Release original LaST-Net checkpoint and usage instructions

README.md

@@ -0,0 +1,44 @@
+---
+tags:
+- audio
+- speech-deepfake-detection
+- pytorch
+- fairseq
+base_model: facebook/wav2vec2-xls-r-300m
+---
+
+# LaST-Net
+
+Model weights for **LaST-Net: Length-Aware Layer and Scale-Adaptive Temporal Network for Speech Deepfake Detection**.
+
+**Training and inference code:** [JunXue-tech/LaST-Net v1.0.0](https://github.com/JunXue-tech/LaST-Net/tree/v1.0.0).
+
+Release **v1.0.0** pairs this original checkpoint with the same version tag in the code repository.
+
+## Checkpoint
+
+`best.pt` is the original epoch-52 checkpoint selected by the lowest mean development EER across 1, 2, 4 and 6 seconds on ASVspoof 2019 LA. It includes the fine-tuned XLS-R 300M frontend, LaST-Net backend, optimizer state and original training metadata. The original checkpoint is distributed without conversion.
+
+The matching code uses strict state-dict loading. Model settings are router temperature 0.8, uniform floor 0.05, and layer dropout 0.1. Paired Prediction Consistency is used only during training.
+
+## Usage
+
+Follow the environment setup in the [code repository](https://github.com/JunXue-tech/LaST-Net). From that repository:
+
+```bash
+python download_model.py
+python infer.py example.wav --checkpoint checkpoints/best.pt \
+  --ssl-path /path/to/xlsr2_300m.pt --seconds 6
+```
+
+The model constructor requires the original fairseq-format XLS-R 300M checkpoint, available from the [official XLS-R repository](https://github.com/facebookresearch/fairseq/tree/main/examples/wav2vec/xlsr), before loading the fine-tuned parameters from `best.pt`.
+
+Input audio must be mono at 16 kHz. The supplied inference code evaluates 1–6 second inputs using prefix cropping and repetition of shorter recordings. Higher `bonafide_log_score` values favor bona fide speech. Scores are not calibrated probabilities.
+
+## Evaluation
+
+Duration-averaged EER (%) across 1–6 second inputs: 19LA **1.29**, 21LA **4.98**, 21DF **3.62**, and In-the-Wild **7.58**. Per-duration results and evaluation commands are provided in the code repository. Results depend on the evaluation protocol and preprocessing.
+
+## Authors
+
+Jun Xue, Qichang Li, Yan Li, Daixian Li, Yanzhen Ren, Yihuan Huang, and Zhuolin Yi. School of Cyber Science and Engineering, Wuhan University. Corresponding author: Yanzhen Ren.

best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fee6178d19b398fac9e1b1bfcf13720c4eb5ff722b5b52e4bc698e0de0fa28b
+size 3811985991