The major changes compared to v1.2.0 are:
- AI Engine (AIE) Backend support as external plugin by @dimdano in #1390
- ☢️ Libero backend for rad-hard PolarFire line of FPGAs by @vloncar in #1240
- Einsum and EinsumDense for oneAPI by @laurilaatu in #1424
- vivado/vitis support sample broadcasting merge ops by @calad0i in #1426
- pytest based synthesis tests by @marco66colombo in #1257
The full list of changes is:
- Bump version to 1.2.0 and switch to ruff for formatting by @JanFSchulte in #1395
- AI Engine (AIE) Backend support as external plugin by @dimdano in #1390
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1397
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1399
- Update funding section in README.md by @bo3z in #1403
- Add NGT logo to README by @jmduarte in #1404
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1402
- Manually increment checkout to version v6 by @JanFSchulte in #1405
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1408
- Backend predict hook by @dimdano in #1409
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1411
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1414
- pytest based synthesis tests by @marco66colombo in #1257
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1416
- Remove parametrized fixtures from pytests for pytest 9 compatability by @JanFSchulte in #1417
- Bump upload-artifact version by @JanFSchulte in #1413
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1420
- allow partial config def by @calad0i in #1419
- Add keras3 environment in tests by @marco66colombo in #1412
- ☢️ Libero backend for rad-hard PolarFire line of FPGAs by @vloncar in #1240
- minor fixes to CONTRIBUTING.md by @jmitrevs in #1423
- Einsum and EinsumDense for oneAPI by @laurilaatu in #1424
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1425
- Add LHC trigger use case context to README by @siddardhadesu in #1418
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1430
- hgq2 homogeneous quant fix by @calad0i in #1427
- Da custom layer by @calad0i in #1429
- Fix Keras v3 model conversion in numerical profiling by @Abubakar-rashid in #1421
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1434
- Uniform test root path by @jmitrevs in #1435
- Pytest folder naming update by @nghielme in #1437
- Fix parsing of activations in Keras v3 non-activation layers by @bo3z in #1440
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1441
- vivado/vitis support sample broadcasting merge ops by @calad0i in #1426
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1447
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1450
- Bump upload-artifact version by @JanFSchulte in #1449
- fix autoprecision fallback for generic weights by @calad0i in #1453
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #1452
New Contributors
- @siddardhadesu made their first contribution in #1418
- @Abubakar-rashid made their first contribution in #1421
Full Changelog: v1.2.0...v1.3.0