angelo-ortiz commited on
Commit
6d87aba
·
verified ·
1 Parent(s): 14bda42

Upload folder using huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +16 -3
README.md CHANGED
@@ -4,9 +4,7 @@ license: mit
4
  tags:
5
  - acoustic-unit-discovery
6
  - hubert
7
- - model_hub_mixin
8
- - phonetics
9
- - pytorch_model_hub_mixin
10
  - speech
11
  language_bcp47:
12
  - ab
@@ -83,6 +81,21 @@ model = MauBERTFeat.from_pretrained("coml/maubert-feat")
83
  Check out the [`maubert`](https://github.com/bootphon/maubert) repository if you are interested in further retraining
84
  or fine-tuning, and the [corresponding paper](https://aclanthology.org/2026.acl-long.24/) for our results.
85
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
86
  ## Citing
87
 
88
  ```bibtex
 
4
  tags:
5
  - acoustic-unit-discovery
6
  - hubert
7
+ - maubert
 
 
8
  - speech
9
  language_bcp47:
10
  - ab
 
81
  Check out the [`maubert`](https://github.com/bootphon/maubert) repository if you are interested in further retraining
82
  or fine-tuning, and the [corresponding paper](https://aclanthology.org/2026.acl-long.24/) for our results.
83
 
84
+ ## Files:
85
+
86
+ - `model.safetensors` and `config.json`: HuggingFace Transformers checkpoint and config.
87
+
88
+ The manifests, phone alignments, language table and phone inventories used to train this model are released separately
89
+ in [`coml/vox-communis-artifacts`](https://huggingface.co/datasets/coml/vox-communis-artifacts). The inventories are
90
+ required to reduce the phone head when extracting features from the phone projection layer:
91
+
92
+ ```python
93
+ from huggingface_hub import snapshot_download
94
+
95
+ data_dir = snapshot_download("coml/vox-communis-artifacts", repo_type="dataset")
96
+ # then pass f"{data_dir}/canonical_inventories.jsonl" as `inventory_file`
97
+ ```
98
+
99
  ## Citing
100
 
101
  ```bibtex