Sentence Similarity
sentence-transformers
PyTorch
Rust
Safetensors
Transformers
English
bert
feature-extraction
Inference Endpoints
text-embeddings-inference
5 papers
nicokuzak commited on
Commit
5c7f88e
1 Parent(s): 6b02c09

Can't spin up an endpoint without this part of the config

Browse files

When trying to spin up a `sentence embedding` endpoint, I get the error `Error: Failed to parse config.json` `missing field architectures at line 21 column 1`. The L6 model has this and doesn't have any problems. I'm emulating the L6 version here

Files changed (1) hide show
  1. config.json +3 -0
config.json CHANGED
@@ -1,5 +1,8 @@
1
  {
2
  "_name_or_path": "microsoft/MiniLM-L12-H384-uncased",
 
 
 
3
  "attention_probs_dropout_prob": 0.1,
4
  "gradient_checkpointing": false,
5
  "hidden_act": "gelu",
 
1
  {
2
  "_name_or_path": "microsoft/MiniLM-L12-H384-uncased",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
  "attention_probs_dropout_prob": 0.1,
7
  "gradient_checkpointing": false,
8
  "hidden_act": "gelu",