Spaces:
Runtime error
Runtime error
Update app to register interactions in an argilla dataset
Browse files- app.py +139 -47
- requirements.txt +2 -1
app.py
CHANGED
@@ -1,4 +1,4 @@
|
|
1 |
-
from typing import Optional,
|
2 |
import os
|
3 |
from pathlib import Path
|
4 |
import tarfile
|
@@ -11,12 +11,14 @@ from huggingface_hub.file_download import hf_hub_download
|
|
11 |
from huggingface_hub import InferenceClient, login
|
12 |
from transformers import AutoTokenizer
|
13 |
import gradio as gr
|
|
|
|
|
14 |
|
15 |
|
16 |
@dataclass
|
17 |
class Settings:
|
18 |
-
"""Settings class to store useful variables for the App.
|
19 |
-
|
20 |
LANCEDB: str = "lancedb"
|
21 |
LANCEDB_FILE_TAR: str = "lancedb.tar.gz"
|
22 |
TOKEN: str = os.getenv("HF_API_TOKEN")
|
@@ -24,13 +26,29 @@ class Settings:
|
|
24 |
REPO_ID: str = "plaguss/argilla_sdk_docs_queries"
|
25 |
TABLE_NAME: str = "docs"
|
26 |
MODEL_NAME: str = "plaguss/bge-base-argilla-sdk-matryoshka"
|
27 |
-
DEVICE: str =
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
MODEL_ID: str = "meta-llama/Meta-Llama-3-70B-Instruct"
|
|
|
|
|
|
|
|
|
29 |
|
30 |
settings = Settings()
|
31 |
|
32 |
login(token=settings.TOKEN)
|
33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
def untar_file(source: Path) -> Path:
|
36 |
"""Untar and decompress files which have passed by `make_tarfile`.
|
@@ -51,7 +69,7 @@ def download_database(
|
|
51 |
repo_id: str,
|
52 |
lancedb_file: str = "lancedb.tar.gz",
|
53 |
local_dir: Path = Path.home() / ".cache/argilla_sdk_docs_db",
|
54 |
-
token: str = os.getenv("HF_API_TOKEN")
|
55 |
) -> Path:
|
56 |
"""Helper function to download the database. Will download a compressed lancedb stored
|
57 |
in a Hugging Face repository.
|
@@ -69,18 +87,18 @@ def download_database(
|
|
69 |
"""
|
70 |
lancedb_download = Path(
|
71 |
hf_hub_download(
|
72 |
-
repo_id,
|
73 |
-
lancedb_file,
|
74 |
-
repo_type="dataset",
|
75 |
-
token=token,
|
76 |
-
local_dir=local_dir
|
77 |
)
|
78 |
)
|
79 |
return untar_file(lancedb_download)
|
80 |
|
81 |
|
82 |
# Get the model to create the embeddings
|
83 |
-
model =
|
|
|
|
|
|
|
|
|
84 |
|
85 |
|
86 |
class Database:
|
@@ -90,7 +108,12 @@ class Database:
|
|
90 |
the expected location. Once ready, the only functionality available is
|
91 |
to retrieve the doc chunks to be used as examples for the LLM.
