Optionally use `flash-attn`'s CE loss for metrics

[snarayan21]

What does this PR do?

Resubmission of #3214 -- using FA's CE Loss results in lower peak reserved memory usage and higher throughput. We are not adding flash attention as an optional dependency to composer since this makes installs and correct builds messy & take a lot longer.

Fixed a small typo where the torch 3.11 CPU tests were using the GPU image with flash attn installed by accident.

Also modified DeviceGPU class so that it instantiates a gloo backend for CPU tensors, if gloo is available. This handles cases where users may want to perform distributed operations with tensors present on CPU even if they are using GPUs.

Manual tests:

• Run started on dev (13b-dense-fsdp-fullshard-hsdp-adam-shardedckpt-start-5PtEdK), resumed with this branch (13b-dense-fsdp-fullshard-hsdp-adam-shardedckpt-resume-E5SieL)
• Run started on this branch (13b-dense-fsdp-fullshard-hsdp-adam-shardedckpt-start-0g8uD4), resumed with dev branch (13b-dense-fsdp-fullshard-hsdp-adam-shardedckpt-resume-TSGoUC)

4th time's the charm :0

Run with torch CE loss (green): tiny-sp-dtms1-32h-wCFWfa Run with FA CE loss (tan): tiny-sp-dtms1-32h-jOfIPL

What issue(s) does this change relate to?

Before submitting

• [ ] Have you read the contributor guidelines?
• [ ] Is this change a documentation change or typo fix? If so, skip the rest of this checklist.
• [ ] Was this change discussed/approved in a GitHub issue first? It is much more likely to be merged if so.
• [ ] Did you update any related docs and document your change?
• [ ] Did you update any related tests and add any new tests related to your change? (see testing)
• [ ] Did you run the tests locally to make sure they pass?
• [ ] Did you run pre-commit on your change? (see the pre-commit section of prerequisites)

[ShashankMosaicML]

Can we offer a flag to gate as well? IIRC there are occasionally numerics issues for long seq...

@ShashankMosaicML do u remember

Flash attention fixed the long seq issue in this PR: https://github.com/Dao-AILab/flash-attention/commit/c79de85ffa0d19b80fa468f90c5086e837499d72

[snarayan21]

Seeing the error below on CPU tests:

>       assert input.is_cuda and target.is_cuda, "Only support CUDA tensors"
E       AssertionError: Only support CUDA tensors

So i'm gonna add a check for torch.cuda.is_available()

[snarayan21]

jk. The torch 3.11 cpu tests were using the cuda image on accident, causing this problem. It was only the torch 3.11 tests too. Fixed that in this PR as well.

[snarayan21]

Added manual test names to PR description