Download optimizations (#2786)

* download_model_files metadata writing improvement

* line swap

* reduce line length

* safer download and greater block size

* Minor changes by pycodestyle

---------

Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>
This commit is contained in:
Peter Sofronas 2023-06-20 22:14:18 -04:00 committed by GitHub
parent 447569e31a
commit b22c7199c9
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -77,7 +77,6 @@ class ModelDownloader:
if os.getenv('HF_USER') is not None and os.getenv('HF_PASS') is not None:
self.s.auth = (os.getenv('HF_USER'), os.getenv('HF_PASS'))
def sanitize_model_and_branch_names(self, model, branch):
if model[-1] == '/':
model = model[:-1]
@ -92,7 +91,6 @@ class ModelDownloader:
return model, branch
def get_download_links_from_huggingface(self, model, branch, text_only=False):
base = "https://huggingface.co"
page = f"/api/models/{model}/tree/{branch}"
@ -163,7 +161,6 @@ class ModelDownloader:
return links, sha256, is_lora
def get_output_folder(self, model, branch, is_lora, base_folder=None):
if base_folder is None:
base_folder = 'models' if not is_lora else 'loras'
@ -174,10 +171,11 @@ class ModelDownloader:
output_folder = Path(base_folder) / output_folder
return output_folder
def get_single_file(self, url, output_folder, start_from_scratch=False):
filename = Path(url.rsplit('/', 1)[1])
output_path = output_folder / filename
headers = {}
mode = 'wb'
if output_path.exists() and not start_from_scratch:
# Check if the file has already been downloaded completely
r = self.s.get(url, stream=True, timeout=20)
@ -187,50 +185,45 @@ class ModelDownloader:
# Otherwise, resume the download from where it left off
headers = {'Range': f'bytes={output_path.stat().st_size}-'}
mode = 'ab'
else:
headers = {}
mode = 'wb'
r = self.s.get(url, stream=True, headers=headers, timeout=20)
with open(output_path, mode) as f:
with self.s.get(url, stream=True, headers=headers, timeout=20) as r:
r.raise_for_status() # Do not continue the download if the request was unsuccessful
total_size = int(r.headers.get('content-length', 0))
# Every 4MB we report an update
block_size = 4*1024*1024
with tqdm.tqdm(total=total_size, unit='iB', unit_scale=True, bar_format='{l_bar}{bar}| {n_fmt:6}/{total_fmt:6} {rate_fmt:6}') as t:
count = 0
for data in r.iter_content(block_size):
t.update(len(data))
f.write(data)
if self.progress_bar is not None:
count += len(data)
self.progress_bar(float(count)/float(total_size), f"Downloading {filename}")
block_size = 1024 * 1024 # 1MB
with open(output_path, mode) as f:
with tqdm.tqdm(total=total_size,
unit='iB',
unit_scale=True,
bar_format='{l_bar}{bar}| {n_fmt:6}/{total_fmt:6} {rate_fmt:6}'
) as t:
count = 0
for data in r.iter_content(block_size):
t.update(len(data))
f.write(data)
if self.progress_bar is not None:
count += len(data)
self.progress_bar(float(count) / float(total_size), f"Downloading {filename}")
def start_download_threads(self, file_list, output_folder, start_from_scratch=False, threads=1):
thread_map(lambda url: self.get_single_file(url, output_folder, start_from_scratch=start_from_scratch), file_list, max_workers=threads, disable=True)
def download_model_files(self, model, branch, links, sha256, output_folder, progress_bar = None, start_from_scratch=False, threads=1):
def download_model_files(self, model, branch, links, sha256, output_folder, progress_bar=None, start_from_scratch=False, threads=1):
self.progress_bar = progress_bar
# Creating the folder and writing the metadata
if not output_folder.exists():
output_folder.mkdir(parents=True, exist_ok=True)
with open(output_folder / 'huggingface-metadata.txt', 'w') as f:
f.write(f'url: https://huggingface.co/{model}\n')
f.write(f'branch: {branch}\n')
f.write(f'download date: {str(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S"))}\n')
sha256_str = ''
for i in range(len(sha256)):
sha256_str += f' {sha256[i][1]} {sha256[i][0]}\n'
if sha256_str != '':
f.write(f'sha256sum:\n{sha256_str}')
output_folder.mkdir(parents=True, exist_ok=True)
metadata = f'url: https://huggingface.co/{model}\n' \
f'branch: {branch}\n' \
f'download date: {datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")}\n'
sha256_str = '\n'.join([f' {item[1]} {item[0]}' for item in sha256])
if sha256_str:
metadata += f'sha256sum:\n{sha256_str}'
metadata += '\n'
(output_folder / 'huggingface-metadata.txt').write_text(metadata)
# Downloading the files
print(f"Downloading the model to {output_folder}")
self.start_download_threads(links, output_folder, start_from_scratch=start_from_scratch, threads=threads)
def check_model_files(self, model, branch, links, sha256, output_folder):
# Validate the checksums
validated = True