|
92 |
"""
|
|
|
93 |
def __init__(self, settings: Settings) -> None:
|
|
|
|
|
|
|
|
|
94 |
self.settings = settings
|
95 |
self._table: lancedb.table.LanceTable = self.get_table_from_db()
|
96 |
|
@@ -110,39 +133,56 @@ class Database:
|
|
110 |
self.settings.REPO_ID,
|
111 |
lancedb_file=self.settings.LANCEDB_FILE_TAR,
|
112 |
local_dir=self.settings.LOCAL_DIR,
|
113 |
-
token=self.settings.TOKEN
|
114 |
)
|
115 |
|
116 |
db = lancedb.connect(str(lancedb_db_path))
|
117 |
table = db.open_table(self.settings.TABLE_NAME)
|
118 |
return table
|
119 |
|
120 |
-
def retrieve_doc_chunks(
|
121 |
-
|
122 |
-
|
123 |
-
|
|
|
124 |
|
125 |
Args:
|
126 |
-
query
|
127 |
-
limit
|
128 |
-
hard_limit
|
|
|
|
|
|
|
|
|
129 |
|
130 |
Returns:
|
131 |
-
|
132 |
"""
|
133 |
-
#
|
134 |
embedded_query = model.generate_embeddings([query])
|
135 |
field_to_retrieve = "text"
|
136 |
retrieved = (
|
137 |
-
self._table
|
138 |
-
|
139 |
-
|
140 |
-
|
141 |
-
|
142 |
-
.to_list()
|
143 |
)
|
144 |
-
|
145 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
146 |
responses = []
|
147 |
unique_responses = set()
|
148 |
|
@@ -164,8 +204,7 @@ database = Database(settings=settings)
|
|
164 |
|
165 |
|
166 |
def get_client_and_tokenizer(
|
167 |
-
model_id: str = settings.MODEL_ID,
|
168 |
-
tokenizer_id: Optional[str] = None
|
169 |
) -> tuple[InferenceClient, AutoTokenizer]:
|
170 |
"""Obtains the inference client and the tokenizer corresponding to the model.
|
171 |
|
@@ -182,14 +221,9 @@ def get_client_and_tokenizer(
|
|
182 |
tokenizer_id = model_id
|
183 |
|
184 |
client = InferenceClient()
|
185 |
-
base_url = client._resolve_url(
|
186 |
-
model=model_id, task="text-generation"
|
187 |
-
)
|
188 |
# Note: We could move to the AsyncClient
|
189 |
-
client = InferenceClient(
|
190 |
-
model=base_url,
|
191 |
-
token=os.getenv("HF_API_TOKEN")
|
192 |
-
)
|
193 |
|
194 |
tokenizer = AutoTokenizer.from_pretrained(tokenizer_id)
|
195 |
return client, tokenizer
|
@@ -204,7 +238,9 @@ client_kwargs = {
|
|
204 |
"temperature": 0.3,
|
205 |
"top_p": None,
|
206 |
"top_k": None,
|
207 |
-
"stop_sequences": ["<|eot_id|>", "<|end_of_text|>"]
|
|
|
|
|
208 |
"seed": None,
|
209 |
}
|
210 |
|
@@ -313,6 +349,42 @@ def prepare_input(message: str, history: list[tuple[str, str]]) -> str:
|
|
313 |
)[0]
|
314 |
|
315 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
316 |
def chatty(message: str, history: list[tuple[str, str]]) -> Generator[str, None, None]:
|
317 |
"""Main function of the app, contains the interaction with the LLM.
|
318 |
|
@@ -326,28 +398,48 @@ def chatty(message: str, history: list[tuple[str, str]]) -> Generator[str, None,
|
|
326 |
"""
|
327 |
prompt = prepare_input(message, history)
|
328 |
|
329 |
-
|
330 |
-
for token_stream in client.text_generation(prompt=prompt, **client_kwargs):
|
331 |
-
partial_message += token_stream
|
332 |
-
yield partial_message
|
333 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
334 |
|
335 |
|
336 |
if __name__ == "__main__":
|
337 |
-
|
338 |
import gradio as gr
|
339 |
|
340 |
gr.ChatInterface(
|
341 |
chatty,
|
342 |
-
chatbot=gr.Chatbot(height=
|
343 |
-
textbox=gr.Textbox(
|
|
|
|
|
344 |
title="Argilla SDK Chatbot",
|
345 |
description="Ask a question about Argilla SDK",
|
346 |
theme="soft",
|
347 |
examples=[
|
348 |
"How can I connect to an argilla server?",
|
349 |
"How can I access a dataset?",
|
350 |
-
"How can I get the current user?"
|
351 |
],
|
352 |
cache_examples=True,
|
353 |
retry_btn=None,
|
|
|
1 |
+
from typing import Optional, Generator
|
2 |
import os
|
3 |
from pathlib import Path
|
4 |
import tarfile
|
|
|
11 |
from huggingface_hub import InferenceClient, login
|
12 |
from transformers import AutoTokenizer
|
13 |
import gradio as gr
|
14 |
+
import argilla as rg
|
15 |
+
import uuid
|
16 |
|
17 |
|
18 |
@dataclass
|
19 |
class Settings:
|
20 |
+
"""Settings class to store useful variables for the App."""
