changes to symlinks to account for hf stuff
Browse files- .gitattributes +4 -0
- dronescapes_reader/multitask_dataset.py +13 -4
- symlinks_from_txt_list.py +26 -0
.gitattributes
CHANGED
@@ -1,2 +1,6 @@
|
|
1 |
*.jpg filter=lfs diff=lfs merge=lfs -text
|
2 |
*.png filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
1 |
*.jpg filter=lfs diff=lfs merge=lfs -text
|
2 |
*.png filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
5 |
+
data/ filter=lfs diff=lfs merge=lfs -text
|
6 |
+
|
dronescapes_reader/multitask_dataset.py
CHANGED
@@ -128,13 +128,19 @@ class MultiTaskDataset(Dataset):
|
|
128 |
all_repr_dirs: list[str] = [x.name for x in self.path.iterdir() if x.is_dir()]
|
129 |
for repr_dir_name in all_repr_dirs:
|
130 |
dir_name = self.path / repr_dir_name
|
131 |
-
|
132 |
-
|
|
|
|
|
|
|
|
|
|
|
133 |
assert not any(len(x) == 0 for x in in_files.values()), f"{ [k for k, v in in_files.items() if len(v) == 0] }"
|
134 |
return in_files
|
135 |
|
136 |
def _build_dataset_drop(self) -> BuildDatasetTuple:
|
137 |
in_files = self._get_all_npz_files()
|
|
|
138 |
common = set(x.name for x in next(iter(in_files.values())))
|
139 |
nodes = in_files.keys()
|
140 |
for node in nodes:
|
@@ -142,11 +148,13 @@ class MultiTaskDataset(Dataset):
|
|
142 |
assert len(common) > 0, f"Node '{node}' made the intersection null"
|
143 |
common = natsorted(list(common))
|
144 |
logger.info(f"Found {len(common)} data points for each node ({len(nodes)} nodes).")
|
145 |
-
files_per_repr = {node: [
|
|
|
146 |
return files_per_repr, common
|
147 |
|
148 |
def _build_dataset_fill_none(self) -> BuildDatasetTuple:
|
149 |
in_files = self._get_all_npz_files()
|
|
|
150 |
all_files = set(x.name for x in next(iter(in_files.values())))
|
151 |
nodes = in_files.keys()
|
152 |
for node in nodes:
|
@@ -158,8 +166,9 @@ class MultiTaskDataset(Dataset):
|
|
158 |
in_file_names = {node: [f.name for f in in_files[node]] for node in nodes}
|
159 |
for node in nodes:
|
160 |
for file_name in all_files:
|
161 |
-
file_path =
|
162 |
files_per_repr[node].append(file_path)
|
|
|
163 |
return files_per_repr, all_files
|
164 |
|
165 |
def _build_dataset(self) -> BuildDatasetTuple:
|
|
|
128 |
all_repr_dirs: list[str] = [x.name for x in self.path.iterdir() if x.is_dir()]
|
129 |
for repr_dir_name in all_repr_dirs:
|
130 |
dir_name = self.path / repr_dir_name
|
131 |
+
if all(f.is_dir() for f in dir_name.iterdir()): # dataset is stored as repr/part_x/0.npz, ..., part_k/n.npz
|
132 |
+
all_files = []
|
133 |
+
for part in dir_name.iterdir():
|
134 |
+
all_files.extend(part.glob(f"*.{self.suffix}"))
|
135 |
+
else: # dataset is stored as repr/0.npz, ..., repr/n.npz
|
136 |
+
all_files = dir_name.glob(f"*.{self.suffix}")
|
137 |
+
in_files[repr_dir_name] = natsorted(all_files, key=lambda x: x.name) # important: use natsorted() here
|
138 |
assert not any(len(x) == 0 for x in in_files.values()), f"{ [k for k, v in in_files.items() if len(v) == 0] }"
|
139 |
return in_files
|
140 |
|
141 |
def _build_dataset_drop(self) -> BuildDatasetTuple:
|
142 |
in_files = self._get_all_npz_files()
|
143 |
+
name_to_node_path = {k: {_v.name: _v for _v in v} for k, v in in_files.items()} # {node: {name: path}}
|
144 |
common = set(x.name for x in next(iter(in_files.values())))
|
145 |
nodes = in_files.keys()
|
146 |
for node in nodes:
|
|
|
148 |
assert len(common) > 0, f"Node '{node}' made the intersection null"
|
149 |
common = natsorted(list(common))
|
150 |
logger.info(f"Found {len(common)} data points for each node ({len(nodes)} nodes).")
