Fix Darwin 35B load: CPU offload flag + spaces pkg + persistent cache (#6)

Fix: Move model load inside @spaces.GPU to prevent CPU OOM crash (#4)

PR #2 requirements: Remove llama.cpp, add transformers and full ML stack (#3)

PR #2: Replace llama.cpp with transformers + spaces.GPU + temp grid (#2)

Fix libcudart.so.12: force CPU-only llama-cpp build (#1)

Update README.md

Delete app-backup2.py

Delete app-backup.py

Update app.py

Update app.py

Update index.html

Update requirements.txt

Update app.py

Update app.py

Update requirements.txt

Update app.py

Update index.html

Update requirements.txt

Create app.py

Rename app.py to app-backup2.py

Update app.py

Update app.py

Update app.py

Update index.html

Update app.py

Update app.py

Update index.html

Update app.py

Create app.py

Rename app.py to app-backup.py

Delete app-backup.py

Update README.md

Update index.html

Update app.py

Update requirements.txt

Update index.html

Update index.html

Update app.py

Update index.html

Update app.py

Update app.py

Update index.html

Update app.py

Create app.py

Rename app.py to app-backup.py

Update index.html

Update app.py

Upload 3 files

initial commit