|
21 |
+
|
22 |
LANCEDB: str = "lancedb"
|
23 |
LANCEDB_FILE_TAR: str = "lancedb.tar.gz"
|
24 |
TOKEN: str = os.getenv("HF_API_TOKEN")
|
|
|
26 |
REPO_ID: str = "plaguss/argilla_sdk_docs_queries"
|
27 |
TABLE_NAME: str = "docs"
|
28 |
MODEL_NAME: str = "plaguss/bge-base-argilla-sdk-matryoshka"
|
29 |
+
DEVICE: str = (
|
30 |
+
"mps"
|
31 |
+
if torch.backends.mps.is_available()
|
32 |
+
else "cuda"
|
33 |
+
if torch.cuda.is_available()
|
34 |
+
else "cpu"
|
35 |
+
)
|
36 |
MODEL_ID: str = "meta-llama/Meta-Llama-3-70B-Instruct"
|
37 |
+
ARGILLA_URL = r"https://plaguss-argilla-sdk-chatbot.hf.space"
|
38 |
+
ARGILLA_API_KEY = os.getenv("ARGILLA_CHATBOT_API_KEY")
|
39 |
+
ARGILLA_DATASET = "chatbot_interactions"
|
40 |
+
|
41 |
|
42 |
settings = Settings()
|
43 |
|
44 |
login(token=settings.TOKEN)
|
45 |
|
46 |
+
client_rg = rg.Argilla(
|
47 |
+
api_url=settings.ARGILLA_URL,
|
48 |
+
api_key=settings.ARGILLA_API_KEY
|
49 |
+
)
|
50 |
+
argilla_dataset = client_rg.datasets(settings.ARGILLA_DATASET)
|
51 |
+
|
52 |
|
53 |
def untar_file(source: Path) -> Path:
|
54 |
"""Untar and decompress files which have passed by `make_tarfile`.
|
|
|
69 |
repo_id: str,
|
70 |
lancedb_file: str = "lancedb.tar.gz",
|
71 |
local_dir: Path = Path.home() / ".cache/argilla_sdk_docs_db",
|
72 |
+
token: str = os.getenv("HF_API_TOKEN"),
|
73 |
) -> Path:
|
74 |
"""Helper function to download the database. Will download a compressed lancedb stored
|
75 |
in a Hugging Face repository.
|
|
|
87 |
"""
|
88 |
lancedb_download = Path(
|
89 |
hf_hub_download(
|
90 |
+
repo_id, lancedb_file, repo_type="dataset", token=token, local_dir=local_dir
|
|
|
|
|
|
|
|
|
91 |
)
|
92 |
)
|
93 |
return untar_file(lancedb_download)
|
94 |
|
95 |
|
96 |
# Get the model to create the embeddings
|
97 |
+
model = (
|
98 |
+
get_registry()
|
99 |
+
.get("sentence-transformers")
|
100 |
+
.create(name=settings.MODEL_NAME, device=settings.DEVICE)
|
101 |
+
)
|
102 |
|
103 |
|
104 |
class Database:
|
|
|
108 |
the expected location. Once ready, the only functionality available is
|
109 |
to retrieve the doc chunks to be used as examples for the LLM.
|
110 |
"""
|
111 |
+
|
112 |
def __init__(self, settings: Settings) -> None:
|
113 |
+
"""
|
114 |
+
Args:
|
115 |
+
settings: Instance of the settings.
