Fix Too Many Requests issue from Zenodo by hosting data (#3)

- Track data files by Git LFS (4d9c7b841c1fccdc37212729afbadbae7a25cf35)
- Host java data files (aa966aaf5ebad85f99df93e187f055781cde9f76)
- Update loading script (b728d7705b17067470a961f63e939c2c5f862b76)
- Add citation for java data (e21f8104961f28eb4f291d86130de1e5091854c0)
- Delete legacy metadata JSON file (0b606472eacef94022d2d0d57e3cd71fef95f352)

.gitattributes

@@ -25,3 +25,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zstandard filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+data/java/java_test_pre filter=lfs diff=lfs merge=lfs -text
+data/java/java_training_pre filter=lfs diff=lfs merge=lfs -text
+data/java/java_validation_pre filter=lfs diff=lfs merge=lfs -text

README.md

@@ -236,6 +236,24 @@ Computational Use of Data Agreement (C-UDA) License.
     }
 ```
 
+The data for "java" configuration comes from:
+```
+@dataset{rafael_michael_karampatsis_2020_3628665,
+  author       = {Rafael - Michael Karampatsis and
+                  Hlib Babii and
+                  Romain Robbes and
+                  Charles Sutton and
+                  Andrea Janes},
+  title        = {Preprocessed Java Code Corpus},
+  month        = jan,
+  year         = 2020,
+  publisher    = {Zenodo},
+  version      = {1.0},
+  doi          = {10.5281/zenodo.3628665},
+  url          = {https://doi.org/10.5281/zenodo.3628665}
+}
+```
+
 ### Contributions
 
 Thanks to @madlag (and partly also @ncoop57) for adding this dataset.

code_x_glue_cc_code_completion_token.py

@@ -27,6 +27,20 @@ _CITATION = """@article{raychev2016probabilistic,
     pages={207--216},
     year={2013},
     organization={IEEE}
+}
+@dataset{rafael_michael_karampatsis_2020_3628665,
+  author       = {Rafael - Michael Karampatsis and
+                  Hlib Babii and
+                  Romain Robbes and
+                  Charles Sutton and
+                  Andrea Janes},
+  title        = {Preprocessed Java Code Corpus},
+  month        = jan,
+  year         = 2020,
+  publisher    = {Zenodo},
+  version      = {1.0},
+  doi          = {10.5281/zenodo.3628665},
+  url          = {https://doi.org/10.5281/zenodo.3628665}
 }"""
 
 
@@ -52,7 +66,7 @@ class CodeXGlueCcCodeCompletionTokenJavaImpl(CodeXGlueCcCodeCompletionTokenImpl)
         if language != "java":
             raise RuntimeError(f"Unknown language {language}: should be java.")
 
-        yield "data", f"https://zenodo.org/record/3628665/files/java_{split_name}_pre"
+        yield "data", f"https://huggingface.co/datasets/code_x_glue_cc_code_completion_token/resolve/main/data/java/java_{split_name}_pre"
 
     def _generate_examples(self, split_name, file_paths):
         with open(file_paths["data"], encoding="utf-8") as f:

data/java/java_test_pre

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a88cd5c91c2ed23a928528bef3535f4fc8db1359975447211f2b13926cc38d9d
+size 26969670

data/java/java_training_pre

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:676295d2756adcac22e213fbc3ea0f50669a0d152e9497e23a2929e2e2124905
+size 81051708

data/java/java_validation_pre

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c58a97d96aa7435396581ee5efbb93c0e74a545ba5f795f878098a6e59ab8b3
+size 18835141

dataset_infos.json

@@ -1 +0,0 @@
-{"java": {"description": "CodeXGLUE CodeCompletion-token dataset, available at https://github.com/microsoft/CodeXGLUE/tree/main/Code-Code/CodeCompletion-token\n\nPredict next code token given context of previous tokens. Models are evaluated by token level accuracy.\nCode completion is a one of the most widely used features in software development through IDEs. An effective code completion tool could improve software developers' productivity. We provide code completion evaluation tasks in two granularities -- token level and line level. Here we introduce token level code completion. Token level task is analogous to language modeling. Models should have be able to predict the next token in arbitary types.