enable using python_callable without mapping symbols to their impls

[t-vi]

Currently, we can't use python_callable if we want to retrace because it does half of transform for execution.
We should absolutely get a flag on python_callable for that because we have a workaround in Thunder quite often now (in thunder/core/transforms.py but also in the autograd.Function handling in jit_ext now).
(the ["output"] might not make this example trivial, but still).

lightning-thunder/thunder/core/transforms.py

Lines 1438 to 1445 in ea96657

	# Using trc.python_callable() makes it impossible to retrace the
	# function because the python_callable uses python_ctx which replaces
	# symbol occurrences with its symbol._call_ctx function
	computation_trc = dce(computation_trc)
	@wraps(computation_trc.python_callable())
	def python_callable(*args, **kwargs):
	return eval_trace(computation_trc, *args, **kwargs)["output"]

cc @apaz-cli

[t-vi]

This is another, more prototypical, example the commented line does not work, but it should with a flag (or maybe even by default, so only transform for execution does it differently).

lightning-thunder/thunder/core/jit_ext.py

Lines 690 to 694 in 9580411

	@wraps(trace_of_augmented_fwd.python_callable())
	def core_of_augmented_forward(*args, **kwargs):
	return thunder.core.trace_interpreter.interpret_trace(trace_of_augmented_fwd, *args, **kwargs)
	# core_of_augmented_forward = trace_of_augmented_fwd.python_callable(include_decorators=False)