got an unexpected keyword argument 'trust_remote_code'

[wangguangtao0722]

when I load SentenceTransformersTextEmbedder on GPU via
"pipline.add_component("text_embedder", SentenceTransformersTextEmbedder(device=ComponentDevice.from_str(self.device)))", the following error occurs:

File "/home/.conda/envs/transformer433/lib/python3.8/site-packages/haystack/core/pipeline/base.py", line 604, in warm_up
self.graph.nodes[node]["instance"].warm_up()
File "/home/.conda/envs/transformer433/lib/python3.8/site-packages/haystack/components/embedders/sentence_transformers_document_embedder.py", line 139, in warm_up
self.embedding_backend = _SentenceTransformersEmbeddingBackendFactory.get_embedding_backend(
File "/home/.conda/envs/transformer433/lib/python3.8/site-packages/haystack/components/embedders/backends/sentence_transformers_backend.py", line 29, in get_embedding_backend
embedding_backend = _SentenceTransformersEmbeddingBackend(
File "/home/.conda/envs/transformer433/lib/python3.8/site-packages/haystack/components/embedders/backends/sentence_transformers_backend.py", line 49, in init
self.model = SentenceTransformer(
TypeError: init() got an unexpected keyword argument 'trust_remote_code'

I can run this script on CPU.

[anakin87]

Can you provide more information about the code you used and the Haystack version?

[wangguangtao0722]

This is the current code. If the device = 'cpu', it works well, but if I set device = 'gpu:0', "trust_remote_code" error raises.

[anakin87]

It would be nice if you could provide a shorter reproducible example and also specify the following details: Haystack version, Sentence Transformers version and used model.

[wangguangtao0722]

haystack 0.42
haystack-ai 2.2.1
fastembed-haystack 1.1.0
sentence-transformers 2.2.2
transformers 4.41.2
Model = sentence-transformers/all-MiniLM-L6-v2

[anakin87]

Ok you should upgrade Sentence Transformers to 2.3.0 at least
UKPLab/sentence-transformers#2272

[wangguangtao0722]

Great! It works, thanks very much~