Copy from MultilingualUnigramLM/LangMap-TheStack-java-100M-olmo2 (rename: drop -olmo2 suffix)

README.md

@@ -0,0 +1,15 @@
+---
+configs:
+- config_name: java
+  data_files:
+    - split: train
+      path: data/java-*.parquet
+---
+
+# LangMap-TheStack-java-100M-olmo2
+
+Code finetuning dataset for **java** streamed from [bigcode/the-stack](https://huggingface.co/datasets/bigcode/the-stack).
+
+- **Tokens collected**: 100,000,000 (target: 100,000,000)
+- **Tokenizer**: `allenai/OLMo-2-1124-7B`
+- **Schema**: `{"text": [...]}` (sanitised source code)

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d66d0094187f1b1edd74e738c26ddd5d7b3c33dee897ec43a9db51e39ec2ca0b
+size 137984025