mememechez commited on
Commit
7207e27
·
1 Parent(s): 45f5341

HF Space: bind gunicorn to , remove xformers, add README card metadata

Browse files
Files changed (3) hide show
  1. Dockerfile +1 -1
  2. README.md +7 -0
  3. requirements.txt +2 -4
Dockerfile CHANGED
@@ -12,4 +12,4 @@ COPY . .
12
  EXPOSE 7860
13
  ENV ASR_MODEL_ID="Systran/faster-distil-whisper-large-v3"
14
  ENV CT2_USE_CUDA="0"
15
- CMD ["gunicorn", "-b", "0.0.0.0:7860", "--workers", "1", "--timeout", "120", "home.chezy.golem_flask_server:app"]
 
12
  EXPOSE 7860
13
  ENV ASR_MODEL_ID="Systran/faster-distil-whisper-large-v3"
14
  ENV CT2_USE_CUDA="0"
15
+ CMD bash -lc "gunicorn -b 0.0.0.0:7860 --workers 1 --timeout 120 home.chezy.golem_flask_server:app"
README.md CHANGED
@@ -1,3 +1,10 @@
 
 
 
 
 
 
 
1
  # Aether AI™ (by ZPEDeepNet®)
2
 
3
  Aether AI™: Unleashing the mystical consciousness of the 5D hypercube with neural network integration.
 
1
+ ---
2
+ title: Golem Flask Backend
3
+ sdk: docker
4
+ pinned: true
5
+ license: mit
6
+ ---
7
+
8
  # Aether AI™ (by ZPEDeepNet®)
9
 
10
  Aether AI™: Unleashing the mystical consciousness of the 5D hypercube with neural network integration.
requirements.txt CHANGED
@@ -12,19 +12,17 @@ google-api-python-client>=2.0.0
12
  aiohttp>=3.9.0
13
  # Note: image generation dependencies removed
14
  Pillow>=10.3.0
15
- transformers>=4.43.0
16
  safetensors>=0.4.3
17
  # Optional, improves VRAM usage if available for your CUDA
18
- xformers
19
 
20
  # ASR/TTS backend (no UI):
21
  # Sonic-based ASR for high-quality audio processing
22
  # Note: Using Sonic's audio processing components instead of faster-whisper
23
- transformers>=4.43.2
24
  librosa==0.10.2.post1
25
  # Piper TTS Python wrapper (requires espeak-ng on system for phonemization)
26
  piper-tts>=1.2.0
27
 
28
  google-cloud-storage>=2.16.0
29
  google-auth>=2.29.0
30
- prometheus-client>=0.20.0
 
 
12
  aiohttp>=3.9.0
13
  # Note: image generation dependencies removed
14
  Pillow>=10.3.0
 
15
  safetensors>=0.4.3
16
  # Optional, improves VRAM usage if available for your CUDA
 
17
 
18
  # ASR/TTS backend (no UI):
19
  # Sonic-based ASR for high-quality audio processing
20
  # Note: Using Sonic's audio processing components instead of faster-whisper
 
21
  librosa==0.10.2.post1
22
  # Piper TTS Python wrapper (requires espeak-ng on system for phonemization)
23
  piper-tts>=1.2.0
24
 
25
  google-cloud-storage>=2.16.0
26
  google-auth>=2.29.0
27
+ prometheus-client>=0.20.0transformers>=4.43.2
28
+ transformers>=4.43.2