Hugging Face's logo

microsoft
/
unispeech-sat-base-plus-sv

Transformers PyTorch English unispeech-sat audio-xvector speech Inference Endpoints 4 papers
Model card Files Community
1
speech-test commited on
Commit
64191a1
•
1 Parent(s): 3f5e6b0

Upload

Browse files
Files changed (2) hide show
  1. config.json +3 -1
  2. pytorch_model.bin +1 -1
config.json CHANGED
@@ -2483,8 +2483,10 @@
2483
  "layer_norm_eps": 1e-05,
2484
  "layerdrop": 0.1,
2485
  "mask_feature_length": 10,
 
2486
  "mask_feature_prob": 0.0,
2487
  "mask_time_length": 10,
 
2488
  "mask_time_prob": 0.05,
2489
  "model_type": "unispeech-sat",
2490
  "num_attention_heads": 12,
@@ -2521,7 +2523,7 @@
2521
  ],
2522
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
2523
  "torch_dtype": "float32",
2524
- "transformers_version": "4.13.0.dev0",
2525
  "use_weighted_layer_sum": true,
2526
  "vocab_size": 32,
2527
  "xvector_output_dim": 512
2483
  "layer_norm_eps": 1e-05,
2484
  "layerdrop": 0.1,
2485
  "mask_feature_length": 10,
2486
+ "mask_feature_min_masks": 0,
2487
  "mask_feature_prob": 0.0,
2488
  "mask_time_length": 10,
2489
+ "mask_time_min_masks": 2,
2490
  "mask_time_prob": 0.05,
2491
  "model_type": "unispeech-sat",
2492
  "num_attention_heads": 12,
2523
  ],
2524
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
2525
  "torch_dtype": "float32",
2526
+ "transformers_version": "4.14.0.dev0",
2527
  "use_weighted_layer_sum": true,
2528
  "vocab_size": 32,
2529
  "xvector_output_dim": 512
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b8c5f13a76fcbc75aa7ca49659bfe3365e79a9382712486f11982ee98c8361b8
3
  size 404496951
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a52dc0c7104acafe08376e4cb97369b2d9ec4875b9b26ac3f134b27e1144a199
3
  size 404496951