Fix gratuitous, parasitic, endlessly repeated, pointless menu in version 3.2.0 (#4864)

## What type of PR is this? (check all applicable)

- [ ] Refactor
- [ ] Feature
- [X] Bug Fix
- [ ] Optimization
- [ ] Documentation Update
- [ ] Community Node Submission


## Have you discussed this change with the InvokeAI team?
- [X] Yes
- [ ] No, because:

      
## Have you updated all relevant documentation?
- [X] Yes
- [ ] No


## Description

A regression in 3.2.0 causes a seemingly nonsensical multiple choice
menu to appear when importing an SD-1 checkpoint model from the
autoimport directory. The menu asks the user to identify which type of
SD-2 model they are trying to import, which makes no sense.

In fact, the menu is popping up because there are now both "epsilon" and
"vprediction" SchedulerPredictionTypes for SD-1 as well as SD-2 models,
and the prober can't determine which prediction type to use. This PR
does two things:

1) rewords the menu as shown below
2) defaults to the most likely choice -- epsilon for v1 models and
vprediction for v2s

Here is the revised multiple-choice menu:
```
Please select the scheduler prediction type of the checkpoint named v1-5-pruned-emaonly.safetensors:
[1] "epsilon" - most v1.5 models and v2 models trained on 512 pixel images
[2] "vprediction" - v2 models trained on 768 pixel images and a few v1.5 models
[3] Accept the best guess;  you can fix it in the Web UI later

select [3]> 
```

Note that one can also put the appropriate config file into the same
directory as the checkpoint you wish to import. Give it the same name as
the model file, but with the extension `.yaml`. For example
`v1-5-pruned-emaonly.yaml`. The system will notice the yaml file and use
that, suppressing the quiz entirely.

## Related Tickets & Documents
- Closes #4768
- Closes #4827

invokeai/frontend/install/model_install.py

@@ -20,6 +20,7 @@ from multiprocessing import Process
 from multiprocessing.connection import Connection, Pipe
 from pathlib import Path
 from shutil import get_terminal_size
+from typing import Optional
 
 import npyscreen
 import torch
@@ -630,21 +631,23 @@ def ask_user_for_prediction_type(model_path: Path, tui_conn: Connection = None)
         return _ask_user_for_pt_cmdline(model_path)
 
 
-def _ask_user_for_pt_cmdline(model_path: Path) -> SchedulerPredictionType:
+def _ask_user_for_pt_cmdline(model_path: Path) -> Optional[SchedulerPredictionType]:
     choices = [SchedulerPredictionType.Epsilon, SchedulerPredictionType.VPrediction, None]
     print(
         f"""
-Please select the type of the V2 checkpoint named {model_path.name}:
-[1] A model based on Stable Diffusion v2 trained on 512 pixel images (SD-2-base)
-[2] A model based on Stable Diffusion v2 trained on 768 pixel images (SD-2-768)
-[3] Skip this model and come back later.
+Please select the scheduler prediction type of the checkpoint named {model_path.name}:
+[1] "epsilon" - most v1.5 models and v2 models trained on 512 pixel images
+[2] "vprediction" - v2 models trained on 768 pixel images and a few v1.5 models
+[3] Accept the best guess;  you can fix it in the Web UI later
 """
     )
     choice = None
     ok = False
     while not ok:
         try:
-            choice = input("select> ").strip()
+            choice = input("select [3]> ").strip()
+            if not choice:
+                return None
             choice = choices[int(choice) - 1]
             ok = True
         except (ValueError, IndexError):
@@ -655,22 +658,18 @@ Please select the type of the V2 checkpoint named {model_path.name}:
 
 
 def _ask_user_for_pt_tui(model_path: Path, tui_conn: Connection) -> SchedulerPredictionType:
-    try:
-        tui_conn.send_bytes(f"*need v2 config for:{model_path}".encode("utf-8"))
-        # note that we don't do any status checking here
-        response = tui_conn.recv_bytes().decode("utf-8")
-        if response is None:
-            return None
-        elif response == "epsilon":
-            return SchedulerPredictionType.epsilon
-        elif response == "v":
-            return SchedulerPredictionType.VPrediction
-        elif response == "abort":
-            logger.info("Conversion aborted")
-            return None
-        else:
-            return response
-    except Exception:
+    tui_conn.send_bytes(f"*need v2 config for:{model_path}".encode("utf-8"))
+    # note that we don't do any status checking here
+    response = tui_conn.recv_bytes().decode("utf-8")
+    if response is None:
+        return None
+    elif response == "epsilon":
+        return SchedulerPredictionType.epsilon
+    elif response == "v":
+        return SchedulerPredictionType.VPrediction
+    elif response == "guess":
+        return None
+    else:
         return None
 
 

invokeai/frontend/install/widgets.py

@@ -381,12 +381,12 @@ def select_stable_diffusion_config_file(
     wrap: bool = True,
     model_name: str = "Unknown",
 ):
-    message = f"Please select the correct base model for the V2 checkpoint named '{model_name}'. Press <CANCEL> to skip installation."
+    message = f"Please select the correct prediction type for the checkpoint named '{model_name}'. Press <CANCEL> to skip installation."
     title = "CONFIG FILE SELECTION"
     options = [
-        "An SD v2.x base model (512 pixels; no 'parameterization:' line in its yaml file)",
-        "An SD v2.x v-predictive model (768 pixels; 'parameterization: \"v\"' line in its yaml file)",
-        "Skip installation for now and come back later",
+        "'epsilon' - most v1.5 models and v2 models trained on 512 pixel images",
+        "'vprediction' - v2 models trained on 768 pixel images and a few v1.5 models)",
+        "Accept the best guess; you can fix it in the Web UI later",
     ]
 
     F = ConfirmCancelPopup(
@@ -410,7 +410,7 @@ def select_stable_diffusion_config_file(
     choice = F.add(
         npyscreen.SelectOne,
         values=options,
-        value=[0],
+        value=[2],
         max_height=len(options) + 1,
         scroll_exit=True,
     )
@@ -420,5 +420,5 @@ def select_stable_diffusion_config_file(
     if not F.value:
         return None
     assert choice.value[0] in range(0, 3), "invalid choice"
-    choices = ["epsilon", "v", "abort"]
+    choices = ["epsilon", "v", "guess"]
     return choices[choice.value[0]]