CLI: add --launcher option, support launcher args, cleanup, refactor (#2924)

* add --launcher option; explicit True/False bool args; small cleanup

* refactor

* add torchrun, accelerate cli args

* add rdzv arg default + tests

* update _quarto

* coderabbit

* fix

* we can't set rdvz_id independently across nodes

* coderabbit

* fix tests

tests/cli/test_cli_base.py

@@ -17,16 +17,23 @@ class BaseCliTest:
             command: Command to test (train/evaluate)
         """
         # Test missing config file
-        result = cli_runner.invoke(cli, [command, "--no-accelerate"])
+        result = cli_runner.invoke(cli, [command, "--launcher", "python"])
         assert result.exit_code != 0
 
         # Test non-existent config file
-        result = cli_runner.invoke(cli, [command, "nonexistent.yml", "--no-accelerate"])
+        result = cli_runner.invoke(
+            cli, [command, "nonexistent.yml", "--launcher", "python"]
+        )
         assert result.exit_code != 0
         assert "Error: Invalid value for 'CONFIG'" in result.output
 
     def _test_basic_execution(
-        self, cli_runner, tmp_path: Path, valid_test_config: str, command: str
+        self,
+        cli_runner,
+        tmp_path: Path,
+        valid_test_config: str,
+        command: str,
+        train: bool = True,
     ):
         """Test basic execution with accelerate.
 
@@ -35,6 +42,7 @@ class BaseCliTest:
             tmp_path: Temporary path fixture
             valid_test_config: Valid config fixture
             command: Command to test (train/evaluate)
+            train: Whether to test training (default) or evaluation
         """
         config_path = tmp_path / "config.yml"
         config_path.write_text(valid_test_config)
@@ -43,15 +51,21 @@ class BaseCliTest:
             result = cli_runner.invoke(cli, [command, str(config_path)])
 
             assert mock.called
-            assert mock.call_args.args[0] == [
+
+            expected = [
                 "accelerate",
                 "launch",
                 "-m",
                 f"axolotl.cli.{command}",
                 str(config_path),
-                "--debug-num-examples",
-                "0",
+                "--debug=False",
+                "--debug-text-only=False",
+                "--debug-num-examples=0",
             ]
+            if train:
+                expected.append("--shard=False")
+
+            assert mock.call_args.args[0] == expected
             assert mock.call_args.kwargs == {"check": True}
             assert result.exit_code == 0