minimize_stateless: avoid reusing initialization seed

Discovered by running tests with `jax_enable_key_reuse_checks=True`.

PiperOrigin-RevId: 614737610

Author: vanderplas

tensorflow_probability/python/math/minimize.py

@@ -138,6 +138,9 @@ def run_jitted_minimize():
     seed_is_none = seed is None
     if not seed_is_none:
       seed = samplers.sanitize_seed(seed, salt='minimize')
+      init_seed, seed = samplers.split_seed(seed, n=2)
+    else:
+      init_seed = None
 
     if not return_full_length_trace:
       # Augment trace to record convergence info, so we can truncate it later.
@@ -153,7 +156,7 @@ def run_jitted_minimize():
      initial_optimizer_state) = optimizer_step_fn(
          parameters=initial_parameters,
          optimizer_state=initial_optimizer_state,
-         seed=seed)
+         seed=init_seed)
 
     initial_convergence_criterion_state = ()
     if convergence_criterion is not None: