent_cat = df_cat[['id', 'name']]
+parent_cat.columns = ['temp_id', 'parent_name']
+df_cat = pd.merge(df_cat, parent_cat, left_on='parent_id', right_on='temp_id', how='left').drop('temp_id', axis=1)
+
+id2catname = {item['id']:item['name'] for item in df_cat[['id', 'name']].to_dict(orient='records')}
+
+parent_cat_map = {}
+for item in df_cat[['parent_name', 'name']].to_dict(orient='records'):
+    if item['parent_name'] in parent_cat_map:
+        parent_cat_map[item['parent_name']].append(item['name'])
+    else:
+        parent_cat_map[item['parent_name']] = [item['name']]
+
+with gr.Blocks() as demo:
+    prediction_results = gr.State({})
+    with gr.Tab(label="Predict by title") as t1:
+        title = gr.Textbox(label='Service Title', placeholder='Please enter service title')
+        d1 = gr.Dropdown(choices = list(), label="Cat 1")
+        d2 = gr.Dropdown(choices = list(), label='Cat 2')
+        d3 = gr.Dropdown(choices = list(), label="CID")
+        
+        b1 = gr.Button()
+        b1.click(run_prediction, [title, prediction_results], [d1, d2, d3, prediction_results])
+        d1.select(filter_cat2, [d1, prediction_results], [d2, prediction_results])
+        d2.select(filter_cid, [d2, prediction_results], d3)
+#     with gr.Tab(label="Predict by title and description") as t2:
+#         title = gr.Textbox(label='Service Title', placeholder='Please enter service title')
+#         description = gr.Textbox(label='Service Description', placeholder="Please enter service description")
+#         d1 = gr.Dropdown(choices = list(locations.keys()), label="Country")
+#         d2 = gr.Dropdown(choices = list(), label='State')
+#         d3 = gr.Dropdown(choices = list(), label="City")
+        
+#         b1 = gr.Button()
+#         b1.click(predict_with_title_and_description, [title, description], d1)
+#         d1.change(filter_states, d1, d2)
+#         d2.change(filter_cities, [d1, d2], d3)
+    
+demo.queue(max_size=5).launch()

requirements.txt

@@ -0,0 +1 @@
+setfit