export.py: fix custom SDPA type conversion logic & re-enable for bfloat16

[swolchok]

Stack from ghstack (oldest at bottom):

• -> export.py: fix custom SDPA type conversion logic & re-enable for bfloat16 #1171
• Default max_seq_length to 128 for ExecuTorch export #1170
• Also use default max_seq_length of 128 for ExecuTorch native runner #1186
• Use default max_seq_length of 128 when loading ExecuTorch models #1184
• add ConvertToLinear, disable custom SDPA for bfloat16 #1169
• parallelize ExecuTorch build #1168
• Update ExecuTorch pin to pick up bfloat16 fixes #1181

q.dtype is always torch.float at this point. Clearly meant to use the input dtype.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchchat/1171

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 4831e83 with merge base 2cf4016 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[swolchok]

Ugh, some ghstack learning pains on my end here.