Hello,
I encountered a new issue when loading any model (even the original ones) from a local file.
Example:
model, tokenizer = FastLanguageModel.from_pretrained(
model_name = "models/input/unsloth-gpt-oss-20b-mxfp4",
dtype = dtype,
max_seq_length = max_seq_length,
load_in_4bit = True,
full_finetuning = False,
) I get this error:
AttributeError: 'GptOssTopKRouter' object has no attribute 'weight'
It looks like something breaks when loading from a local path.
#### Unsloth: `hf_xet==1.1.10` and `ipykernel>6.30.1` breaks progress bars. Disabling for now in XET.
#### Unsloth: To re-enable progress bars, please downgrade to `ipykernel==6.30.1` or wait for a fix to
https://github.com/huggingface/xet-core/issues/526
🦥 Unsloth: Will patch your computer to enable 2x faster free finetuning.
Skipping import of cpp extensions due to incompatible torch version 2.8.0+cu129 for torchao version 0.14.0 Please see GitHub issue #2919 for more info
🦥 Unsloth Zoo will now patch everything to make training faster!
==((====))== Unsloth 2025.10.3: Fast Gpt_Oss patching. Transformers: 4.57.0.dev0.
\\ /| NVIDIA GeForce RTX 5060 Ti. Num GPUs = 1. Max memory: 15.928 GB. Platform: Linux.
O^O/ \_/ \ Torch: 2.8.0+cu129. CUDA: 12.0. CUDA Toolkit: 12.9. Triton: 3.4.0
\ / Bfloat16 = TRUE. FA [Xformers = None. FA2 = False]
"-____-" Free license: http://github.com/unslothai/unsloth
Unsloth: Fast downloading is enabled - ignore downloading bars which are red colored!
INFO:httpx: HTTP Request: HEAD https://huggingface.co/unslothai/other/resolve/main/config.json "HTTP/1.1 307 Temporary Redirect"
INFO:httpx: HTTP Request: HEAD https://huggingface.co/api/resolve-cache/models/unslothai/other/43d9e0f2f19a5d7836895f648dc0e762816acf77/config.json "HTTP/1.1 200 OK"
INFO:httpx: HTTP Request: GET https://huggingface.co/api/resolve-cache/models/unslothai/other/43d9e0f2f19a5d7836895f648dc0e762816acf77/config.json "HTTP/1.1 200 OK"
INFO:httpx: HTTP Request: HEAD https://huggingface.co/unslothai/other/resolve/main/config.json "HTTP/1.1 307 Temporary Redirect"
.....
INFO:httpx: HTTP Request: GET https://huggingface.co/api/resolve-cache/models/unslothai/1/7ec782b7604cd9ea0781c23a4270f031650f5617/config.json "HTTP/1.1 200 OK"
INFO:httpx: HTTP Request: HEAD https://huggingface.co/unslothai/1/resolve/main/config.json "HTTP/1.1 307 Temporary Redirect"
INFO:httpx: HTTP Request: HEAD https://huggingface.co/api/resolve-cache/models/unslothai/1/7ec782b7604cd9ea0781c23a4270f031650f5617/config.json "HTTP/1.1 200 OK"
INFO:httpx: HTTP Request: GET https://huggingface.co/api/resolve-cache/models/unslothai/1/7ec782b7604cd9ea0781c23a4270f031650f5617/config.json "HTTP/1.1 200 OK"
Traceback (most recent call last):
File "/mnt/c/AiUtils/train/train_lora_model.py", line 25, in <module>
model, tokenizer = FastLanguageModel.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
....
File "/mnt/c/AiUtils/train/gpt-oss/lib/python3.11/site-packages/transformers/models/gpt_oss/modeling_gpt_oss.py", line 433, in _init_weights
module.weight.data.normal_(mean=0.0, std=std)
^^^^^^^^^^^^^
File "/mnt/c/AiUtils/train/gpt-oss/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1962, in __getattr__
raise AttributeError(
AttributeError: 'GptOssTopKRouter' object has no attribute 'weight'
Hello,
I encountered a new issue when loading any model (even the original ones) from a local file.
Example:
I get this error:
It looks like something breaks when loading from a local path.