\n", "citation": "@article{raychev2016probabilistic,\n    title={Probabilistic Model for Code with Decision Trees},\n    author={Raychev, Veselin and Bielik, Pavol and Vechev, Martin},\n    journal={ACM SIGPLAN Notices},\n    pages={731--747},\n    year={2016},\n    publisher={ACM New York, NY, USA}\n}\n@inproceedings{allamanis2013mining,\n    title={Mining Source Code Repositories at Massive Scale using Language Modeling},\n    author={Allamanis, Miltiadis and Sutton, Charles},\n    booktitle={2013 10th Working Conference on Mining Software Repositories (MSR)},\n    pages={207--216},\n    year={2013},\n    organization={IEEE}\n}", "homepage": "https://github.com/madlag/CodeXGLUE/tree/main/Code-Code/CodeCompletion-token", "license": "", "features": {"id": {"dtype": "int32", "id": null, "_type": "Value"}, "code": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "code_x_glue_cc_code_completion_token", "config_name": "java", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 128312061, "num_examples": 12934, "dataset_name": "code_x_glue_cc_code_completion_token"}, "validation": {"name": "validation", "num_bytes": 30259174, "num_examples": 7189, "dataset_name": "code_x_glue_cc_code_completion_token"}, "test": {"name": "test", "num_bytes": 43027956, "num_examples": 8268, "dataset_name": "code_x_glue_cc_code_completion_token"}}, "download_checksums": {"https://zenodo.org/record/3628665/files/java_training_pre": {"num_bytes": 81051708, "checksum": "676295d2756adcac22e213fbc3ea0f50669a0d152e9497e23a2929e2e2124905"}, "https://zenodo.org/record/3628665/files/java_validation_pre": {"num_bytes": 18835141, "checksum": "0c58a97d96aa7435396581ee5efbb93c0e74a545ba5f795f878098a6e59ab8b3"}, "https://zenodo.org/record/3628665/files/java_test_pre": {"num_bytes": 26969670, "checksum": "a88cd5c91c2ed23a928528bef3535f4fc8db1359975447211f2b13926cc38d9d"}}, "download_size": 126856519, "post_processing_size": null, "dataset_size": 201599191, "size_in_bytes": 328455710}, "python": {"description": "CodeXGLUE CodeCompletion-token dataset, available at https://github.com/microsoft/CodeXGLUE/tree/main/Code-Code/CodeCompletion-token\n\nPredict next code token given context of previous tokens. Models are evaluated by token level accuracy.\nCode completion is a one of the most widely used features in software development through IDEs. An effective code completion tool could improve software developers' productivity. We provide code completion evaluation tasks in two granularities -- token level and line level. Here we introduce token level code completion. Token level task is analogous to language modeling. Models should have be able to predict the next token in arbitary types.\n", "citation": "@article{raychev2016probabilistic,\n    title={Probabilistic Model for Code with Decision Trees},\n    author={Raychev, Veselin and Bielik, Pavol and Vechev, Martin},\n    journal={ACM SIGPLAN Notices},\n    pages={731--747},\n    year={2016},\n    publisher={ACM New York, NY, USA}\n}\n@inproceedings{allamanis2013mining,\n    title={Mining Source Code Repositories at Massive Scale using Language Modeling},\n    author={Allamanis, Miltiadis and Sutton, Charles},\n    booktitle={2013 10th Working Conference on Mining Software Repositories (MSR)},\n    pages={207--216},\n    year={2013},\n    organization={IEEE}\n}", "homepage": "https://github.com/madlag/CodeXGLUE/tree/main/Code-Code/CodeCompletion-token", "license": "", "features": {"id": {"dtype": "int32", "id": null, "_type": "Value"}, "path": {"dtype": "string", "id": null, "_type": "Value"}, "code": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "code_x_glue_cc_code_completion_token", "config_name": "python", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 684319575, "num_examples": 100000, "dataset_name": "code_x_glue_cc_code_completion_token"}, "test": {"name": "test", "num_bytes": 333978088, "num_examples": 50000, "dataset_name": "code_x_glue_cc_code_completion_token"}}, "download_checksums": {"http://files.srl.inf.ethz.ch/data/py150_files.tar.gz": {"num_bytes": 199067128, "checksum": "73be7f7a78e549845cf80cf779a3bcc3a9cf351ff7017e07b89e8d1c82b8d389"}}, "download_size": 199067128, "post_processing_size": null, "dataset_size": 1018297663, "size_in_bytes": 1217364791}}