Hello,
at the very begin of the notebook (on the second executable cell) I get the following error and cannot proceed any further with the assignment.
“No sentence-transformers model found with name BAAI/bge-base-en-v1.5. Creating a new one with mean pooling.”
I already tried multiple times to reboot kernel, retboot server and get latest version but nothing seems to work
:
Hi @bottarelli.lorenzo , welcome to the RAG course support zone 
I’ve launched the C1M3_Assignment myself and currently the issue you noticed is not reproducible, so could be some temporary issues.. Notice you have both HF_HUB_OFFLINE and TRANSFORMERS_OFFLINE set to 1 (which is a default). I believe these env. variables control the caching mechanism. I’ve seen issues when setting these to 0 (but in your case this is not probably a case).
Can you please try to go to right top menu of the Jupyter notebook and restore the default environment and try again. I think that should resolve some problem with the existing data.
Let me pls. know if this has resolved it, I cold raise the issue that is visible to me back to developers still.
can you post the screenshot of full error, we did have issues labs, but updated just yesterday. This is seems to be completely new error especially with Local entry not found error.
as @ondrous mentioned Restoring to original version should get you the updated notebook. As another checkpoint, sometimes what I do, delete the old assignment file, and reboot the environment, so no previous cache files corrupt the new updated environment for the notebook codes to run successfully. Clearing your cache and browsing history also is better step before you restore updated notebook.
Clock on 3 dots on right upper corner and click Restore original version.
I have already tried multiple times restoring of the original version. I also deleted all the content of my lab to be sure that the restoring was creating all new files, but nothing seams to work.
As for the full error it is way too much to fit in a screenshot, here is a full copy:
No sentence-transformers model found with name BAAI/bge-base-en-v1.5. Creating a new one with mean pooling.
---------------------------------------------------------------------------
LocalEntryNotFoundError Traceback (most recent call last)
File /usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py:470, 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)
468 if len(full_filenames) == 1:
469 # This is slightly better for only 1 file
--> 470 hf_hub_download(
471 path_or_repo_id,
472 filenames[0],
473 subfolder=None if len(subfolder) == 0 else subfolder,
474 repo_type=repo_type,
475 revision=revision,
476 cache_dir=cache_dir,
477 user_agent=user_agent,
478 force_download=force_download,
479 proxies=proxies,
480 resume_download=resume_download,
481 token=token,
482 local_files_only=local_files_only,
483 )
484 else:
File /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py:114, in validate_hf_hub_args.<locals>._inner_fn(*args, **kwargs)
112 kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=has_token, kwargs=kwargs)
--> 114 return fn(*args, **kwargs)
File /usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py:1008, 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)
1007 else:
-> 1008 return _hf_hub_download_to_cache_dir(
1009 # Destination
1010 cache_dir=cache_dir,
1011 # File info
1012 repo_id=repo_id,
1013 filename=filename,
1014 repo_type=repo_type,
1015 revision=revision,
1016 # HTTP info
1017 endpoint=endpoint,
1018 etag_timeout=etag_timeout,
1019 headers=hf_headers,
1020 proxies=proxies,
1021 token=token,
1022 # Additional options
1023 local_files_only=local_files_only,
1024 force_download=force_download,
1025 )
File /usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py:1115, 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)
1114 # Otherwise, raise appropriate error
-> 1115 _raise_on_head_call_error(head_call_error, force_download, local_files_only)
1117 # From now on, etag, commit_hash, url and size are not None.
File /usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py:1636, in _raise_on_head_call_error(head_call_error, force_download, local_files_only)
1635 if local_files_only:
-> 1636 raise LocalEntryNotFoundError(
1637 "Cannot find the requested files in the disk cache and outgoing traffic has been disabled. To enable"
1638 " hf.co look-ups and downloads online, set 'local_files_only' to False."
1639 )
1640 elif isinstance(head_call_error, (RepositoryNotFoundError, GatedRepoError)) or (
1641 isinstance(head_call_error, HfHubHTTPError) and head_call_error.response.status_code == 401
1642 ):
1643 # Repo not found or gated => let's raise the actual error
1644 # Unauthorized => likely a token issue => let's raise the actual error
LocalEntryNotFoundError: Cannot find the requested files in the disk cache and outgoing traffic has been disabled. To enable hf.co look-ups and downloads online, set 'local_files_only' to False.