|
116 |
+
"""
|
117 |
self.settings = settings
|
118 |
self._table: lancedb.table.LanceTable = self.get_table_from_db()
|
119 |
|
|
|
133 |
self.settings.REPO_ID,
|
134 |
lancedb_file=self.settings.LANCEDB_FILE_TAR,
|
135 |
local_dir=self.settings.LOCAL_DIR,
|
136 |
+
token=self.settings.TOKEN,
|
137 |
)
|
138 |
|
139 |
db = lancedb.connect(str(lancedb_db_path))
|
140 |
table = db.open_table(self.settings.TABLE_NAME)
|
141 |
return table
|
142 |
|
143 |
+
def retrieve_doc_chunks(
|
144 |
+
self, query: str, limit: int = 12, hard_limit: int = 4
|
145 |
+
) -> str:
|
146 |
+
"""Search for similar queries in the database, and return the context to be passed
|
147 |
+
to the LLM.
|
148 |
|
149 |
Args:
|
150 |
+
query: Query from the user.
|
151 |
+
limit: Number of similar items to retrieve. Defaults to 12.
|
152 |
+
hard_limit: Limit of responses to take into account.
|
153 |
+
As we generated repeated questions initially, the database may contain
|
154 |
+
repeated chunks of documents, in the initial `limit` selection, using
|
155 |
+
`hard_limit` we limit to this number the total of unique retrieved chunks.
|
156 |
+
Defaults to 4.
|
157 |
|
158 |
Returns:
|
159 |
+
The context to be used by the model to generate the response.
|
160 |
"""
|
161 |
+
# Embed the query to use our custom model instead of the default one.
|
162 |
embedded_query = model.generate_embeddings([query])
|
163 |
field_to_retrieve = "text"
|
164 |
retrieved = (
|
165 |
+
self._table.search(embedded_query[0])
|
166 |
+
.metric("cosine")
|
167 |
+
.limit(limit)
|
168 |
+
.select([field_to_retrieve]) # Just grab the chunk to use for context
|
169 |
+
.to_list()
|
|
|
170 |
)
|
171 |
+
return self._prepare_context(retrieved, hard_limit)
|
172 |
+
|
173 |
+
@staticmethod
|
174 |
+
def _prepare_context(retrieved: list[dict[str, str]], hard_limit: int) -> str:
|
175 |
+
"""Prepares the examples to be used in the LLM prompt.
|
176 |
+
|
177 |
+
Args:
|
178 |
+
retrieved: The list of retrieved chunks.
|
179 |
+
hard_limit: Max number of doc pieces to return.
|
180 |
+
|
181 |
+
Returns:
|
182 |
+
Context to be used by the LLM.
|
183 |
+
"""
|
184 |
+
# We have repeated questions (up to 4) for a given chunk, so we may get repeated chunks.
|
185 |
+
# Request more than necessary and filter them afterwards
|
186 |
responses = []
|
187 |
unique_responses = set()
|
188 |
|
|
|
204 |
|
205 |
|
206 |
def get_client_and_tokenizer(
|
207 |
+
model_id: str = settings.MODEL_ID, tokenizer_id: Optional[str] = None
|
|
|
208 |
) -> tuple[InferenceClient, AutoTokenizer]:
|
209 |
"""Obtains the inference client and the tokenizer corresponding to the model.
|
210 |
|
|
|
221 |
tokenizer_id = model_id
|
222 |
|
223 |
client = InferenceClient()
|
224 |
+
base_url = client._resolve_url(model=model_id, task="text-generation")
|
|
|
|
|
225 |
# Note: We could move to the AsyncClient
|
226 |
+
client = InferenceClient(model=base_url, token=os.getenv("HF_API_TOKEN"))
|
|
|
|
|
|
|
227 |
|
228 |
tokenizer = AutoTokenizer.from_pretrained(tokenizer_id)
|
229 |
return client, tokenizer
|
|
|
238 |
"temperature": 0.3,
|
239 |
"top_p": None,
|
240 |
"top_k": None,
|
241 |
+
"stop_sequences": ["<|eot_id|>", "<|end_of_text|>"]
|
242 |
+
if settings.MODEL_ID.startswith("meta-llama/Meta-Llama-3")
|
243 |
+
else None,
|
244 |
"seed": None,
|
245 |
}
|
246 |
|
|
|
349 |
)[0]
|
350 |
|
351 |
|
352 |
+
def create_chat_html(history: list[tuple[str, str]]) -> str:
|
353 |
+
"""Helper function to create a conversation in HTML in argilla.
