Township: per-profile LoRA training, Matches page, modals; fix SDXL LoRA dtype

LoRA training (codai/api/loras.py):
- Fix instant crash "mat1 and mat2 have the same dtype (Half != float)":
  components loaded in their checkpoint's native dtype (fp16 text encoders +
  fp32 UNet). Force consistent fp32 (stable LoRA reference) for both SD1.5 and
  SDXL, and free VAE/text encoders after pre-encoding to keep memory in budget.

Web UI (tools/gen_township_fighters.py):
- Per-profile LoRA training launched from the Characters/Environments cards
  (steps/rank, live server progress, status badge); the "Train LoRAs" step now
  trains regardless of the consistency setting and logs what it does. Profiles
  are uploaded to CoderAI first so server-side training has reference images.
- New Matches page: final short/long videos, single clips, and outcome outputs
  per match. Re-render clips/outputs with the video model (reuses the render
  path, honoring keyframe/LoRA settings) or reassemble finals from existing
  clips (no model).
- Replace browser alert/confirm/prompt with proper in-page modal dialogs
  (uiConfirm/uiAlert/uiPrompt) on every page.
- Apply changed connection/model settings (image/video/text model) to the live
  session on Save config / Start, so per-profile jobs use them immediately.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

codai/api/loras.py

@@ -311,10 +311,12 @@ def _train_sd15(req, base_path, images, instance_prompt,
     name = req.name
     g = torch.Generator(device=device).manual_seed(seed)
 
+    # Consistent fp32 precision (see _train_sdxl) to avoid mixed-dtype crashes.
+    weight_dtype = torch.float32
     tokenizer = CLIPTokenizer.from_pretrained(base_path, subfolder="tokenizer")
-    text_encoder = CLIPTextModel.from_pretrained(base_path, subfolder="text_encoder").to(device)
-    vae = AutoencoderKL.from_pretrained(base_path, subfolder="vae").to(device)
-    unet = UNet2DConditionModel.from_pretrained(base_path, subfolder="unet").to(device)
+    text_encoder = CLIPTextModel.from_pretrained(base_path, subfolder="text_encoder").to(device, dtype=weight_dtype)
+    vae = AutoencoderKL.from_pretrained(base_path, subfolder="vae").to(device, dtype=weight_dtype)
+    unet = UNet2DConditionModel.from_pretrained(base_path, subfolder="unet").to(device, dtype=weight_dtype)
     noise_scheduler = DDPMScheduler.from_pretrained(base_path, subfolder="scheduler")
 
     vae.requires_grad_(False)
@@ -338,6 +340,13 @@ def _train_sd15(req, base_path, images, instance_prompt,
                         return_tensors="pt").input_ids.to(device)
         encoder_hidden_states = text_encoder(tok)[0]
 
+    # VAE + text encoder are done; free them so only the UNet trains resident.
+    del vae, text_encoder
+    try:
+        torch.cuda.empty_cache()
+    except Exception:
+        pass
+
     _set_progress(status="training", message="training (SD1.5)")
     unet.train()
     n = len(latents_list)
@@ -379,7 +388,7 @@ def _train_sd15(req, base_path, images, instance_prompt,
     _write_meta(name, req, base_path, len(images), "sd15", instance_prompt)
 
     # Release training tensors.
-    del unet, vae, text_encoder, optimizer, latents_list
+    del unet, optimizer, latents_list
     try:
         torch.cuda.empty_cache()
     except Exception:
@@ -405,12 +414,19 @@ def _train_sdxl(req, base_path, images, instance_prompt,
     name = req.name
     g = torch.Generator(device=device).manual_seed(seed)
 
+    # Train in a single consistent precision. Checkpoints can store each
+    # component in a different native dtype (e.g. fp16 text encoders + fp32
+    # UNet), which otherwise crashes with "mat1 and mat2 have the same dtype"
+    # (Half != float) in cross-attention. fp32 is the stable reference for
+    # LoRA fine-tuning.
+    weight_dtype = torch.float32
+
     tokenizer_1 = CLIPTokenizer.from_pretrained(base_path, subfolder="tokenizer")
     tokenizer_2 = CLIPTokenizer.from_pretrained(base_path, subfolder="tokenizer_2")
-    text_encoder_1 = CLIPTextModel.from_pretrained(base_path, subfolder="text_encoder").to(device)
-    text_encoder_2 = CLIPTextModelWithProjection.from_pretrained(base_path, subfolder="text_encoder_2").to(device)
-    vae = AutoencoderKL.from_pretrained(base_path, subfolder="vae").to(device)
-    unet = UNet2DConditionModel.from_pretrained(base_path, subfolder="unet").to(device)
+    text_encoder_1 = CLIPTextModel.from_pretrained(base_path, subfolder="text_encoder").to(device, dtype=weight_dtype)
+    text_encoder_2 = CLIPTextModelWithProjection.from_pretrained(base_path, subfolder="text_encoder_2").to(device, dtype=weight_dtype)
+    vae = AutoencoderKL.from_pretrained(base_path, subfolder="vae").to(device, dtype=weight_dtype)
+    unet = UNet2DConditionModel.from_pretrained(base_path, subfolder="unet").to(device, dtype=weight_dtype)
     noise_scheduler = DDPMScheduler.from_pretrained(base_path, subfolder="scheduler")
 
     for m in (vae, text_encoder_1, text_encoder_2, unet):
@@ -458,6 +474,14 @@ def _train_sdxl(req, base_path, images, instance_prompt,
         device=device, dtype=prompt_embeds.dtype,
     )
 
+    # VAE + text encoders are no longer needed during the training loop; free
+    # them so only the UNet stays resident (keeps SDXL fp32 training in budget).
+    del vae, text_encoder_1, text_encoder_2
+    try:
+        torch.cuda.empty_cache()
+    except Exception:
+        pass
+
     _set_progress(status="training", message="training (SDXL)")
     unet.train()
     n = len(latents_list)
@@ -498,7 +522,7 @@ def _train_sdxl(req, base_path, images, instance_prompt,
                                                 safe_serialization=True)
     _write_meta(name, req, base_path, len(images), "sdxl", instance_prompt)
 
-    del unet, vae, text_encoder_1, text_encoder_2, optimizer, latents_list
+    del unet, optimizer, latents_list
     try:
         torch.cuda.empty_cache()
     except Exception: