[Regressions] ThunderFX is slower than 2 weeks ago for 3 models

Nobody has claimed this yet.

## 🐛 Bug

Recently found regressions:

falcon-40b -> mentioned in https://github.com/Lightning-AI/lightning-thunder/issues/1366
falcon-180B -> mentioned in https://github.com/Lightning-AI/lightning-thunder/issues/1484
falcon-7b -> New, regression probably due to MBS difference w.r.t torch.compile
pythia-12b -> New, fp8 regression, bf16 regression probably due to MBS difference w.r.t torch.compile
Platypus-30B -> mentioned in https://github.com/Lightning-AI/lightning-thunder/issues/1366
Gemma-7b -> mentioned in https://github.com/Lightning-AI/lightning-thunder/issues/1484
vicuna-33b-v1.3 -> mentioned in https://github.com/Lightning-AI/lightning-thunder/issues/1366
pythia-14m -> New, regression probably due to MBS difference w.r.t torch.compile

### To Reproduce

Tested on `pjnl-20241205` (as in the Latest image date in the screenshot).

### Environment

system.device_product_name DGXH100
system.gpu_driver_version 535.129.03
libraries.cuda 12.8.33
libraries.pip.lightning 2.4.0.dev20240728
libraries.pip.lightning-thunder 0.2.0.dev0
libraries.pip.lightning-utilities 0.11.9
libraries.pip.litgpt 0.4.11
libraries.pip.nvfuser 0.2.23+git1dda106
libraries.pip.pytorch-lightning 2.4.0
libraries.pip.torch 2.6.0a0+git45ed7c1
libraries.pip.torchao NaN
libraries.pip.torchmetrics 1.6.0
libraries.pip.torchvision 0.19.0a0+d23a6e1

Start by reproducing the listed model benchmarks on pjnl-20241205 with the reported DGXH100, CUDA, PyTorch, and Lightning Thunder versions, then compare results with the performance from two weeks earlier. Review the related issues 1366 and 1484 for existing investigations; done means identifying and resolving the regressions and confirming improved results for the affected models.

Written by the indexing model from the issue text.