datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
curated-elixir-the-stack-v3
Curated Elixir Stack v3
A curated, syntax-verified Elixir source code dataset derived from HuggingFaceCode/stack-v3-full. This was also an exploration for me to learn how datasets are curated.
Dataset Summary
Source: Derived directly from the public Elixir partition of The Stack v3 by BigCode.
File Count: ~564,000 files.
Target Size Bounds: Filtered to files between 1.5 KB and 25 KB post-cleaning.
Syntax Parsing: Validated using tree-sitter-elixir (AST parse… See the full description on the dataset page: https://huggingface.co/datasets/beyond369/curated-elixir-the-stack-v3.elixirDatasetsstack_elixir
Dataset Card for "stack_elixir"
More Information needed
elixir-trainning-sharegptelixir-thinking-v3
elixir-thinking-v3
1,518 ShareGPT-format Elixir/OTP examples with <think> reasoning traces,
generated with DeepSeek V4 Pro (low reasoning effort) on OpenRouter.
Splits
Split
Examples
train
1,442
validation
76
Random 95/5 split, seed=42.
Format
Each row is a conversations array with system / user / assistant
messages. The assistant message wraps native reasoning in <think>...</think>
and follows with idiomatic Elixir code + a short explanation.… See the full description on the dataset page: https://huggingface.co/datasets/nmaroulis/elixir-thinking-v3.elixir_generated
Dataset Card for elixir_generated
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/joelkoch/elixir_generated/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/joelkoch/elixir_generated.buzz_sources_138_elixirelixir-golden-dataset