Skip to page content
SAVRN
Search Plan a project

Independent publisher

Dvislobokov

dvislobokov

Models in Library0
Datasets in Library2
Models on Hugging Face2
Followers3

Compare the published evidence

This profile connects the 2 model and dataset records shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.

Published tasks: Text generation.

Reported licenses: other. Open a record to review its specific license and conditions.

Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.

Datasets

Dataset · Text generation

csharp-ml-complation

Dvislobokov

Caret-based full-line completion samples extracted from permissively licensed C# repositories with csharp-dataset-prepare. Each sample is an exact editor position: leftcontext ends at the caret, targettext is the rest of the physical line (no newline), rightcontext follows it. Original source is reconstructable from offsets; every row carries repositoryid, revision, relativepath and license for attribution. Splits are assigned per repository group (forks/near-copies together); cross-repository duplicate files are kept once. No code was built or executed. See the builder repository for the data contract, leakage policy and benchmarks. For every sample the builder removes the hidden target…

Publicly accessible other

Dataset · Text generation

go-ml-complation

Dvislobokov

Caret-based full-line completion samples extracted from permissively licensed Go repositories with the goflc builder (Go go/parser + go/types). Each sample is an exact editor position: leftcontext ends at the caret, targettext is the rest of the physical line (no newline, trailing whitespace excluded), rightcontext follows it. Original source is reconstructable from offsets (byte offsets as used by Go tooling, plus UTF-16 offsets for IDEs); every row carries repositoryid, revision, relativepath and license for attribution. Splits are assigned per repository group (forks/near-copies together); cross-repository duplicate files are kept once. Generated code (// Code generated... DO NOT…

Publicly accessible other