Card without manifest.json; manifest.json removed

README.md

@@ -9,7 +9,7 @@ Text for pretraining a Python and C++ code model from scratch: 178,887,291 docum
 
 ## Layout
 
-One folder per source, each with `train/` and `val/` Parquet shards. Validation holds 1% of each source: whole repositories in the repository folders (`stack_edu_*`, `ultradata_stack_v3_l2_*`, `ultradata_code_l2_*`) and `codexglue_python`, whole problems in `codeforces_cots`, documents elsewhere. A repository folder's shard holds whole repositories, each cut into samples of files that import one another (`sample`); any other folder's train rows are in `id` order, so its leading shards are a uniform sample. `manifest.json` records each source's origin and each folder's rows.
+One folder per source, each with `train/` and `val/` Parquet shards. Validation holds 1% of each source: whole repositories in the repository folders (`stack_edu_*`, `ultradata_stack_v3_l2_*`, `ultradata_code_l2_*`) and `codexglue_python`, whole problems in `codeforces_cots`, documents elsewhere. A repository folder's shard holds whole repositories, each cut into samples of files that import one another (`sample`); any other folder's train rows are in `id` order, so its leading shards are a uniform sample.
 
 ## Schema