[bug]: "Unknown Qwen3 variant" Error when generating offline using Anima.

[PFJackalope33]

Is there an existing issue for this problem?

• I have searched the existing issues

Install method

Invoke's Launcher

Operating system

Linux

GPU vendor

Nvidia (CUDA)

GPU model

NVIDIA GeForce RTX 3060 Lite Hash Rate

GPU VRAM

12GB

Version number

v6.13.5

Browser

No response

System Information

Linux Mint 22.1 x86_64
Cinnamon 6.4.8

What happened

This issue happens when Invoke is not connected to the internet and an generation is attempted with Anima Base V1 and finetune checkpoints. The error prevents image generation for starting.

[2026-07-04 20:29:30,645]::[Qwen3EncoderCheckpointLoader]::WARNING --> Unknown Qwen3 variant: embed_hidden_size=1024, layers=28. Attempting to detect configuration from weights...
[2026-07-04 20:29:30,645]::[Qwen3EncoderCheckpointLoader]::INFO --> Qwen3 config: hidden_size=1024, layers=28, heads=16, kv_heads=8, intermediate=3072
[2026-07-04 20:29:31,174]::[InvokeAI]::ERROR --> Error while invoking session 2e62fec8-dd49-412f-820d-ead7e948eebd, invocation ddecaa32-cce6-4b99-b2c2-bcbf9f965321 (anima_text_encoder): [Errno -2] Name or service not known
[2026-07-04 20:29:31,174]::[InvokeAI]::ERROR --> Traceback (most recent call last):
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_transports/default.py", line 101, in map_httpcore_exceptions
    yield
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_transports/default.py", line 250, in handle_request
    resp = self._pool.handle_request(req)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_sync/connection_pool.py", line 256, in handle_request
    raise exc from None
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_sync/connection_pool.py", line 236, in handle_request
    response = connection.handle_request(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_sync/connection.py", line 101, in handle_request
    raise exc
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_sync/connection.py", line 78, in handle_request
    stream = self._connect(request)
             ^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_sync/connection.py", line 124, in _connect
    stream = self._network_backend.connect_tcp(**kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_backends/sync.py", line 207, in connect_tcp
    with map_exceptions(exc_map):
         ^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 158, in __exit__
    self.gen.throw(value)
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpcore/_exceptions.py", line 14, in map_exceptions
    raise to_exc(exc) from exc
httpcore.ConnectError: [Errno -2] Name or service not known

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/services/session_processor/session_processor_default.py", line 130, in run_node
    output = invocation.invoke_internal(context=context, services=self._services)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/invocations/baseinvocation.py", line 244, in invoke_internal
    output = self.invoke(context)
             ^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 124, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/invocations/anima_text_encoder.py", line 84, in invoke
    qwen3_embeds, t5xxl_ids, t5xxl_weights = self._encode_prompt(context)
                                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/invocations/anima_text_encoder.py", line 122, in _encode_prompt
    tokenizer_info = context.models.load(self.qwen3_encoder.tokenizer)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/services/shared/invocation_context.py", line 397, in load
    return self._services.model_manager.load.load_model(model, submodel_type)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/app/services/model_load/model_load_default.py", line 71, in load_model
    ).load_model(model_config, submodel_type)
      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/backend/model_manager/load/load_default.py", line 89, in load_model
    cache_record = self._load_and_cache(model_config, submodel_type)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/backend/model_manager/load/load_default.py", line 134, in _load_and_cache
    loaded_model = self._load_model(config, submodel_type)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/backend/model_manager/load/model_loaders/z_image.py", line 554, in _load_model
    return self._load_tokenizer_with_offline_fallback()
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/invokeai/backend/model_manager/load/model_loaders/z_image.py", line 569, in _load_tokenizer_with_offline_fallback
    return AutoTokenizer.from_pretrained(self.DEFAULT_TOKENIZER_SOURCE, local_files_only=True)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/models/auto/tokenization_auto.py", line 792, in from_pretrained
    return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 1729, in from_pretrained
    return cls._from_pretrained(
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 1918, in _from_pretrained
    tokenizer = cls(*init_inputs, **init_kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/models/qwen2/tokenization_qwen2.py", line 89, in __init__
    super().__init__(
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/tokenization_utils_tokenizers.py", line 477, in __init__
    self._tokenizer = self._patch_mistral_regex(
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/tokenization_utils_tokenizers.py", line 1297, in _patch_mistral_regex
    is_local or (not is_local and is_base_mistral(pretrained_model_name_or_path))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/transformers/tokenization_utils_tokenizers.py", line 1287, in is_base_mistral
    model = model_info(model_id)
            ^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/huggingface_hub/utils/_validators.py", line 88, in _inner_fn
    return fn(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/huggingface_hub/hf_api.py", line 3313, in model_info
    r = get_session().get(path, headers=headers, timeout=timeout, params=params)
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 1053, in get
    return self.request(
           ^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 825, in request
    return self.send(request, auth=auth, follow_redirects=follow_redirects)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 914, in send
    response = self._send_handling_auth(
               ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 942, in _send_handling_auth
    response = self._send_handling_redirects(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 979, in _send_handling_redirects
    response = self._send_single_request(request)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_client.py", line 1014, in _send_single_request
    response = transport.handle_request(request)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_transports/default.py", line 249, in handle_request
    with map_httpcore_exceptions():
         ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/MYNAME/.local/share/uv/python/cpython-3.12.9-linux-x86_64-gnu/lib/python3.12/contextlib.py", line 158, in __exit__
    self.gen.throw(value)
  File "/home/MYNAME/InvokeAI/.venv/lib/python3.12/site-packages/httpx/_transports/default.py", line 118, in map_httpcore_exceptions
    raise mapped_exc(message) from exc
httpx.ConnectError: [Errno -2] Name or service not known

