Leandro von Werra
Leandro von Werra
## Describe the bug Loading a JSON dataset with `load_dataset` can get stuck when running on a machine with many CPUs. This is especially an issue when loading a large...
Since there hasn't been a release of `rouge_score` in a while (#1199) I tried to install the package from source with: ```bash pip install git+https://github.com/google-research/google-research.git#subdirectory=rouge ``` Which returns: ```bash (env)...
# Performance analysis Following up on speed comments in #689 I made a more systematic test to see the download speed using `Repository` and `git clone`. I also added `load_dataset`...
The following error occurs when uploading ~1000 files (300GB total): ```bash LFS: Authorization error: https://s3.us-east-1.amazonaws.com/lfs.huggingface.co/datasets/lvwerra/github-code/XXXXXX Check that you have proper access to the repository ``` I am using the following...
Currently there are several different inputs/output formats possible in `Metrics`. We should standardize them as much as possible and respecting the following principle: - inputs/outputs are easy to understand and...
This PR reworks the config/kwargs logic for evaluation modules (closes #169). The structure is the following: 1. The allowed fields and their defaults are defined in a `Config` (a `dataclass`)...
This PR adds a `KerasCallback` class which lets a user wrap an evaluation in a `keras` compatible wrapper. Following the discussions in https://github.com/huggingface/evaluate/issues/10#issuecomment-1116365811. ```py import evaluate recall = evaluate.load("recall") transform...
If a dataset is passed as a string the loading of the data will fail since the rest of the pipeline expects a `dataset.Dataset` and not a `dataset.DatasetDict` which is...
This PR makes sure that the spaces are always on the same version as the `evaluate` repository. Also fixed the colors of the measurements.
Currently, when `compute` is called all data is loaded into memory and passed as a list. This can pose a bottleneck especially for data intensive modalities (e.g. images) or measurements...