Cannot load model or tokenizer
#33
by IJSDerbashi - opened
Good day!
Last Friday, I have loaded the model without problems, and did some tests. Today (Monday) I am not able to run the same code.
Additionally, I ran the code in the readme, and the suggested Colab cell verbatim. Everytime I get the same error, whether I load tokeniser after the model or model after tokenizer.
The error is:
tokenizer_config.json: 100%
1.16M/1.16M [00:00<00:00, 8.50MB/s]
tokenizer.model: 0%
0.00/4.69M [00:28<?, ?B/s]
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
/usr/local/lib/python3.12/dist-packages/transformers/tokenization_utils_base.py in from_pretrained(cls, pretrained_model_name_or_path, cache_dir, force_download, local_files_only, token, revision, trust_remote_code, *init_inputs, **kwargs)
2059 try:
-> 2060 resolved_vocab_files[file_id] = cached_file(
2061 pretrained_model_name_or_path,
10 frames
/usr/local/lib/python3.12/dist-packages/transformers/utils/hub.py in cached_file(path_or_repo_id, filename, **kwargs)
321 """
--> 322 file = cached_files(path_or_repo_id=path_or_repo_id, filenames=[filename], **kwargs)
323 file = file[0] if file is not None else file
/usr/local/lib/python3.12/dist-packages/transformers/utils/hub.py in cached_files(path_or_repo_id, filenames, cache_dir, force_download, resume_download, proxies, token, revision, local_files_only, subfolder, repo_type, user_agent, _raise_exceptions_for_gated_repo, _raise_exceptions_for_missing_entries, _raise_exceptions_for_connection_errors, _commit_hash, **deprecated_kwargs)
566 elif not isinstance(e, EntryNotFoundError):
--> 567 raise e
568
/usr/local/lib/python3.12/dist-packages/transformers/utils/hub.py in cached_files(path_or_repo_id, filenames, cache_dir, force_download, resume_download, proxies, token, revision, local_files_only, subfolder, repo_type, user_agent, _raise_exceptions_for_gated_repo, _raise_exceptions_for_missing_entries, _raise_exceptions_for_connection_errors, _commit_hash, **deprecated_kwargs)
478 # This is slightly better for only 1 file
--> 479 hf_hub_download(
480 path_or_repo_id,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/utils/_validators.py in _inner_fn(*args, **kwargs)
113
--> 114 return fn(*args, **kwargs)
115
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in hf_hub_download(repo_id, filename, subfolder, repo_type, revision, library_name, library_version, cache_dir, local_dir, user_agent, force_download, proxies, etag_timeout, token, local_files_only, headers, endpoint, resume_download, force_filename, local_dir_use_symlinks)
1009 else:
-> 1010 return _hf_hub_download_to_cache_dir(
1011 # Destination
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in _hf_hub_download_to_cache_dir(cache_dir, repo_id, filename, repo_type, revision, endpoint, etag_timeout, headers, proxies, token, local_files_only, force_download)
1170 with WeakFileLock(lock_path):
-> 1171 _download_to_tmp_and_move(
1172 incomplete_path=Path(blob_path + ".incomplete"),
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in _download_to_tmp_and_move(incomplete_path, destination_path, url_to_download, proxies, headers, expected_size, filename, force_download, etag, xet_file_data)
1722 logger.debug("Xet Storage is enabled for this repo. Downloading file from Xet Storage..")
-> 1723 xet_get(
1724 incomplete_path=incomplete_path,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in xet_get(incomplete_path, xet_file_data, headers, expected_size, displayed_filename, _tqdm_bar)
628
--> 629 download_files(
630 xet_download_info,
RuntimeError: Data processing error: CAS service error : Reqwest Error: HTTP status server error (500 Internal Server Error), domain: https://cas-server.xethub.hf.co/reconstructions/a81fa217b67ef4a1992b48a47651c27a2a19df419eafd1aad9c0bbd5ff49bde3
During handling of the above exception, another exception occurred:
OSError Traceback (most recent call last)
/tmp/ipython-input-1108580225.py in <cell line: 0>()
2 from transformers import AutoTokenizer, AutoModelForCausalLM
3
----> 4 tokenizer = AutoTokenizer.from_pretrained("google/gemma-3-1b-it")
5 model = AutoModelForCausalLM.from_pretrained("google/gemma-3-1b-it")
6 messages = [
/usr/local/lib/python3.12/dist-packages/transformers/models/auto/tokenization_auto.py in from_pretrained(cls, pretrained_model_name_or_path, *inputs, **kwargs)
1138 f"Tokenizer class {tokenizer_class_candidate} does not exist or is not currently imported."
