Aller au contenu
ModelPrésentéExterne
Browse files

Citympc Txvit

Stripped Lightning checkpoints for the CityMPC TXViT model from the NeurIPS 2026 paper. One checkpoint per city, selected by the seed-sweep stability filter (Kendall weighting σpresence < 1e-3 + good-verdict on...

Sourcehuggingface.co/neurips2026citympc/citympc-txvit/tree/3c7656d7066307d87ab76f69ef2472a61da2ebb4
TypeOtherVersionRevision 3c7656dBClasse

Source du modèle

Type
Other
Taille du fichier
60,6 MB
Version
vRevision 3c7656d
Mis à jour
10 oct. 2026
Téléchargements
0
Source
Hugging Face

Description de la source

CityMPC — TX_ViT pre-trained checkpoints (5 cities)

Stripped Lightning checkpoints for the CityMPC TX_ViT model from the NeurIPS 2026 paper. One checkpoint per city, selected by the seed-sweep stability filter (Kendall weighting σ_presence < 1e-3 + good-verdict on test-set outlier rule), then ranked by lowest Path Loss MAE / best Presence F1.

Per-city metrics (test set)

CityPath Loss MAE (dB)ToF MAE (ns)Avg Delay MAE (ns)Presence F1Seed
tx_austin1.1075.1816.360.8942024
tx_dallas1.4515.299.480.879123
tx_denver1.9763.9810.150.863123
tx_fortworth1.7364.0711.130.881123
tx_newyork2.4818.5919.960.87442

How to use

These checkpoints are stripped — inference only, not training-resumable. After cloning the CityMPC repo:

# download a single city
hf download neurips2026citympc/citympc-txvit \
  outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt \
  outputs/tx_austin/city_10_austin_3p5_s/final/config.yaml \
  --local-dir .

# evaluate
python scripts/analyze.py \
  --ckpt outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt

Or open notebooks/02_analyze_trained_model.ipynb and paste the downloaded run path into TXVIT_RUN_DIR.

File layout

outputs/
  tx_austin/city_10_austin_3p5_s/final/{ckpts/last.ckpt, config.yaml}
  tx_dallas/city_8_dallas_3p5_s/final/{ckpts/last.ckpt, config.yaml}
  tx_denver/city_18_denver_3p5_s/final/{ckpts/last.ckpt, config.yaml}
  tx_fortworth/city_12_fortworth_3p5_s/final/{ckpts/last.ckpt, config.yaml}
  tx_newyork/city_0_newyork_3p5_s/final/{ckpts/last.ckpt, config.yaml}

Integrity (SHA-256)

FileSizeSHA-256
outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt60.6 MB2c4b31591eb347caab9acf11623dcc2d8a184515fbf8c4186b4958a2c5fe95a3
outputs/tx_austin/city_10_austin_3p5_s/final/config.yaml0.0 MBf3e75a17d3d7726cfd01a75df635b26aac3ca54edd386af2b604eaaa7e3057c1
outputs/tx_dallas/city_8_dallas_3p5_s/final/ckpts/last.ckpt60.6 MB397327907665bb7abb70335a0cd581677a8b6b55e9d3a4fbe95092d7fd9d12ad
outputs/tx_dallas/city_8_dallas_3p5_s/final/config.yaml0.0 MB9b60b5483f2580737052ba8ed6e4d89fe86dd52161095b50a14afcbe7256a88d
outputs/tx_denver/city_18_denver_3p5_s/final/ckpts/last.ckpt60.6 MB13e496219b7210f9092996023f16ef0aed53b066ac46b322cf8b664085706dee
outputs/tx_denver/city_18_denver_3p5_s/final/config.yaml0.0 MBb03ea4570bbd423a719f91919293ed23617d48476476f7f398fb1ab0d3f51ea0
outputs/tx_fortworth/city_12_fortworth_3p5_s/final/ckpts/last.ckpt60.6 MB6f6ce2e43b025edf422de5b94edc8c476dbc4d604137d487f00cf494c88f9317
outputs/tx_fortworth/city_12_fortworth_3p5_s/final/config.yaml0.0 MBd7d124c9ff9d28c4ac8c7da2c3273863a4466e567480fae382593b5c9bd32e34
outputs/tx_newyork/city_0_newyork_3p5_s/final/ckpts/last.ckpt60.6 MB98de5771b763445dd04d8c0bdf070777492710ffb0e65fa7807e599871b00e74
outputs/tx_newyork/city_0_newyork_3p5_s/final/config.yaml0.0 MBc33baa28a3d6ae02626a46cd64f3161378cb53e4d93968528cc46091c861065d

Notes

  • Architecture: TX_ViT conditional VAE (CityMPC). See the paper for details.
  • Trained on the CityMPC dataset.
  • Lightning state stripped (optimizer / scheduler / callback) — checkpoints are not training-resumable.
  • σ-gate: every shipped run had weighting/sigma/presence < 1e-3 at its last logged step.

