Spaces:
Sleeping
Sleeping
Ajout
Browse files- .env +4 -0
- .gitignore +2 -0
- app.py +66 -0
- requirements.txt +170 -0
- s3_utils.py +66 -0
.env
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
QDRANT_URL=https://006817a4-0b45-4db8-a4e5-1f916808e19b.us-east4-0.gcp.cloud.qdrant.io:6333
|
2 |
+
QDRANT_KEY=d9MUWQOMUA7JGBdpoFRtmou-h4Pf-e9uCrr0jMtUfAvvk4osqL_JtA
|
3 |
+
AWS_ACCESS_KEY_ID=AKIAWOUASMWP5DM6RZG2
|
4 |
+
AWS_SECRET_ACCESS_KEY=HfD73+MKijEgNlVRAkTEgRuNeivyFeYdrtLUqOmq
|
.gitignore
ADDED
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
1 |
+
# Ignore le fichier .env
|
2 |
+
.env
|
app.py
ADDED
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import gradio as gr
|
3 |
+
from qdrant_client import QdrantClient
|
4 |
+
from transformers import ClapModel, ClapProcessor
|
5 |
+
from dotenv import load_dotenv
|
6 |
+
import requests
|
7 |
+
|
8 |
+
# Charger les variables d'environnement à partir du fichier .env
|
9 |
+
load_dotenv()
|
10 |
+
|
11 |
+
# Récupérer les variables d'environnement
|
12 |
+
QDRANT_URL = os.getenv('QDRANT_URL')
|
13 |
+
QDRANT_KEY = os.getenv('QDRANT_KEY')
|
14 |
+
|
15 |
+
# Vérifier les valeurs récupérées
|
16 |
+
print(f"QDRANT_URL: {QDRANT_URL}")
|
17 |
+
print(f"QDRANT_KEY: {QDRANT_KEY}")
|
18 |
+
|
19 |
+
try:
|
20 |
+
# Tester la connexion à l'URL de Qdrant
|
21 |
+
response = requests.get(QDRANT_URL)
|
22 |
+
print(f"Test de la connexion à Qdrant: {response.status_code}")
|
23 |
+
|
24 |
+
# Vérifier que les variables sont correctement récupérées
|
25 |
+
if not QDRANT_URL or not QDRANT_KEY:
|
26 |
+
raise ValueError("Les variables d'environnement QDRANT_URL ou QDRANT_KEY ne sont pas définies")
|
27 |
+
|
28 |
+
# Connexion au client Qdrant
|
29 |
+
client = QdrantClient(QDRANT_URL, api_key=QDRANT_KEY)
|
30 |
+
print("[INFO] Client created...")
|
31 |
+
|
32 |
+
# Chargement du modèle
|
33 |
+
print("[INFO] Loading the model...")
|
34 |
+
model_name = "laion/larger_clap_general"
|
35 |
+
model = ClapModel.from_pretrained(model_name)
|
36 |
+
processor = ClapProcessor.from_pretrained(model_name)
|
37 |
+
|
38 |
+
# Interface Gradio
|
39 |
+
max_results = 10
|
40 |
+
|
41 |
+
def sound_search(query):
|
42 |
+
text_inputs = processor(text=query, return_tensors="pt")
|
43 |
+
text_embed = model.get_text_features(**text_inputs)[0]
|
44 |
+
|
45 |
+
hits = client.search(
|
46 |
+
collection_name="demo_spaces_db",
|
47 |
+
query_vector=text_embed,
|
48 |
+
limit=max_results,
|
49 |
+
)
|
50 |
+
return [
|
51 |
+
gr.Audio(
|
52 |
+
hit.payload['audio_path'],
|
53 |
+
label=f"style: {hit.payload['style']} -- score: {hit.score}")
|
54 |
+
for hit in hits
|
55 |
+
]
|
56 |
+
|
57 |
+
with gr.Blocks() as demo:
|
58 |
+
gr.Markdown("# Sound search database")
|
59 |
+
inp = gr.Textbox(placeholder="What sound are you looking for ?")
