PAI dataset support

[bcostarendon]

• Add make instruction for PAI dataset creation
• Add workflow to ingest PAI dataset into CDS
• Update py dependencies, include usage of physical_ai_av library

Run the sample script scripts/run_cds_on_pai.sh using query: "sunny city"

$ ./scripts/run_cds_on_pai.sh 
Building all Docker images... 
[...]
[+] Building 0.7s (15/15) FINISHED                                                                                                                                                        [...]
[+] Building 0.6s (15/15) FINISHED                                                                                                                                                        [...]
Installing all dependencies from pyproject.toml... 
[...]
✓ CDS CLI installed successfully! 
[...]
Starting services... 
cd deploy/standalone && docker compose -f docker-compose.build.yml up -d
[+] up 11/11
[...]
Preparing dataset using scripts/pai_test_100.yaml... 
[...]
Encoding 100 videos to H264 using 24 workers
[...]
Ingesting PAI dataset into collection... 
[...]
│   Processed files: 100/100 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:01:31 0:00:00 ││    Status code 200: 100/100 100% │
[22:32:55] 🚀 Finished processing job queue!                                                                                                                                                client.py:723
           Processed 100 files successfully                                                                                                                                                 [...]
04c5a3f4-7d2f-499a-ba05-b57194594735
d81becb2-5333-4808-a01e-14ce2f2f1027
259b1529-0d28-4ea5-b822-d8691e91f062
95ad1d0a-7713-4899-91dd-08176123823f
275c9acf-8cb4-46ca-8335-5a7b80a501fe