Hugging Face's logo

SpeechAntiSpoofingBenchmarks
/
Codecfake-W2V2-AASIST-CoTrain

PyTorch
audio
anti-spoofing
deepfake-detection
speech-anti-spoofing-arena
Model card Files
xet
 Community
Codecfake-W2V2-AASIST-CoTrain / implementation-notes.md
korallll's picture
korallll
Wrapper code (speech-spoof-bench)
26da0bd verified
preview code
|
Raw History Blame Contribute Delete
5.05 kB

Codecfake-W2V2-AASIST-CoTrain β€” implementation notes

Arena wrapper for the co-trained W2V2-AASIST countermeasure of The Codecfake Dataset and Countermeasures for the Universally Detection of Deepfake Audio (Xie et al., arXiv 2405.04880), built 2026-11-01 under speech-spoof-bench/docs/plans/2026-11-01-nine-model-batch-arena-submission.md (model #2). Sibling model #8 (codec-trained) lives in ../Codecfake-W2V2-AASIST-Codec/ with an identical copy of _net.py.

Source

  • Code: https://github.com/xieyuankun/Codecfake @ 72c32e2840eef95fbeade349425fa9c5391fa11e β€” no LICENSE (so no LICENSE.source; weights not re-hosted).
  • Checkpoint: pretrained_model/cotrain_w2v2aasist_CSAM/anti-spoofing_feat_model.pt (1.8 MB, sha256 86876e2c…c644a), trained on ASVspoof2019 LA train + Codecfake train with the CSAM sampler. It is torch.save(model) of the whole W2VAASIST module. Pickle globals inspected with pickletools first: only model.{W2VAASIST,GraphAttentionLayer,HtrgGraphAttentionLayer,GraphPool,Residual_block}, torch.nn.modules.*, torch storage rebuilders, OrderedDict, set.
  • Frozen front-end: HF facebook/wav2vec2-xls-r-300m @ 1a640f32ac3e39899438a2931f9924c02f080a54 (main).

Weight acquisition (_net.head_state)

First load(): download from the pinned raw URL https://raw.githubusercontent.com/xieyuankun/Codecfake/72c32e2…/pretrained_model/<sub>/anti-spoofing_feat_model.pt into ../_ckpt_dl/xieyuankun__Codecfake/<rev>/…, check sha256, unpickle with an allow-list unpickler that maps the source's model.* classes to a bare nn.Module stub (no upstream code imported or run), take its state_dict(), save head_<sub>.safetensors in the model dir (gitignored). Later loads read the safetensors. Pickled hyper-params checked against the port: GAT temperatures 2/2/100Γ—4, pool ratios 0.5Γ—6; module was saved in eval mode. Strict load: 245 keys, 0 missing / 0 unexpected.

Port (_net.py)

  • W2VAASIST + layers ported verbatim (prints/unused imports removed). Residual_block: the source computes bn1β†’selu for non-first blocks then applies conv1 to x (result discarded) β€” dead code dropped, bn1 kept.
  • Pipeline as generate_score.py: pad_dataset (first 64 600 samples, int(cut/len)+1 tiles) β†’ Wav2Vec2FeatureExtractor normalisation (do_normalize; reproduced per row in torch, max |Ξ”| 5e-7) β†’ XLS-R hidden_states[5] β†’ unsqueeze(1).transpose(2,3) β†’ (B,1,1024,201) β†’ head.
  • XLS-R truncated to 5 transformer layers (XLSRLayer5: feature encoder β†’ projection β†’ pos-conv β†’ layers 0–4, no final LayerNorm). hidden_states[5] of the stable-layer-norm encoder is the input of layer 5 = output of layer 4 before encoder.layer_norm; verified bit-identical (torch.equal, max |Ξ”| 0.0) against the full 24-layer HF model (eager attention) on 3 random clips. Load asserts 0 missing keys and only layers 5–23 + pre-training heads unexpected.
  • Attention: eager (as transformers 4.36 used by the authors).
  • TF32 disabled in load() (cuDNN TF32 convs otherwise moved scores by ~2e-3 between batch compositions).

Score

Class 0 = bona fide (source writes softmax(out)[:, 0]). Arena score = logits[:,0] - logits[:,1] = logit of the authors' probability (monotone, not saturated). Higher = bona fide.

Params

76,554,634 = XLS-R used part 76,107,392 + head 447,242 (meta.yaml 76.554634).

Reproduction vs the authors' score files (parity.py; authors' result/*.txt = softmax[:,0])

set (full) authors EER ours EER Pearson (logit) Spearman median |Ξ” logit|
ASVspoof2019_LA eval (71 237) 0.312 0.312 0.99994 0.999995 0.002
InTheWild (31 779) 4.681 4.562 0.9986 0.9995 0.046

Paper: ITW 4.68, 19LA 0.31 β€” matched. ITW residual explained (parity_float_vs_flac.py, 512 clips): scoring the original float wavs (release_in_the_wild/) gives median |Ξ”| 0.0035 vs the authors (0.043 from the arena's 16-bit FLAC copy); i.e. the port is exact, the gap is the dataset packaging.

Throughput / TRT (2026-11-01, GPUs shared with other agents' jobs)

PyTorch sweep: RTX 4070 Ti SUPER 288 utt/s @32, RTX 5060 Ti 104 utt/s @16 (< 300 on the 5060 β†’ TRT built). TRT (trt_codecfake_w2v2_aasist_cotrain.py, DPDA-GA-TCM pattern, sidecar trt_codecfake-w2v2-aasist-cotrain.json):

card precision best batch utt/s build Spearman (64 ITW) J-SPAW_LA EER PT β†’ TRT Spearman vs PT
RTX 4070 Ti SUPER FP32 (FP16 0.9992 < 0.9999) 32 396 0.99991 0.0626 β†’ 0.0626 0.99999
RTX 5060 Ti FP32 (FP16 0.9991) 16 306 0.99977 0.0626 β†’ 0.0626 0.99999
Both accepted. Note J-SPAW_LA EER is ~0 for this model, so Ξ”EER is a weak check; Spearman carries it.

Files

run_sets.sh (PyTorch full-set runs, both dirs), trt_export_both.sh, trt_jspaw.sh (both dirs), parity.py, parity_float_vs_flac.py. results_pt/{J-SPAW_LA,InTheWild,ASVspoof2019_LA}, results_trt2/3/J-SPAW_LA.