1139 )
-> 1140 return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
1141
1142 # Otherwise we have to be creative.
/usr/local/lib/python3.12/dist-packages/transformers/tokenization_utils_base.py in from_pretrained(cls, pretrained_model_name_or_path, cache_dir, force_download, local_files_only, token, revision, trust_remote_code, *init_inputs, **kwargs)
2078 except Exception:
2079 # For any other exception, we throw a generic error.
-> 2080 raise OSError(
2081 f"Can't load tokenizer for '{pretrained_model_name_or_path}'. If you were trying to load it from "
2082 "'https://huggingface.co/models', make sure you don't have a local directory with the same name. "
OSError: Can't load tokenizer for 'google/gemma-3-1b-it'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'google/gemma-3-1b-it' is the correct path to a directory containing all relevant files for a GemmaTokenizerFast tokenizer.
This is a fresh Colab runtime so there are no directories with that name.
Additionally I tried to download the dataset manually using:
from huggingface_hub import snapshot_download
model_id = "google/gemma-3-1b-it"
local_dir = "/models/gemma3_1b"
snapshot_download(
model_id,
local_dir=local_dir,
max_workers=8,
)
And I get back the following:
Fetching 10 files: 50%
5/10 [01:15<00:03, 1.66it/s]
generation_config.json: 100%
215/215 [00:00<00:00, 3.49kB/s]
.gitattributes: 100%
1.68k/1.68k [00:00<00:00, 34.4kB/s]
README.md: 100%
24.3k/24.3k [00:00<00:00, 302kB/s]
added_tokens.json: 100%
35.0/35.0 [00:00<00:00, 379B/s]
config.json: 100%
899/899 [00:00<00:00, 12.1kB/s]
special_tokens_map.json: 100%
662/662 [00:00<00:00, 20.4kB/s]
tokenizer.json: 0%
0.00/33.4M [01:02<?, ?B/s]
model.safetensors: 0%
0.00/2.00G [01:14<?, ?B/s]
tokenizer_config.json: 100%
1.16M/1.16M [00:00<00:00, 2.58MB/s]
tokenizer.model: 0%
0.00/4.69M [00:26<?, ?B/s]
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
/tmp/ipython-input-1573216716.py in <cell line: 0>()
4
5 local_dir = "/models/gemma3_1b"
----> 6 snapshot_download(
7 model_id,
8 local_dir=local_dir,
16 frames
/usr/local/lib/python3.12/dist-packages/huggingface_hub/utils/_validators.py in _inner_fn(*args, **kwargs)
112 kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=has_token, kwargs=kwargs)
113
--> 114 return fn(*args, **kwargs)
115
116 return _inner_fn # type: ignore
/usr/local/lib/python3.12/dist-packages/huggingface_hub/_snapshot_download.py in snapshot_download(repo_id, repo_type, revision, cache_dir, local_dir, library_name, library_version, user_agent, proxies, etag_timeout, force_download, token, local_files_only, allow_patterns, ignore_patterns, max_workers, tqdm_class, headers, endpoint, local_dir_use_symlinks, resume_download)
330 _inner_hf_hub_download(file)
331 else:
--> 332 thread_map(
333 _inner_hf_hub_download,
334 filtered_repo_files,
/usr/local/lib/python3.12/dist-packages/tqdm/contrib/concurrent.py in thread_map(fn, *iterables, **tqdm_kwargs)
67 """
68 from concurrent.futures import ThreadPoolExecutor
---> 69 return _executor_map(ThreadPoolExecutor, fn, *iterables, **tqdm_kwargs)
70
71
/usr/local/lib/python3.12/dist-packages/tqdm/contrib/concurrent.py in _executor_map(PoolExecutor, fn, *iterables, **tqdm_kwargs)
49 with PoolExecutor(max_workers=max_workers, initializer=tqdm_class.set_lock,
50 initargs=(lk,)) as ex:
---> 51 return list(tqdm_class(ex.map(fn, *iterables, chunksize=chunksize), **kwargs))
52
53
/usr/local/lib/python3.12/dist-packages/tqdm/notebook.py in __iter__(self)
248 try:
249 it = super().__iter__()
--> 250 for obj in it:
251 # return super(tqdm...) will not catch exception
252 yield obj
/usr/local/lib/python3.12/dist-packages/tqdm/std.py in __iter__(self)
1179
1180 try:
-> 1181 for obj in iterable:
1182 yield obj
1183 # Update and possibly print the progressbar.