|
60 |
+
out = [gr.Audio(label=f"{x}") for x in range(max_results)]
|
61 |
+
inp.change(sound_search, inp, out)
|
62 |
+
|
63 |
+
demo.launch()
|
64 |
+
|
65 |
+
except Exception as e:
|
66 |
+
print(f"[ERROR] Failed to create Qdrant client: {e}")
|
requirements.txt
ADDED
@@ -0,0 +1,170 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
aioboto3==12.3.0
|
2 |
+
aiobotocore==2.11.2
|
3 |
+
aiofiles==23.2.1
|
4 |
+
aiohttp==3.9.3
|
5 |
+
aioitertools==0.11.0
|
6 |
+
aiosignal==1.3.1
|
7 |
+
altair==5.3.0
|
8 |
+
annotated-types==0.6.0
|
9 |
+
antlr4-python3-runtime==4.9.3
|
10 |
+
anyio==4.3.0
|
11 |
+
appdirs==1.4.4
|
12 |
+
async-timeout==4.0.3
|
13 |
+
attrs==23.2.0
|
14 |
+
audioread==3.0.1
|
15 |
+
boto3==1.34.34
|
16 |
+
botocore==1.34.34
|
17 |
+
braceexpand==0.1.7
|
18 |
+
certifi==2024.2.2
|
19 |
+
cffi==1.16.0
|
20 |
+
charset-normalizer==3.3.2
|
21 |
+
click==8.1.7
|
22 |
+
cloudpickle==3.0.0
|
23 |
+
contourpy==1.2.1
|
24 |
+
cycler==0.12.1
|
25 |
+
datasets==2.19.1
|
26 |
+
decorator==5.1.1
|
27 |
+
demucs==4.0.1
|
28 |
+
dill==0.3.8
|
29 |
+
diskcache==5.6.3
|
30 |
+
docker-pycreds==0.4.0
|
31 |
+
dora_search==0.1.12
|
32 |
+
einops==0.7.0
|
33 |
+
essentia==2.1b6.dev1110
|
34 |
+
exceptiongroup==1.2.0
|
35 |
+
fastapi==0.110.1
|
36 |
+
ffmpy==0.3.2
|
37 |
+
filelock==3.13.3
|
38 |
+
filetype==1.2.0
|
39 |
+
fonttools==4.51.0
|
40 |
+
frozenlist==1.4.1
|
41 |
+
fsspec==2024.3.1
|
42 |
+
ftfy==6.2.0
|
43 |
+
gitdb==4.0.11
|
44 |
+
GitPython==3.1.43
|
45 |
+
gradio==4.26.0
|
46 |
+
gradio_client==0.15.1
|
47 |
+
grpcio==1.62.1
|
48 |
+
grpcio-tools==1.62.1
|
49 |
+
h11==0.14.0
|
50 |
+
h2==4.1.0
|
51 |
+
h5py==3.10.0
|
52 |
+
hpack==4.0.0
|
53 |
+
httpcore==1.0.5
|
54 |
+
httpx==0.27.0
|
55 |
+
huggingface-hub==0.22.2
|
56 |
+
hyperframe==6.0.1
|
57 |
+
idna==3.6
|
58 |
+
importlib_resources==6.4.0
|
59 |
+
Jinja2==3.1.3
|
60 |
+
jmespath==1.0.1
|
61 |
+
joblib==1.3.2
|
62 |
+
jsonschema==4.21.1
|
63 |
+
jsonschema-specifications==2023.12.1
|
64 |
+
julius==0.2.7
|
65 |
+
kiwisolver==1.4.5
|
66 |
+
laion-clap==1.1.4
|
67 |
+
lameenc==1.7.0
|
68 |
+
lazy_loader==0.4
|
69 |
+
librosa==0.10.1
|
70 |
+
llvmlite==0.42.0
|
71 |
+
markdown-it-py==3.0.0
|
72 |
+
MarkupSafe==2.1.5
|
73 |
+
matplotlib==3.8.4
|
74 |
+
mdurl==0.1.2
|
75 |
+
miditoolkit==1.0.1
|
76 |
+
mido==1.3.2
|
77 |
+
mpmath==1.3.0
|
78 |
+
msgpack==1.0.8
|
79 |
+
multidict==6.0.5
|
80 |
+
multiprocess==0.70.16
|
81 |
+
networkx==3.3