Sources

1 source
Hugging FacePrincipalVersion 3c7656d7066307d87ab76f69ef2472a61da2ebb4
huggingface.co/neurips2026citympc/citympc-txvit/tree/3c7656d7066307d87ab76f69ef2472a61da2ebb4

Vérifié 15 sept.

Download files

Artefacts du modèle

5 artefacts

outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt

ckpt · 60,6 MB · SHA-256 2c4b31591eb3…95a3 · Hugging Face

Download file

outputs/tx_dallas/city_8_dallas_3p5_s/final/ckpts/last.ckpt

ckpt · 60,6 MB · SHA-256 397327907665…12ad · Hugging Face

Download file

outputs/tx_denver/city_18_denver_3p5_s/final/ckpts/last.ckpt

ckpt · 60,6 MB · SHA-256 13e496219b72…6dee · Hugging Face

Download file

outputs/tx_fortworth/city_12_fortworth_3p5_s/final/ckpts/last.ckpt

ckpt · 60,6 MB · SHA-256 6f6ce2e43b02…9317 · Hugging Face

Download file

outputs/tx_newyork/city_0_newyork_3p5_s/final/ckpts/last.ckpt

ckpt · 60,6 MB · SHA-256 98de5771b763…0e74 · Hugging Face

Download file

Extraits de sources

2 extraits
Readme31 juil.huggingface.co/neurips2026citympc/citympc-txvit

--- license: cc-by-4.0 library_name: pytorch tags: - channel-modeling - ray-tracing - cir-prediction - vae - cvae - lightning - sionna --- # CityMPC — TX_ViT pre-trained checkpoints (5 cities) Stripped Lightning checkpoints for the **CityMPC TX_ViT** model from the NeurIPS 2026 paper. One checkpoint per city, selected by the seed-sweep stability filter (Kendall weighting σ_presence < 1e-3 + good-verdict on test-set outlier rule), then ranked by lowest Path Loss MAE / best Presence F1. ## Per-city metrics (test set) | City | Path Loss MAE (dB) | ToF MAE (ns) | Avg Delay MAE (ns) | Presence F1 | Seed | |---|---|---|---|---|---| | tx_austin | 1.107 | 5.18 | 16.36 | 0.894 | 2024 | | tx_dallas | 1.451 | 5.29 | 9.48 | 0.879 | 123 | | tx_denver | 1.976 | 3.98 | 10.15 | 0.863 | 123 | | tx_fortworth | 1.736 | 4.07 | 11.13 | 0.881 | 123 | | tx_newyork | 2.481 | 8.59 | 19.96 | 0.874 | 42 | ## How to use These checkpoints are stripped — **inference only**, not training-resumable. After cloning the [CityMPC repo](https://github.com/neurips2026citympc/CityMPC): ```bash # download a single city hf download neurips2026citympc/citympc-txvit \ outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt \ outputs/tx_austin/city_10_austin_3p5_s/final/config.yaml \ --local-dir . # evaluate python scripts/analyze.py \ --ckpt outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt ``` Or open `notebooks/02_analyze_trained_model.ipynb` and paste the downloaded run path into `TXVIT_RUN_DIR`. ## File layout ``` outputs/ tx_austin/city_10_austin_3p5_s/final/{ckpts/last.ckpt, config.yaml} tx_dallas/city_8_dallas_3p5_s/final/{ckpts/last.ckpt, config.yaml} tx_denver/city_18_denver_3p5_s/final/{ckpts/last.ckpt, config.yaml} tx_fortworth/city_12_fortworth_3p5_s/final/{ckpts/last.ckpt, config.yaml} tx_newyork/city_0_newyork_3p5_s/final/{ckpts/last.ckpt, config.yaml} ``` ## Integrity (SHA-256) | File | Size | SHA-256 | |---|---|---| | `outputs/tx_austin/city_10_austin_3p5_s/final/ckpts/last.ckpt` | 60.6 MB | `2c4b31591eb347caab9acf11623dcc2d8a184515fbf8c4186b4958a2c5fe95a3` | | `outputs/tx_austin/city_10_austin_3p5_s/final/config.yaml` | 0.0 MB | `f3e75a17d3d7726cfd01a75df635b26aac3ca54edd386af2b604eaaa7e3057c1` | | `outputs/tx_dallas/city_8_dallas_3p5_s/final/ckpts/last.ckpt` | 60.6 MB | `397327907665bb7abb70335a0cd581677a8b6b55e9d3a4fbe95092d7fd9d12ad`...

Note31 juil.huggingface.co/neurips2026citympc/citympc-txvit

Source context: 0 downloads · 0 likes · Library pytorch · Repo neurips2026citympc/citympc-txvit