What you expected to happen

Normal image generation.

How to reproduce the problem

Generate image offline using Anima Base V1 or checkpoint with my version of InvokeUI on Linux Mint.

Additional context

Someone reported a similar issue with the previous version of invoke with a different model, I'm not sure if their issue is fully related to this though hence the new bug report.

Discord username

No response

[Pfannkuchensack]

From where is that qwen3 encoder?

[PFJackalope33]

Its the one downloaded through the starter models.

Here is the source URL that invoke provides: Source: https://huggingface.co/circlestone-labs/Anima/resolve/main/split_files/text_encoders/qwen_3_06b_base.safetensors

Update:

I reinstalled the encoder and did a repair install just to see if that would fix it, unfortunately I still get the same error.

Interestingly I still do get the same error when I am connected to the internet it just doesn't seem to stop or inhibit generation.

[2026-07-05 06:24:18,089]::[Qwen3EncoderCheckpointLoader]::WARNING --> Unknown Qwen3 variant: embed_hidden_size=1024, layers=28. Attempting to detect configuration from weights...
[2026-07-05 06:24:18,089]::[Qwen3EncoderCheckpointLoader]::INFO --> Qwen3 config: hidden_size=1024, layers=28, heads=16, kv_heads=8, intermediate=3072
[2026-07-05 06:24:18,994]::[ModelManagerService]::INFO --> [MODEL CACHE] Loaded model '9f2bfc96-4673-47bc-8a69-1de302e3fd8a:text_encoder' (Qwen3ForCausalLM) onto cuda device in 0.28s. Total model size: 1433.62MB, VRAM: 1433.62MB (100.0%)
[2026-07-05 06:24:18,996]::[ModelManagerService]::INFO --> [MODEL CACHE] Loaded model '9f2bfc96-4673-47bc-8a69-1de302e3fd8a:tokenizer' (Qwen2Tokenizer) onto cuda device in 0.00s. Total model size: 0.00MB, VRAM: 0.00MB (0.0%)
[2026-07-05 06:24:20,638]::[ModelManagerService]::INFO --> [MODEL CACHE] Loaded model 'f1f2858e-26df-4e70-9b99-70e7190aa3d1:transformer' (AnimaTransformer) onto cuda device in 0.74s. Total model size: 3988.40MB, VRAM: 3988.40MB (100.0%)
Denoising (Anima):  83%|████████████████████████████████████████████████████████████████████████████████▊                | 25/30 [00:49<00:09,  1.96s/it]

[Pfannkuchensack]

Invoke needs the configuration values ​​for the file, and since these values ​​aren't present in safetensors, they need to be downloaded. That's why it works with an internet connection. Once the configuration has been downloaded, it should work offline because it's cached.

[PFJackalope33]

The con file must not be saving permanently. I run into the same error whenever invoke is restarted when my internet is down.