/usr/lib/python3.12/concurrent/futures/_base.py in result_iterator()
617 # Careful not to keep a reference to the popped future
618 if timeout is None:
--> 619 yield _result_or_cancel(fs.pop())
620 else:
621 yield _result_or_cancel(fs.pop(), end_time - time.monotonic())
/usr/lib/python3.12/concurrent/futures/_base.py in _result_or_cancel(***failed resolving arguments***)
315 try:
316 try:
--> 317 return fut.result(timeout)
318 finally:
319 fut.cancel()
/usr/lib/python3.12/concurrent/futures/_base.py in result(self, timeout)
454 raise CancelledError()
455 elif self._state == FINISHED:
--> 456 return self.__get_result()
457 else:
458 raise TimeoutError()
/usr/lib/python3.12/concurrent/futures/_base.py in __get_result(self)
399 if self._exception:
400 try:
--> 401 raise self._exception
402 finally:
403 # Break a reference cycle with the exception in self._exception
/usr/lib/python3.12/concurrent/futures/thread.py in run(self)
57
58 try:
---> 59 result = self.fn(*self.args, **self.kwargs)
60 except BaseException as exc:
61 self.future.set_exception(exc)
/usr/local/lib/python3.12/dist-packages/huggingface_hub/_snapshot_download.py in _inner_hf_hub_download(repo_file)
304 # have the file locally.
305 def _inner_hf_hub_download(repo_file: str):
--> 306 return hf_hub_download(
307 repo_id,
308 filename=repo_file,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/utils/_validators.py in _inner_fn(*args, **kwargs)
112 kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=has_token, kwargs=kwargs)
113
--> 114 return fn(*args, **kwargs)
115
116 return _inner_fn # type: ignore
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in hf_hub_download(repo_id, filename, subfolder, repo_type, revision, library_name, library_version, cache_dir, local_dir, user_agent, force_download, proxies, etag_timeout, token, local_files_only, headers, endpoint, resume_download, force_filename, local_dir_use_symlinks)
988 )
989
--> 990 return _hf_hub_download_to_local_dir(
991 # Destination
992 local_dir=local_dir,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in _hf_hub_download_to_local_dir(local_dir, repo_id, repo_type, filename, revision, endpoint, etag_timeout, headers, proxies, token, cache_dir, force_download, local_files_only)
1298 with WeakFileLock(paths.lock_path):
1299 paths.file_path.unlink(missing_ok=True) # delete outdated file first
-> 1300 _download_to_tmp_and_move(
1301 incomplete_path=paths.incomplete_path(etag),
1302 destination_path=paths.file_path,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in _download_to_tmp_and_move(incomplete_path, destination_path, url_to_download, proxies, headers, expected_size, filename, force_download, etag, xet_file_data)
1721 if xet_file_data is not None and is_xet_available():
1722 logger.debug("Xet Storage is enabled for this repo. Downloading file from Xet Storage..")
-> 1723 xet_get(
1724 incomplete_path=incomplete_path,
1725 xet_file_data=xet_file_data,
/usr/local/lib/python3.12/dist-packages/huggingface_hub/file_download.py in xet_get(incomplete_path, xet_file_data, headers, expected_size, displayed_filename, _tqdm_bar)
627 progress.update(progress_bytes)
628
--> 629 download_files(
630 xet_download_info,
631 endpoint=connection_info.endpoint,
RuntimeError: Data processing error: CAS service error : Reqwest Error: HTTP status server error (500 Internal Server Error), domain: https://cas-server.xethub.hf.co/reconstructions/7473bad0373d89455623d1039c446aa729eb9d5a5b3a7e6316e188cee4e50a40
Can you assist with that?
UPDATE: This seemed to be an AWS error affecting the data transmission from HF.
In my case, this was fixed in ~4hours after the AWS outage was resolved.
IJSDerbashi changed discussion status to closed