Spaces:
Running
Running
Upload folder using huggingface_hub
Browse files
app.py
CHANGED
@@ -7,6 +7,7 @@ import gradio as gr
|
|
7 |
import faiss
|
8 |
import numpy as np
|
9 |
import torch
|
|
|
10 |
from sentence_transformers import SentenceTransformer
|
11 |
|
12 |
|
@@ -17,7 +18,9 @@ file_example = """Please upload a JSON file with a "text" field (with optional "
|
|
17 |
{"title": "Title A", "text": "This an example text with the title"},
|
18 |
{"title": "Title B", "text": "This an example text with the title"},
|
19 |
]
|
20 |
-
```
|
|
|
|
|
21 |
|
22 |
|
23 |
def create_index(embeddings, use_gpu):
|
@@ -42,13 +45,26 @@ def upload_file_fn(
|
|
42 |
documents = []
|
43 |
for obj in document_data:
|
44 |
text = obj["title"] + "\n" + obj["text"] if obj.get("title") else obj["text"]
|
45 |
-
|
|
|
|
|
|
|
46 |
except Exception as e:
|
47 |
print(e)
|
48 |
-
gr.
|
49 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
|
51 |
-
documents_embeddings = model.encode(documents)
|
52 |
|
53 |
document_index = create_index(documents_embeddings, use_gpu=False)
|
54 |
|
@@ -56,25 +72,30 @@ def upload_file_fn(
|
|
56 |
torch.cuda.empty_cache()
|
57 |
torch.cuda.ipc_collect()
|
58 |
|
59 |
-
|
|
|
|
|
60 |
|
61 |
|
62 |
def clear_file_fn():
|
63 |
-
return None
|
64 |
|
65 |
|
66 |
-
def retrieve_document_fn(question,
|
67 |
-
document_data, document_index = document_states
|
68 |
num_retrieval_doc = 3
|
69 |
-
|
|
|
70 |
gr.Warning("Please upload documents first!")
|
71 |
-
return [None for i in range(num_retrieval_doc)]
|
|
|
|
|
|
|
72 |
|
73 |
-
question_embedding = model.encode([instruct + question])
|
74 |
batch_scores, batch_inxs = document_index.search(question_embedding, k=num_retrieval_doc)
|
75 |
|
76 |
answers = [document_data[i]["text"] for i in batch_inxs[0][:num_retrieval_doc]]
|
77 |
-
return
|
78 |
|
79 |
|
80 |
def main(args):
|
@@ -82,10 +103,8 @@ def main(args):
|
|
82 |
|
83 |
model = SentenceTransformer(args.model_name_or_path)
|
84 |
|
85 |
-
|
86 |
-
document_data = gr.State()
|
87 |
|
88 |
-
|
89 |
with open(Path(__file__).parent / "resources/head.html") as html_file:
|
90 |
head = html_file.read().strip()
|
91 |
with gr.Blocks(theme=gr.themes.Soft(font="sans-serif").set(background_fill_primary="linear-gradient(90deg, #e3ffe7 0%, #d9e7ff 100%)", background_fill_primary_dark="linear-gradient(90deg, #4b6cb7 0%, #182848 100%)",),
|
@@ -98,23 +117,24 @@ def main(args):
|
|
98 |
doc_files_box = gr.File(label="Upload Documents", file_types=[".json"], file_count="single")
|
99 |
retrieval_interface = gr.Interface(
|
100 |
fn=retrieve_document_fn,
|
101 |
-
inputs=["text"],
|
102 |
-
outputs=["
|
103 |
-
additional_inputs=[gr.Textbox("Instruct: Given a query, retrieve documents that answer the query. \n Query: ", label="Instruct of Query")
|
104 |
concurrency_limit=1,
|
105 |
)
|
|
|
106 |
|
107 |
doc_files_box.upload(
|
108 |
upload_file_fn,
|
109 |
[doc_files_box],
|
110 |
-
[
|
111 |
queue=True,
|
112 |
trigger_mode="once"
|
113 |
)
|
114 |
doc_files_box.clear(
|
115 |
-
|
116 |
None,
|
117 |
-
[
|
118 |
queue=True,
|
119 |
trigger_mode="once"
|
120 |
)
|
@@ -123,7 +143,8 @@ def main(args):
|
|
123 |
|
124 |
if __name__ == "__main__":
|
125 |
parser = argparse.ArgumentParser()
|
126 |
-
parser.add_argument("--model_name_or_path", type=str, default="HIT-TMG/KaLM-embedding-multilingual-mini-instruct-v1.5")
|
|
|
127 |
|
128 |
args = parser.parse_args()
|
129 |
main(args)
|
|
|
7 |
import faiss
|
8 |
import numpy as np
|
9 |
import torch
|
10 |
+
from tqdm import tqdm
|
11 |
from sentence_transformers import SentenceTransformer
|
12 |
|
13 |
|
|
|
18 |
{"title": "Title A", "text": "This an example text with the title"},
|
19 |
{"title": "Title B", "text": "This an example text with the title"},
|
20 |
]
|
21 |
+
```
|
22 |
+
Due to the computation resources, please test with small scale data.
