upload

.gitattributes

@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+*.jsonl filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,7 +1,49 @@
-Hello there - If you are looking for the CS-YODAS dataset, please check back after May 19th. I am currently at LREC presenting this work and will complete the data upload promptly after thge conference. Thank you for your interest in our work!
-
--Brian
-
----
-license: cc-by-4.0
----
+# CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech
+
+## Overview
+
+CS-YODAS is a large-scale dataset of code-switched speech extracted from in-the-wild YouTube videos across six language pairs.
+
+For details, see the paper: [CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech](https://lrec.elra.info/lrec2026-main-456)
+
+## Dataset Statistics
+
+| Language | Code | Records |
+|----------|------|---------|
+| Arabic | ara | 245 |
+| Mandarin Chinese | cmn | 439 |
+| French | fra | 30,189 |
+| Hindi | hin | 1,834 |
+| Japanese | jpn | 242 |
+| Russian | rus | 1,081 |
+| **TOTAL** | | **34,030** |
+
+## How to Download
+
+```bash
+git lfs install
+git clone https://huggingface.co/datasets/byan/cs-yodas
+```
+
+Or load in 🤗 Datasets:
+
+```python
+from datasets import load_dataset
+data = load_dataset("byan/cs-yodas")
+```
+
+## License
+
+CC-BY-NC-4.0
+
+## Citation
+
+```bibtex
+@inproceedings{yan2026cs,
+  title={CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech},
+  author={Yan et al.},
+  booktitle={Proceedings of LREC 2026},
+  year={2026},
+  doi={10.63317/3fir5kp98qyt}
+}
+```

ara.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19254013052100a5c780d2342c21185babee874cec6a7cb5f14c67bae1435abb
+size 929359

ara.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b72b946e6cf971f1e4130fad46e7368a28748e696d731722068db550a4b2bcc0
+size 561602560

cmn.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6cdad8b713ea41a598077dd0fedae14824eb2d6998970132d5b2c885201c7697
+size 1175420

cmn.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:481ee979d308c99d3df40dd67cd55a7bbc5bd8fd98d316daa0d6ebd4f4b1cb81
+size 900536320

cs_yodas.py

@@ -0,0 +1,100 @@
+import json
+from pathlib import Path
+import datasets
+
+_CITATION = """\
+@inproceedings{yan2026cs,
+  title={CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech},
+  author={Yan et al.},
+  booktitle={Proceedings of LREC 2026},
+  year={2026},
+  doi={10.63317/3fir5kp98qyt}
+}
+"""
+
+_DESCRIPTION = """\
+CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech
+"""
+
+_HOMEPAGE = "https://lrec.elra.info/lrec2026-main-456"
+
+_LICENSE = "CC-BY-NC-4.0"
+
+_LANGUAGES = ["ara", "cmn", "fra", "hin", "jpn", "rus"]
+
+
+class CSYodasConfig(datasets.BuilderConfig):
+    """Builder config for CS-YODAS dataset."""
+
+    def __init__(self, language, **kwargs):
+        super(CSYodasConfig, self).__init__(**kwargs)
+        self.language = language
+
+
+class CSYodas(datasets.GeneratorBasedBuilder):
+    """CS-YODAS: A Mined Dataset of In-the-Wild Code-Switched Speech"""
+
+    BUILDER_CONFIGS = [
+        CSYodasConfig(
+            name=lang,
+            language=lang,
+            version=datasets.Version("1.0.0"),
+            description=f"CS-YODAS {lang} subset",
+        )
+        for lang in _LANGUAGES
+    ]
+
+    DEFAULT_CONFIG_NAME = "fra"
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "id": datasets.Value("string"),
+                    "text": datasets.Value("string"),
+                    "language": datasets.Value("string"),
+                    "languages": datasets.Sequence(datasets.Value("string")),
+                    "consistent": datasets.Value("int32"),
+                    "Q1": datasets.Value("string"),
+                    "Q2": datasets.Value("string"),
+                    "Q3": datasets.Value("string"),
+                    "Q4": datasets.Value("string"),
+                    "Q5": datasets.Value("string"),
+                    "Comments": datasets.Value("string"),
+                    "context_text": datasets.Value("string"),
+                    "context_text_list": datasets.Sequence(
+                        datasets.Sequence(datasets.Value("string"))
+                    ),
+                    "kaldi_uttid": datasets.Value("string"),
+                    "all_code_switched_texts": datasets.Sequence(
+                        datasets.Value("string")
+                    ),
+                    "wav_path": datasets.Value("string"),
+                }
+            ),
+            supervised_keys=None,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        language = self.config.language
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": f"{language}.jsonl",
+                    "language": language,
+                },
+            )
+        ]
+
+    def _generate_examples(self, filepath, language):
+        """Generate examples from JSONL file."""
+        with open(filepath, encoding="utf-8") as f:
+            for idx, line in enumerate(f):
+                if line.strip():
+                    data = json.loads(line)
+                    yield idx, data

fra.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76f415663519b4898ddd0540bbf7b5149aae25f181e9262a39e89aae843fbfda
+size 87735168

fra.tar.aa

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e53cb4f57f1245c341b10cf466a4173b653658a183a008a75c63b80045d66e9
+size 32212254720

fra.tar.ab

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f489d2a7999765b32e382f3f651300b28dffed9950e90e62ff0bee01b288c409
+size 28187269120

hin.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2533311b72283f436ec9a0aca687b8f66983bd2c1fdd2574704a3091cba3e526
+size 10256431

hin.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:115e7f32893b73e6568144c1d37e6978e76f76f5322305d4f8a504b7db957bc6
+size 4262277120

jpn.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5fca5676a09f07391aee943996ce6619529bea273b17237f756f1e66c7a5844
+size 602051

jpn.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fc886748e2c4a713f9e59e2d5cc1067d008d5390ef4932b35e11cba2bbc08cb
+size 453734400

rus.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f29381d741c611e54e5ad06d7645f4a9fa837508ea8784006d4f600d70be006a
+size 3947866

rus.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cd45336294f62592977dec607bf80dc470339ced892f886d312bd5338c3627
+size 2045777920