Merge pull request #1056 from kohya-ss/dev

fix vram usage in LoRA training

README.md

@@ -249,6 +249,16 @@ ControlNet-LLLite, a novel method for ControlNet with SDXL, is added. See [docum
 
 ## Change History
 
+### Jan 17, 2024 / 2024/1/17: v0.8.1
+
+- Fixed a bug that the VRAM usage without Text Encoder training is larger than before in training scripts for LoRA etc (`train_network.py`, `sdxl_train_network.py`).
+  - Text Encoders were not moved to CPU.
+- Fixed typos. Thanks to akx! [PR #1053](https://github.com/kohya-ss/sd-scripts/pull/1053)
+
+- LoRA 等の学習スクリプト（`train_network.py`、`sdxl_train_network.py`）で、Text Encoder を学習しない場合の VRAM 使用量が以前に比べて大きくなっていた不具合を修正しました。 
+  - Text Encoder が GPU に保持されたままになっていました。
+- 誤字が修正されました。 [PR #1053](https://github.com/kohya-ss/sd-scripts/pull/1053) akx 氏に感謝します。
+
 ### Jan 15, 2024 / 2024/1/15: v0.8.0
 
 - Diffusers, Accelerate, Transformers and other related libraries have been updated. Please update the libraries with [Upgrade](#upgrade).

sdxl_train_network.py

@@ -95,8 +95,8 @@ def cache_text_encoder_outputs_if_needed(
                 unet.to(org_unet_device)
         else:
             # Text Encoderから毎回出力を取得するので、GPUに乗せておく
-            text_encoders[0].to(accelerator.device)
-            text_encoders[1].to(accelerator.device)
+            text_encoders[0].to(accelerator.device, dtype=weight_dtype)
+            text_encoders[1].to(accelerator.device, dtype=weight_dtype)
 
     def get_text_cond(self, args, accelerator, batch, tokenizers, text_encoders, weight_dtype):
         if "text_encoder_outputs1_list" not in batch or batch["text_encoder_outputs1_list"] is None:

train_network.py

@@ -117,7 +117,7 @@ def cache_text_encoder_outputs_if_needed(
         self, args, accelerator, unet, vae, tokenizers, text_encoders, data_loader, weight_dtype
     ):
         for t_enc in text_encoders:
-            t_enc.to(accelerator.device)
+            t_enc.to(accelerator.device, dtype=weight_dtype)
 
     def get_text_cond(self, args, accelerator, batch, tokenizers, text_encoders, weight_dtype):
         input_ids = batch["input_ids"].to(accelerator.device)
@@ -278,6 +278,7 @@ def train(self, args):
             accelerator.wait_for_everyone()
 
         # 必要ならテキストエンコーダーの出力をキャッシュする: Text Encoderはcpuまたはgpuへ移される
+        # cache text encoder outputs if needed: Text Encoder is moved to cpu or gpu
         self.cache_text_encoder_outputs_if_needed(
             args, accelerator, unet, vae, tokenizers, text_encoders, train_dataset_group, weight_dtype
         )
@@ -394,8 +395,7 @@ def train(self, args):
         for t_enc in text_encoders:
             t_enc.requires_grad_(False)
 
-        # acceleratorがなんかよろしくやってくれるらしい
-        # TODO めちゃくちゃ冗長なのでコードを整理する
+        # acceleratorがなんかよろしくやってくれるらしい / accelerator will do something good
         if train_unet:
             unet = accelerator.prepare(unet)
         else:
@@ -407,8 +407,8 @@ def train(self, args):
                 text_encoder = accelerator.prepare(text_encoder)
                 text_encoders = [text_encoder]
         else:
-            for t_enc in text_encoders:
-                t_enc.to(accelerator.device, dtype=weight_dtype)
+            pass  # if text_encoder is not trained, no need to prepare. and device and dtype are already set
+
         network, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(network, optimizer, train_dataloader, lr_scheduler)
 
         if args.gradient_checkpointing:
@@ -685,7 +685,7 @@ def train(self, args):
         if accelerator.is_main_process:
             init_kwargs = {}
             if args.wandb_run_name:
-                init_kwargs['wandb'] = {'name': args.wandb_run_name}
+                init_kwargs["wandb"] = {"name": args.wandb_run_name}
             if args.log_tracker_config is not None:
                 init_kwargs = toml.load(args.log_tracker_config)
             accelerator.init_trackers(