The above exception was the direct cause of the following exception:
OSError Traceback (most recent call last)
Cell In[2], line 1
----> 1 from utils import (
2 generate_with_single_input,
3 print_object_properties,
4 display_widget,
5 kill_processes_on_ports
6 )
7 import unittests
9 # Kill processes on ports before importing flask_app and weaviate_server
10 # WARNING: Running this cell twice may kill the active kernel
File ~/work/utils.py:13
10 from sentence_transformers import SentenceTransformer
12 # Load a pretrained model from Hugging Face
---> 13 model = SentenceTransformer("BAAI/bge-base-en-v1.5", cache_folder = ".models")
15 # Custom transport to bypass SSL verification
16 transport = httpx.HTTPTransport(local_address="0.0.0.0", verify=False)
File /usr/local/lib/python3.10/dist-packages/sentence_transformers/SentenceTransformer.py:339, in SentenceTransformer.__init__(self, model_name_or_path, modules, device, prompts, default_prompt_name, similarity_fn_name, cache_folder, trust_remote_code, revision, local_files_only, token, use_auth_token, truncate_dim, model_kwargs, tokenizer_kwargs, config_kwargs, model_card_data, backend)
327 modules, self.module_kwargs = self._load_sbert_model(
328 model_name_or_path,
329 token=token,
(...)
336 config_kwargs=config_kwargs,
337 )
338 else:
--> 339 modules = self._load_auto_model(
340 model_name_or_path,
341 token=token,
342 cache_folder=cache_folder,
343 revision=revision,
344 trust_remote_code=trust_remote_code,
345 local_files_only=local_files_only,
346 model_kwargs=model_kwargs,
347 tokenizer_kwargs=tokenizer_kwargs,
348 config_kwargs=config_kwargs,
349 has_modules=has_modules,
350 )
352 if modules is not None and not isinstance(modules, OrderedDict):
353 modules = OrderedDict([(str(idx), module) for idx, module in enumerate(modules)])
File /usr/local/lib/python3.10/dist-packages/sentence_transformers/SentenceTransformer.py:2061, in SentenceTransformer._load_auto_model(self, model_name_or_path, token, cache_folder, revision, trust_remote_code, local_files_only, model_kwargs, tokenizer_kwargs, config_kwargs, has_modules)
2058 tokenizer_kwargs = shared_kwargs if tokenizer_kwargs is None else {**shared_kwargs, **tokenizer_kwargs}
2059 config_kwargs = shared_kwargs if config_kwargs is None else {**shared_kwargs, **config_kwargs}
-> 2061 transformer_model = Transformer(
2062 model_name_or_path,
2063 cache_dir=cache_folder,
2064 model_args=model_kwargs,
2065 tokenizer_args=tokenizer_kwargs,
2066 config_args=config_kwargs,
2067 backend=self.backend,
2068 )
2069 pooling_model = Pooling(transformer_model.get_word_embedding_dimension(), "mean")
2070 if not local_files_only:
File /usr/local/lib/python3.10/dist-packages/sentence_transformers/models/Transformer.py:87, in Transformer.__init__(self, model_name_or_path, max_seq_length, model_args, tokenizer_args, config_args, cache_dir, do_lower_case, tokenizer_name_or_path, backend)
84 if config_args is None:
85 config_args = {}
---> 87 config, is_peft_model = self._load_config(model_name_or_path, cache_dir, backend, config_args)
88 self._load_model(model_name_or_path, config, cache_dir, backend, is_peft_model, **model_args)
90 if max_seq_length is not None and "model_max_length" not in tokenizer_args:
File /usr/local/lib/python3.10/dist-packages/sentence_transformers/models/Transformer.py:152, in Transformer._load_config(self, model_name_or_path, cache_dir, backend, config_args)
148 from peft import PeftConfig
150 return PeftConfig.from_pretrained(model_name_or_path, **config_args, cache_dir=cache_dir), True
--> 152 return AutoConfig.from_pretrained(model_name_or_path, **config_args, cache_dir=cache_dir), False
File /usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py:1197, in AutoConfig.from_pretrained(cls, pretrained_model_name_or_path, **kwargs)
1194 trust_remote_code = kwargs.pop("trust_remote_code", None)
1195 code_revision = kwargs.pop("code_revision", None)
-> 1197 config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs)
1198 has_remote_code = "auto_map" in config_dict and "AutoConfig" in config_dict["auto_map"]
1199 has_local_code = "model_type" in config_dict and config_dict["model_type"] in CONFIG_MAPPING
File /usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py:608, in PretrainedConfig.get_config_dict(cls, pretrained_model_name_or_path, **kwargs)
606 original_kwargs = copy.deepcopy(kwargs)
607 # Get config dict associated with the base config file
--> 608 config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs)
609 if config_dict is None:
610 return {}, kwargs
File /usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py:667, in PretrainedConfig._get_config_dict(cls, pretrained_model_name_or_path, **kwargs)
663 configuration_file = kwargs.pop("_configuration_file", CONFIG_NAME) if gguf_file is None else gguf_file
665 try:
666 # Load from local folder or from cache or download from model Hub and cache
--> 667 resolved_config_file = cached_file(
668 pretrained_model_name_or_path,
669 configuration_file,
670 cache_dir=cache_dir,
671 force_download=force_download,
672 proxies=proxies,
673 resume_download=resume_download,
674 local_files_only=local_files_only,
675 token=token,
676 user_agent=user_agent,
677 revision=revision,
678 subfolder=subfolder,
679 _commit_hash=commit_hash,
680 )
681 if resolved_config_file is None:
682 return None, kwargs
File /usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py:312, in cached_file(path_or_repo_id, filename, **kwargs)
254 def cached_file(
255 path_or_repo_id: Union[str, os.PathLike],
256 filename: str,
257 **kwargs,
258 ) -> Optional[str]:
259 """
260 Tries to locate a file in a local folder and repo, downloads and cache it if necessary.
261
(...)
