Tochka-AI commited on
Commit
83302a6
1 Parent(s): b9941f5

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -19,7 +19,7 @@ The model is trained on contexts **up to 2048 tokens** in length, but can be use
19
 
20
  ## Usage
21
 
22
- **Important**: To load the model correctly, you must enable dowloading code from the model's repository: `trust_remote_code=True`, this will download the **modeling_rope_bert.py** script and load the weights into the correct architecture.
23
  Otherwise, you can download this script manually and use classes from it directly to load the model.
24
 
25
  ### Basic usage (no efficient attention)
@@ -27,7 +27,7 @@ Otherwise, you can download this script manually and use classes from it directl
27
  ```python
28
  model_name = 'Tochka-AI/ruRoPEBert-classic-base-2k'
29
  tokenizer = AutoTokenizer.from_pretrained(model_name)
30
- model = AutoModel.from_pretrained(model_name, trust_remote_code=True)
31
  ```
32
 
33
  ### With SDPA (efficient attention)
 
19
 
20
  ## Usage
21
 
22
+ **Important**: 4.37.2 and higher is the recommended version of `transformers`. To load the model correctly, you must enable dowloading code from the model's repository: `trust_remote_code=True`, this will download the **modeling_rope_bert.py** script and load the weights into the correct architecture.
23
  Otherwise, you can download this script manually and use classes from it directly to load the model.
24
 
25
  ### Basic usage (no efficient attention)
 
27
  ```python
28
  model_name = 'Tochka-AI/ruRoPEBert-classic-base-2k'
29
  tokenizer = AutoTokenizer.from_pretrained(model_name)
30
+ model = AutoModel.from_pretrained(model_name, trust_remote_code=True, attn_implementation='eager')
31
  ```
32
 
33
  ### With SDPA (efficient attention)