|
151 |
+
files_per_repr = {node: [name_to_node_path[node][x] for x in common] for node in nodes}
|
152 |
+
assert len(files_per_repr) > 0
|
153 |
return files_per_repr, common
|
154 |
|
155 |
def _build_dataset_fill_none(self) -> BuildDatasetTuple:
|
156 |
in_files = self._get_all_npz_files()
|
157 |
+
name_to_node_path = {k: {_v.name: _v for _v in v} for k, v in in_files.items()}
|
158 |
all_files = set(x.name for x in next(iter(in_files.values())))
|
159 |
nodes = in_files.keys()
|
160 |
for node in nodes:
|
|
|
166 |
in_file_names = {node: [f.name for f in in_files[node]] for node in nodes}
|
167 |
for node in nodes:
|
168 |
for file_name in all_files:
|
169 |
+
file_path = name_to_node_path[node].get(file_name, None)
|
170 |
files_per_repr[node].append(file_path)
|
171 |
+
assert len(files_per_repr) > 0
|
172 |
return files_per_repr, all_files
|
173 |
|
174 |
def _build_dataset(self) -> BuildDatasetTuple:
|
symlinks_from_txt_list.py
CHANGED
@@ -48,6 +48,30 @@ def check_and_gather_all_files(dataset_dir: Path, txt_data: list[tuple[str, str]
|
|
48 |
logger.info(f"Gathered {len(symlinks_to_do)} symlinks to create")
|
49 |
return symlinks_to_do
|
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
def read_txt_data(txt_file: Path) -> list[tuple[str, str]]:
|
52 |
"""reads the data from the txt file with format scene/stem"""
|
53 |
f = open(txt_file, "r")
|
@@ -66,6 +90,7 @@ def get_args() -> Namespace:
|
|
66 |
parser.add_argument("--output_path", "-o", type=lambda p: Path(p).absolute())
|
67 |
parser.add_argument("--overwrite", action="store_true")
|
68 |
parser.add_argument("--copy_files", action="store_true")
|
|
|
69 |
args = parser.parse_args()
|
70 |
assert not args.output_path.exists() or args.overwrite, f"'{args.output_path}' exists. Use --overwrite."
|
71 |
if args.output_path.exists():
|
@@ -80,6 +105,7 @@ def main(args: Namespace):
|
|
80 |
args.output_path.mkdir(exist_ok=False, parents=True)
|
81 |
txt_data = read_txt_data(args.txt_file)
|
82 |
symlinks_to_do = check_and_gather_all_files(args.dataset_dir, txt_data, args.output_path, suffix=".npz")
|
|
|
83 |
for in_file, out_file in tqdm(symlinks_to_do.items(), desc="symlinks"):
|
84 |
Path(out_file).parent.mkdir(exist_ok=True, parents=True)
|
85 |
if args.copy_files:
|
|
|
48 |
logger.info(f"Gathered {len(symlinks_to_do)} symlinks to create")
|
49 |
return symlinks_to_do
|
50 |
|
51 |
+
def make_partitions_if_needed(symlinks_to_do: dict[str, str], partition_max_size: int) -> dict[str, str]:
|
52 |
+
"""updated from out_dir/repr/0.npz to out_dir/repr/part_0/0.npz if needed"""
|
53 |
+
symlinks_by_repr = {} # gather as {repr: {in_file: out_file}}
|
54 |
+
for k, v in symlinks_to_do.items():
|
55 |
+
_repr = v.parent.name
|
56 |
+
if _repr not in symlinks_by_repr:
|
57 |
+
symlinks_by_repr[_repr] = {}
|
58 |
+
symlinks_by_repr[_repr][k] = v
|
59 |
+
|
60 |
+
new_symlinks_to_do = {}
|
61 |
+
for _repr, repr_files in symlinks_by_repr.items():
|
62 |
+
if (count := len(repr_files)) <= partition_max_size:
|
63 |
+
new_symlinks_to_do = {**new_symlinks_to_do, **repr_files}
|
64 |
+
else:
|
65 |
+
logger.info(f"Representation {_repr} has {count} items which > than {partition_max_size}. Partitioning.")
|
66 |
+
n_parts = (count // partition_max_size) + (count % partition_max_size != 0)
|
67 |
+
repr_files_as_tuple = tuple(repr_files.items())
|
68 |
+
for i in range(n_parts):
|
69 |
+
part = repr_files_as_tuple[i * partition_max_size: (i + 1) * partition_max_size]
|
70 |
+
for in_file, out_file in part: # add the partition subdir
|
71 |
+
new_symlinks_to_do[in_file] = out_file.parent / f"part{i}" / out_file.name
|
72 |
+
assert (a := len(new_symlinks_to_do)) == (b := len(symlinks_to_do)), (a, b)
|
73 |
+
return new_symlinks_to_do
|
74 |
+
|
75 |
def read_txt_data(txt_file: Path) -> list[tuple[str, str]]:
|
76 |
"""reads the data from the txt file with format scene/stem"""
|
77 |
f = open(txt_file, "r")
|
|
|
90 |
parser.add_argument("--output_path", "-o", type=lambda p: Path(p).absolute())
|
91 |
parser.add_argument("--overwrite", action="store_true")
|
92 |
parser.add_argument("--copy_files", action="store_true")
|
93 |
+
parser.add_argument("--partition_max_size", type=int, default=10000) # thanks huggingface for this
|
94 |
args = parser.parse_args()
|
95 |
assert not args.output_path.exists() or args.overwrite, f"'{args.output_path}' exists. Use --overwrite."
|
96 |
if args.output_path.exists():
|
|
|
105 |
args.output_path.mkdir(exist_ok=False, parents=True)
|
106 |
txt_data = read_txt_data(args.txt_file)
|
107 |
symlinks_to_do = check_and_gather_all_files(args.dataset_dir, txt_data, args.output_path, suffix=".npz")
|
108 |
+
symlinks_to_do = make_partitions_if_needed(symlinks_to_do, args.partition_max_size)
|
109 |
for in_file, out_file in tqdm(symlinks_to_do.items(), desc="symlinks"):
|
110 |
Path(out_file).parent.mkdir(exist_ok=True, parents=True)
|
111 |
if args.copy_files:
|