tuieval 0.2.0.dev119__tar.gz → 0.2.0.dev120__tar.gz

{tuieval-0.2.0.dev119 → tuieval-0.2.0.dev120}/CHANGELOG.md

@@ -2,6 +2,7 @@
 
 ## Unreleased
 
+- Fix: `v` on the run screen switches to the last model still running when the stream shows one that has finished.
 - Setup opens with no models ticked: at launch, after New run, and after Back from a run or tune. Packs and options are still remembered.
 - Tuning: a server that refuses the `ignore_eos` request field is asked again without it, so strict servers can be tuned; the profile notes that workload answers may end early.
 - Fix: a server that crashes mid-request (the stream ends with no finish reason) was graded as an empty or cut-short answer; it's a server error now, never counted against the model.

{tuieval-0.2.0.dev119 → tuieval-0.2.0.dev120}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: tuieval
-Version: 0.2.0.dev119
+Version: 0.2.0.dev120
 Summary: Evaluate local and frontier LLMs on your own questions: accuracy, speed, tokens and PASS/FAIL verdicts, in the terminal.
 Project-URL: Homepage, https://github.com/ashe-wb/tuieval
 Project-URL: Issues, https://github.com/ashe-wb/tuieval/issues

{tuieval-0.2.0.dev119 → tuieval-0.2.0.dev120}/src/tuieval/_version.py

@@ -18,7 +18,7 @@ version_tuple: tuple[int | str, ...]
 commit_id: str | None
 __commit_id__: str | None
 
-__version__ = version = '0.2.0.dev119'
-__version_tuple__ = version_tuple = (0, 2, 0, 'dev119')
+__version__ = version = '0.2.0.dev120'
+__version_tuple__ = version_tuple = (0, 2, 0, 'dev120')
 
 __commit_id__ = commit_id = None

{tuieval-0.2.0.dev119 → tuieval-0.2.0.dev120}/src/tuieval/tui.py

@@ -2779,8 +2779,8 @@ class RunScreen(Screen):
     def action_watch_next(self):
         """Stream the next model running alongside (from its next answer)."""
         active = self.active()
-        if len(active) < 2:
-            self.notify("Only one model is running right now.")
+        if not active or active == [self.watch]:
+            self.notify("Only one model is running right now." if active else "No model is running right now.")
             return
         self.watch = active[(active.index(self.watch) + 1) % len(active)] if self.watch in active else active[0]
         with self.buf_lock:

{tuieval-0.2.0.dev119 → tuieval-0.2.0.dev120}/tests/test_tuieval.py

@@ -847,6 +847,17 @@ class Parallel(unittest.TestCase):
         self.assertEqual({out["first"], out["switched"]}, {"a", "b"})   # v streams the other model
         self.assertTrue(any(r.get("ran_alongside") for r in self.results("a")["results"]))
 
+    def test_watch_next_moves_to_the_last_model_running(self):
+        from tuieval.tui import RunScreen
+        run = unittest.mock.MagicMock(watch="a", buf={"reasoning": [], "answer": []}, buf_lock=threading.Lock())
+        run.active.return_value = ["b"]          # a finished; only b is still running
+        RunScreen.action_watch_next(run)
+        self.assertEqual(run.watch, "b")
+        run.active.return_value = ["b"]
+        RunScreen.action_watch_next(run)          # already watching the only one
+        self.assertEqual(run.watch, "b")
+        run.notify.assert_called_with("Only one model is running right now.")
+
     def test_skip_one_model_keeps_the_other(self):
         e, events = self.engine()
         jobs = e.plan(["a", "b"], ["apps"], None, "certify", False, {})