|
354 |
+
|
355 |
+
Args:
|
356 |
+
history: History of messages with the chatbot.
|
357 |
+
|
358 |
+
Returns:
|
359 |
+
HTML formatted conversation.
|
360 |
+
"""
|
361 |
+
chat_html = ""
|
362 |
+
alignments = ["right", "left"]
|
363 |
+
colors = ["#c2e3f7", "#f5f5f5"]
|
364 |
+
|
365 |
+
for turn in history:
|
366 |
+
# Create the HTML message div with inline styles
|
367 |
+
message_html = ""
|
368 |
+
|
369 |
+
# To include message still not answered
|
370 |
+
(user, assistant) = turn
|
371 |
+
if assistant is None:
|
372 |
+
turn = (user, )
|
373 |
+
|
374 |
+
for i, content in enumerate(turn):
|
375 |
+
message_html += f'<div style="display: flex; justify-content: {alignments[i]}; margin: 10px;">'
|
376 |
+
message_html += f'<div style="background-color: {colors[i]}; padding: 10px; border-radius: 10px; max-width: 70%; word-wrap: break-word;">{content}</div>'
|
377 |
+
message_html += "</div>"
|
378 |
+
|
379 |
+
# Add the message to the chat HTML
|
380 |
+
chat_html += message_html
|
381 |
+
|
382 |
+
return chat_html
|
383 |
+
|
384 |
+
|
385 |
+
conv_id = str(uuid.uuid4())
|
386 |
+
|
387 |
+
|
388 |
def chatty(message: str, history: list[tuple[str, str]]) -> Generator[str, None, None]:
|
389 |
"""Main function of the app, contains the interaction with the LLM.
|
390 |
|
|
|
398 |
"""
|
399 |
prompt = prepare_input(message, history)
|
400 |
|
401 |
+
partial_response = ""
|
|
|
|
|
|
|
402 |
|
403 |
+
for token_stream in client.text_generation(prompt=prompt, **client_kwargs):
|
404 |
+
partial_response += token_stream
|
405 |
+
yield partial_response
|
406 |
+
|
407 |
+
global conv_id
|
408 |
+
new_conversation = len(history) == 0
|
409 |
+
if new_conversation:
|
410 |
+
conv_id = str(uuid.uuid4())
|
411 |
+
else:
|
412 |
+
history.append((message, None))
|
413 |
+
|
414 |
+
# Register to argilla dataset
|
415 |
+
argilla_dataset.records.log(
|
416 |
+
[
|
417 |
+
{
|
418 |
+
"instruction": create_chat_html(history) if history else message,
|
419 |
+
"response": partial_response,
|
420 |
+
"conv_id": conv_id,
|
421 |
+
"turn": len(history)
|
422 |
+
},
|
423 |
+
]
|
424 |
+
)
|
425 |
|
426 |
|
427 |
if __name__ == "__main__":
|
|
|
428 |
import gradio as gr
|
429 |
|
430 |
gr.ChatInterface(
|
431 |
chatty,
|
432 |
+
chatbot=gr.Chatbot(height=700),
|
433 |
+
textbox=gr.Textbox(
|
434 |
+
placeholder="Ask me about the new argilla SDK", container=False, scale=7
|
435 |
+
),
|
436 |
title="Argilla SDK Chatbot",
|
437 |
description="Ask a question about Argilla SDK",
|
438 |
theme="soft",
|
439 |
examples=[
|
440 |
"How can I connect to an argilla server?",
|
441 |
"How can I access a dataset?",
|
442 |
+
"How can I get the current user?",
|
443 |
],
|
444 |
cache_examples=True,
|
445 |
retry_btn=None,
|
requirements.txt
CHANGED
@@ -1,3 +1,4 @@
|
|
1 |
torch==2.3.1
|
2 |
lancedb==0.8.2
|
3 |
-
sentence-transformers==3.0.1
|
|
|
|
1 |
torch==2.3.1
|
2 |
lancedb==0.8.2
|
3 |
+
sentence-transformers==3.0.1
|
4 |
+
argilla==2.0.0rc1
|