310 ```
311 """
--> 312 file = cached_files(path_or_repo_id=path_or_repo_id, filenames=[filename], **kwargs)
313 file = file[0] if file is not None else file
314 return file
File /usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py:543, 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)
540 # Here we only raise if both flags for missing entry and connection errors are True (because it can be raised
541 # even when `local_files_only` is True, in which case raising for connections errors only would not make sense)
542 elif _raise_exceptions_for_missing_entries:
--> 543 raise OSError(
544 f"We couldn't connect to '{HUGGINGFACE_CO_RESOLVE_ENDPOINT}' to load the files, and couldn't find them in the"
545 f" cached files.\nCheck your internet connection or see how to run the library in offline mode at"
546 " 'https://huggingface.co/docs/transformers/installation#offline-mode'."
547 ) from e
548 # snapshot_download will not raise EntryNotFoundError, but hf_hub_download can. If this is the case, it will be treated
549 # later on anyway and re-raised if needed
550 elif isinstance(e, HTTPError) and not isinstance(e, EntryNotFoundError):
OSError: We couldn't connect to 'https://huggingface.co' to load the files, and couldn't find them in the cached files.
Check your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'.
I just ran down the codes after restoring original version, i didn’t encounter any error, please see the screenshot 
Notice in the error feedback mentions there could be internet connection issue,
Could you try a different browser, and follow these steps.
Let me know if you still encounter this error.
regards
Dr. Deepti
Sorry I forgot to mention and maybe it is a foundamental information, but I am doing this course on Coursera. Maybe the problem is not solved yet on that platform ?
That’s understandable with the confusion about platform.
It was informed to me last night only that all labs in Coursera as well as dlai learning platform including ungraded labs were updated. I also have run down lab for module 3 assignment successfully after restoring to original version as you can see here.
for coursera, please follow these steps.
First delete your cache and browsing history from your system/laptop.
Re-login in the Coursera, open notebook from your classroom page.
Delete the assignment file, by clicking on File==> Open section.
Click 
, then click Reboot, then click Get the Latest Version.
Again wait until the notebook loads in the environment and do not switch browsers.
Once loaded, connect your kernel.
Start running down cells individually one by one.
Remember the import cells usually takes time to run down completely, so wait for the cell to down completely, do not double click the running down cell, if you see * for longer time, as this kill the kernel and throws error.
Please try these steps and let me know.
Regards
Dr. Deepti.
I have tried a couple of times the steps you suggested but unfortunately the problem persist
i will inform staff and get back to you.
I am also having the exact same issue. DO we have a solution or a workaround?
kindly wait for staff update.
Hi! I fixed this issue in Coursera today earlier, you will need to refresh your workspace in order to get the new utils.py file.
Let me know if this helps you.
Sorry but it doesn’t work still.
Just to be clear:
Rebooted
Got the latest version. Now all files and folders have been regenated to the latest version. See screenshot
The problem still persist. See screenshot
I tried refreshing the workspace and the problem still persists
Kind reminder that the problem is still there @Deepti_Prasad @lucas.coutinho
If it would help my Lab ID is omsckaqsrypy
I am able to run the code successfully on dlai, sadly I am not able to have access on coursera right for this course to know why it is causing issue still for some learners.
But I wanted to try as best I could help.
So when I checked your earlier detailed error, it is pointing it couldn’t find any cache files to download the model when the import utils codes were run down.
I then checked utils.py file, to see if the model was downloaded offline, and it was.
So in dlai it shows it does download the cache file as cache folder, see the screenshot
So can you open the notebook from your classroom page, and open utils.py file, and take the same screenshot of the same part and post here, so I can check it there is any difference between the two, as to reasoning behind why you are still encountering this error.
Regards
Dr. Deepti
Ohh 
, thank you @mannyiyer for the image.
Then I think we will have to wait for @lucas.coutinho to look at this thread as I am still not able to open rag course on Coursera, i am unable to mimic the same error.
For sure the issue or error mention, to set local_dir to set to false, so the model can be downloaded online, that’s another way to work this code around, but lucas must have chosen this to be downloaded as offline for preventing killing of port processors.
So let us wait for lucas to respond.
If he doesn’t respond, one of you try to send the utils.py file from coursera via personal DM, I will check and try to look at the issue.
But I have 
guarantee lucas will resolve this issue when he will check your both comments.
Apologies for this setback. as courses and codes are dependent of multiple channels, changes at one end causes a constant correction, reflection on recurring issues.
Regards
Dr. Deepti
