Minor changes

flux/README.md

@@ -21,8 +21,9 @@ The dependencies are minimal, namely:
 
 - `huggingface-hub` to download the checkpoints.
 - `regex` for the tokenization
-- `tqdm`, `PIL`, and `numpy` for the `txt2image.py` script
+- `tqdm`, `PIL`, and `numpy` for the scripts
 - `sentencepiece` for the T5 tokenizer
+- `datasets` for using an HF dataset directly
 
 You can install all of the above with the `requirements.txt` as follows:
 

flux/flux/datasets.py

@@ -5,51 +5,69 @@ from PIL import Image
 
 
 class Dataset:
-    def __init__(self, data):
+    def __getitem__(self, index: int):
-        self._data = data
+        raise NotImplementedError()
 
     def __len__(self):
-        if self._data is None:
+        raise NotImplementedError()
-            return 0
-        return len(self._data)
 
 
 class LocalDataset(Dataset):
+    prompt_key = "prompt"
 
     def __init__(self, dataset: str, data_file):
         self.dataset_base = Path(dataset)
         with open(data_file, "r") as fid:
             self._data = [json.loads(l) for l in fid]
 
-        super().__init__(self._data)
+    def __len__(self):
+        return len(self._data)
 
     def __getitem__(self, index: int):
         item = self._data[index]
         image = Image.open(self.dataset_base / item["image"])
-        return image, item["prompt"]
+        return image, item[self.prompt_key]
+
+
+class LegacyDataset(LocalDataset):
+    prompt_key = "text"
+
+    def __init__(self, dataset: str):
+        self.dataset_base = Path(dataset)
+        with open(self.dataset_base / "index.json") as f:
+            self._data = json.load(f)["data"]
 
 
 class HuggingFaceDataset(Dataset):
 
     def __init__(self, dataset: str):
-        from datasets import load_dataset
+        from datasets import load_dataset as hf_load_dataset
 
-        df = load_dataset(dataset)["train"]
+        self._df = hf_load_dataset(dataset)["train"]
-        self._data = df.data
+
-        super().__init__(df)
+    def __len__(self):
+        return len(self._df)
 
     def __getitem__(self, index: int):
-        item = self._data[index]
+        item = self._df[index]
         return item["image"], item["prompt"]
 
 
 def load_dataset(dataset: str):
     dataset_base = Path(dataset)
     data_file = dataset_base / "train.jsonl"
+    legacy_file = dataset_base / "index.json"
 
     if data_file.exists():
         print(f"Load the local dataset {data_file} .", flush=True)
         dataset = LocalDataset(dataset, data_file)
+    elif legacy_file.exists():
+        print(f"Load the local dataset {legacy_file} .")
+        print()
+        print("     WARNING: 'index.json' is deprecated in favor of 'train.jsonl'.")
+        print("              See the README for details.")
+        print(flush=True)
+        dataset = LegacyDataset(dataset)
     else:
         print(f"Load the Hugging Face dataset {dataset} .", flush=True)
         dataset = HuggingFaceDataset(dataset)

flux/flux/trainer.py

@@ -26,15 +26,15 @@ class Trainer:
         # Random crop the input image between 0.8 to 1.0 of its original dimensions
         crop_size = (
             max((0.8 + 0.2 * a) * width, resolution[0]),
-            max((0.8 + 0.2 * a) * height, resolution[1]),
+            max((0.8 + 0.2 * b) * height, resolution[1]),
         )
         pan = (width - crop_size[0], height - crop_size[1])
         img = img.crop(
             (
-                pan[0] * b,
+                pan[0] * c,
-                pan[1] * c,
+                pan[1] * d,
-                crop_size[0] + pan[0] * b,
+                crop_size[0] + pan[0] * c,
-                crop_size[1] + pan[1] * c,
+                crop_size[1] + pan[1] * d,
             )
         )