[Summer beauty] Prepare upgrade to Torch 2.x and Lightning 2.x

[mpariente]

Release before moving to asteroid 0.7.x with torch and lightning upgrades.

Breaking

• [install] Limit torch and pytorch-lightning to 1.x (Limit torch and pytorch-lightning to 1.x #671)

Added

• [egs] Raise error if storage_dir not specified in LibriMix recipes ([egs] Raise error if storage_dir not specified in LibriMix recipes #626)
• [egs] Create Multi-Decoder DPRNN README.md (Create README.md #632)
• [egs] Update Multi-Decoder DPRNN with inference function on a single file out-of-the-box (Update Multi-Decoder DPRNN with inference function on a single file out-of-the-box #653)
• [egs] Add a new pre-trained model (X-UMXL) (Add a new pre-trained model (X-UMXL) #665)
• [egs] Add a notebook demo, separate.py script for MD-DPRNN. Easier use from_pretrained (Add a colab notebook demo, and a separate script. Support from_pretrained for easier use. #668)

Changed

• [src] Use torch.linalg for solve and cholesky (Use torch.linalg for solve and cholesky #623)
• [install] Remove torchmetrics version limit
• [CI] Lint with black 22.3.0 and fix linting CI python3.*8 ([CI] Lint with black 22.3.0 and fix linting CI python3.*8 #624)

Fixed

• [hub] Upgrade deprecated huggingface_hub.cached_download (Upgrade deprecated huggingface_hub.cached_download #645)
• [src] Fix cumulative normalization (Fix cumulative normalization #649)
• [src] Fix argparse key based on Python version (KeyError: Options in training stage for recipes #628) ([src] Fix argparse key based on Python version (#628) #657)
• [ci] Fix tests in continuous integration (Fix continuous integration #672)

Many thanks to all the contributors @actuallyaswin, @mystlee, @JunzheJosephZhu, @jbartolewska, @LeonieBorne, @mattiadg, @r-sawata and @zmolikova ! 💪 🤩 🙏

---

This discussion was created from the release [Summer beauty] Prepare upgrade to Torch 2.x and Lightning 2.x.