Unitxt 1.14.0 - Faster Unitxt

What's Changed

• Simplify qa example by @yoavkatz in #1234
• allow multiple references for f1 strings metric by @ShirApp in #1225
• Add bluebench recipes by @shachardon in #1237
• Allow templates dicts to be python dicts and fix a bug in the TemplatesDict definition by @elronbandel in #1240
• Deep copy artifacts that fetched twice by @elronbandel in #1239
• Adding of ANLS metric to doc_vqa and info_vqa datasets by @alfassy in #1241
• Update README.md by @elronbandel in #1242
• Update version to 1.13.1 by @elronbandel in #1244
• Enhancements to inference engines by @lilacheden in #1243
• add post processor to convert log probs dictionary to probabilities of a specific class by @lilacheden in #1247
• CI for metrics other than main + Bugfix in RetrievalAtK by @lilacheden in #1246
• Add huggingface cache disabling option to unitxt settings by @elronbandel in #1250
• Make F1Strings faster by @elronbandel in #1248
• Fix duplicate column deletion bug in pandas serializer by @elronbandel in #1249
• revived no_deep just to compare performance by @dafnapension in #1254
• fixed scigen post-processor by @csrajmohan in #1253
• Add prediction length metric by @perlitz in #1252
• Fix faithfulness confidence intervals by @matanor in #1257
• Allow role names to be captialized in SerializeOpenAiFormatDialog by @yoavkatz in #1259
• Accelerate image example 1000X by @elronbandel in #1258
• Fix the empty few-shot target issue when using produce() by @marukaz in #1266
• fix postprocessors in turl_col_type taskcard by @csrajmohan in #1261
• Fix answer correctness confidence intervals by @matanor in #1256
• add BlueBench as a benchmark to the catalog by @shachardon in #1262
• Fix MultipleSourceLoader documentation by @marukaz in #1270
• Ignore unitxt-venv by @marukaz in #1269
• Add mmmu by @elronbandel in #1271
• A fix for a bug in metric pipeline by @elronbandel in #1268
• Added Tablebench taskcard by @csrajmohan in #1273
• Fix missing deep copy in MapInstanceValues by @yoavkatz in #1267
• Add stream name to generation of dataset by @elronbandel in #1276
• Fix demos pool inference by @elronbandel in #1278
• Fix quality github action by @elronbandel in #1281
• add operators for robustness check on tables by @csrajmohan in #1279
• Instruction in SystemFormet demo support. by @piotrhelm in #1274
• change the max_test_instances of bluebench.recipe.attaq_500 to 100 by @shachardon in #1285
• Add documentation for types and serializers by @elronbandel in #1286
• Add example for image processing with different templates by @elronbandel in #1280
• Integrate metrics team LLMaJ with current unitxt implemantation by @lilacheden in #1205
• performance profiler with visualization by @dafnapension in #1255
• Remove split arg to support old hf datasets versions by @elronbandel in #1288
• add post-processors for tablebench taskcard by @csrajmohan in #1289
• recursive copy seems safer here by @dafnapension in #1295
• Fix performance tracking action by @elronbandel in #1296
• try num of instances in nested global scores by @dafnapension in #1282
• Update version to 1.14.0 by @elronbandel in #1298

New Contributors

• @alfassy made their first contribution in #1241
• @piotrhelm made their first contribution in #1274

Full Changelog: 1.13.0...1.14.0