|
82 |
+
numba==0.59.1
|
83 |
+
numpy==1.23.5
|
84 |
+
nvidia-cublas-cu12==12.1.3.1
|
85 |
+
nvidia-cuda-cupti-cu12==12.1.105
|
86 |
+
nvidia-cuda-nvrtc-cu12==12.1.105
|
87 |
+
nvidia-cuda-runtime-cu12==12.1.105
|
88 |
+
nvidia-cudnn-cu12==8.9.2.26
|
89 |
+
nvidia-cufft-cu12==11.0.2.54
|
90 |
+
nvidia-curand-cu12==10.3.2.106
|
91 |
+
nvidia-cusolver-cu12==11.4.5.107
|
92 |
+
nvidia-cusparse-cu12==12.1.0.106
|
93 |
+
nvidia-nccl-cu12==2.19.3
|
94 |
+
nvidia-nvjitlink-cu12==12.4.127
|
95 |
+
nvidia-nvtx-cu12==12.1.105
|
96 |
+
omegaconf==2.3.0
|
97 |
+
openunmix==1.2.1
|
98 |
+
orjson==3.10.0
|
99 |
+
packaging==23.2
|
100 |
+
pandas==2.2.1
|
101 |
+
pillow==10.3.0
|
102 |
+
platformdirs==4.2.0
|
103 |
+
pooch==1.8.1
|
104 |
+
portalocker==2.8.2
|
105 |
+
progressbar==2.5
|
106 |
+
protobuf==4.25.3
|
107 |
+
psutil==5.9.8
|
108 |
+
pyarrow==16.1.0
|
109 |
+
pyarrow-hotfix==0.6
|
110 |
+
pycparser==2.22
|
111 |
+
pydantic==2.6.4
|
112 |
+
pydantic_core==2.16.3
|
113 |
+
pydub==0.25.1
|
114 |
+
Pygments==2.17.2
|
115 |
+
pyparsing==3.1.2
|
116 |
+
python-dateutil==2.9.0.post0
|
117 |
+
python-dotenv==1.0.1
|
118 |
+
python-multipart==0.0.9
|
119 |
+
pytz==2024.1
|
120 |
+
PyYAML==6.0.1
|
121 |
+
qdrant-client==1.8.2
|
122 |
+
referencing==0.34.0
|
123 |
+
regex==2023.12.25
|
124 |
+
requests==2.31.0
|
125 |
+
retrying==1.3.4
|
126 |
+
rich==13.7.1
|
127 |
+
rpds-py==0.18.0
|
128 |
+
ruff==0.3.5
|
129 |
+
s3transfer==0.10.1
|
130 |
+
safetensors==0.4.2
|
131 |
+
scikit-learn==1.4.1.post1
|
132 |
+
scipy==1.13.0
|
133 |
+
semantic-version==2.10.0
|
134 |
+
sentry-sdk==1.44.1
|
135 |
+
setproctitle==1.3.3
|
136 |
+
sf_segmenter==0.0.2
|
137 |
+
shellingham==1.5.4
|
138 |
+
six==1.16.0
|
139 |
+
smmap==5.0.1
|
140 |
+
sniffio==1.3.1
|
141 |
+
soundfile==0.12.1
|
142 |
+
soxr==0.3.7
|
143 |
+
starlette==0.37.2
|
144 |
+
submitit==1.5.1
|
145 |
+
sympy==1.12
|
146 |
+
threadpoolctl==3.4.0
|
147 |
+
tokenizers==0.13.3
|
148 |
+
tomlkit==0.12.0
|
149 |
+
toolz==0.12.1
|
150 |
+
torch==2.2.2
|
151 |
+
torchaudio==2.2.2
|
152 |
+
torchlibrosa==0.1.0
|
153 |
+
torchvision==0.17.2
|
154 |
+
tqdm==4.66.2
|
155 |
+
transformers==4.30.0
|
156 |
+
treetable==0.2.5
|
157 |
+
triton==2.2.0
|
158 |
+
typer==0.12.2
|
159 |
+
typing_extensions==4.11.0
|
160 |
+
tzdata==2024.1
|
161 |
+
urllib3==2.0.7
|
162 |
+
uvicorn==0.29.0
|
163 |
+
wandb==0.16.6
|
164 |
+
wcwidth==0.2.13
|
165 |
+
webdataset==0.2.86
|
166 |
+
websockets==11.0.3