|
23 |
+
"""
|
24 |
|
25 |
|
26 |
def create_index(embeddings, use_gpu):
|
|
|
45 |
documents = []
|
46 |
for obj in document_data:
|
47 |
text = obj["title"] + "\n" + obj["text"] if obj.get("title") else obj["text"]
|
48 |
+
if len(str(text).strip()):
|
49 |
+
documents.append(text)
|
50 |
+
else:
|
51 |
+
documents.append(model.tokenizer.eos_token)
|
52 |
except Exception as e:
|
53 |
print(e)
|
54 |
+
gr.Error("Read the file failed. Please check the data format.")
|
55 |
+
gr.Error(str(e))
|
56 |
+
return None
|
57 |
+
|
58 |
+
if len(documents) < 3:
|
59 |
+
gr.Error("Please upload more than 3 documents.")
|
60 |
+
return None
|
61 |
+
|
62 |
+
gr.Info(f"Upload {len(documents)} documents.")
|
63 |
+
if len(documents) > 2000:
|
64 |
+
gr.Info(f"Cut uploaded documents to 2000.")
|
65 |
+
documents = documents[: 2000]
|
66 |
|
67 |
+
documents_embeddings = model.encode(documents, show_progress_bar=True)
|
68 |
|
69 |
document_index = create_index(documents_embeddings, use_gpu=False)
|
70 |
|
|
|
72 |
torch.cuda.empty_cache()
|
73 |
torch.cuda.ipc_collect()
|
74 |
|
75 |
+
print("upload is OK")
|
76 |
+
document_state = {"document_data": document_data, "document_index": document_index}
|
77 |
+
return document_state,
|
78 |
|
79 |
|
80 |
def clear_file_fn():
|
81 |
+
return None
|
82 |
|
83 |
|
84 |
+
def retrieve_document_fn(question, document_states, instruct):
|
|
|
85 |
num_retrieval_doc = 3
|
86 |
+
|
87 |
+
if document_states is None:
|
88 |
gr.Warning("Please upload documents first!")
|
89 |
+
return [None for i in range(num_retrieval_doc)] + [None]
|
90 |
+
|
91 |
+
print(document_states)
|
92 |
+
document_data, document_index = document_states["document_data"], document_states["document_index"]
|
93 |
|
94 |
+
question_embedding = model.encode([str(instruct) + str(question)])
|
95 |
batch_scores, batch_inxs = document_index.search(question_embedding, k=num_retrieval_doc)
|
96 |
|
97 |
answers = [document_data[i]["text"] for i in batch_inxs[0][:num_retrieval_doc]]
|
98 |
+
return answers[0], answers[1], answers[2], document_states
|
99 |
|
100 |
|
101 |
def main(args):
|
|
|
103 |
|
104 |
model = SentenceTransformer(args.model_name_or_path)
|
105 |
|
106 |
+
document_state = gr.State()
|
|
|
107 |
|
|
|
108 |
with open(Path(__file__).parent / "resources/head.html") as html_file:
|
109 |
head = html_file.read().strip()
|
110 |
with gr.Blocks(theme=gr.themes.Soft(font="sans-serif").set(background_fill_primary="linear-gradient(90deg, #e3ffe7 0%, #d9e7ff 100%)", background_fill_primary_dark="linear-gradient(90deg, #4b6cb7 0%, #182848 100%)",),
|
|
|
117 |
doc_files_box = gr.File(label="Upload Documents", file_types=[".json"], file_count="single")
|
118 |
retrieval_interface = gr.Interface(
|
119 |
fn=retrieve_document_fn,
|
120 |
+
inputs=["text", document_state],
|
121 |
+
outputs=[gr.Text(label="Recall-1"), gr.Text(label="Recall-2"), gr.Text(label="Recall-3"), gr.State()],
|
122 |
+
additional_inputs=[gr.Textbox("Instruct: Given a query, retrieve documents that answer the query. \n Query: ", label="Instruct of Query")],
|
123 |
concurrency_limit=1,
|
124 |
)
|
125 |
+
|
126 |
|
127 |
doc_files_box.upload(
|
128 |
upload_file_fn,
|
129 |
[doc_files_box],
|
130 |
+
[document_state],
|
131 |
queue=True,
|
132 |
trigger_mode="once"
|
133 |
)
|
134 |
doc_files_box.clear(
|
135 |
+
clear_file_fn,
|
136 |
None,
|
137 |
+
[document_state],
|
138 |
queue=True,
|
139 |
trigger_mode="once"
|
140 |
)
|
|
|
143 |
|
144 |
if __name__ == "__main__":
|
145 |
parser = argparse.ArgumentParser()
|
146 |
+
# parser.add_argument("--model_name_or_path", type=str, default="HIT-TMG/KaLM-embedding-multilingual-mini-instruct-v1.5")
|
147 |
+
parser.add_argument("--model_name_or_path", type=str, default="/raid/hxs/Checkpoints/huggingface_models/bge-base-en-v1.5")
|
148 |
|
149 |
args = parser.parse_args()
|
150 |
main(args)
|