|
167 |
+
wget==3.2
|
168 |
+
wrapt==1.16.0
|
169 |
+
xxhash==3.4.1
|
170 |
+
yarl==1.9.4
|
s3_utils.py
ADDED
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import hashlib
|
2 |
+
from enum import Enum
|
3 |
+
|
4 |
+
import boto3
|
5 |
+
from botocore.client import BaseClient
|
6 |
+
|
7 |
+
|
8 |
+
# S3 HANDLING ######################################################################################
|
9 |
+
def get_md5(fpath):
|
10 |
+
with open(fpath, "rb") as f:
|
11 |
+
file_hash = hashlib.md5()
|
12 |
+
while chunk := f.read(8192):
|
13 |
+
file_hash.update(chunk)
|
14 |
+
return file_hash.hexdigest()
|
15 |
+
|
16 |
+
|
17 |
+
def upload_file_to_bucket(s3_client, file_obj, bucket, s3key):
|
18 |
+
"""Upload a file to an S3 bucket
|
19 |
+
:param file_obj: File to upload
|
20 |
+
:param bucket: Bucket to upload to
|
21 |
+
:param s3key: s3key
|
22 |
+
:param object_name: S3 object name. If not specified then file_name is used
|
23 |
+
:return: True if file was uploaded, else False
|
24 |
+
"""
|
25 |
+
# Upload the file
|
26 |
+
return s3_client.upload_fileobj(
|
27 |
+
file_obj, bucket, s3key,
|
28 |
+
ExtraArgs={"ACL": "public-read", "ContentType": "Content-Type: audio/mpeg"}
|
29 |
+
)
|
30 |
+
|
31 |
+
|
32 |
+
def s3_auth(aws_access_key_id, aws_secret_access_key, region_name) -> BaseClient:
|
33 |
+
s3 = boto3.client(
|
34 |
+
service_name='s3',
|
35 |
+
aws_access_key_id=aws_access_key_id,
|
36 |
+
aws_secret_access_key=aws_secret_access_key,
|
37 |
+
region_name=region_name
|
38 |
+
)
|
39 |
+
return s3
|
40 |
+
|
41 |
+
|
42 |
+
def get_list_of_buckets(s3: BaseClient):
|
43 |
+
response = s3.list_buckets()
|
44 |
+
buckets = {}
|
45 |
+
|
46 |
+
for buckets in response['Buckets']:
|
47 |
+
buckets[response['Name']] = response['Name']
|
48 |
+
|
49 |
+
BucketName = Enum('BucketName', buckets)
|
50 |
+
return BucketName
|
51 |
+
|
52 |
+
|
53 |
+
if __name__ == '__main__':
|
54 |
+
import os
|
55 |
+
|
56 |
+
AWS_ACCESS_KEY_ID = os.environ['AWS_ACCESS_KEY_ID']
|
57 |
+
AWS_SECRET_ACCESS_KEY = os.environ['AWS_SECRET_ACCESS_KEY']
|
58 |
+
S3_BUCKET = "synthia-research"
|
59 |
+
S3_FOLDER = "huggingface_spaces_demo"
|
60 |
+
AWS_REGION = "eu-west-3"
|
61 |
+
|
62 |
+
s3 = s3_auth(AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_REGION)
|
63 |
+
print(s3.list_buckets())
|
64 |
+
|
65 |
+
s3key = f'{S3_FOLDER}/015.WAV'
|
66 |
+
#print(upload_file_to_bucket(s3, file_obj, S3_BUCKET, s3key))
|