Update README.md

README.md

@@ -16,8 +16,9 @@ Check out our [GitHub](https://github.com/HazyResearch/m2/tree/main) for instruc
 
 You can load this model using Hugging Face `AutoModel`:
 ```python
-from transformers import AutoModelForMaskedLM
-model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", trust_remote_code=True)
+from transformers import AutoModelForMaskedLM, BertConfig
+config = BertConfig.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1")
+model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", config=config, trust_remote_code=True)
 ```
 
 This model uses the Hugging Face `bert-base-uncased tokenizer`:
@@ -30,11 +31,12 @@ tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
 
 This model generates embeddings for retrieval. The embeddings have a dimensionality of 768:
 ```
-from transformers import AutoTokenizer, AutoModelForMaskedLM
+from transformers import AutoTokenizer, AutoModelForMaskedLM, BertConfig
 
 max_seq_length = 32768
 testing_string = "Every morning, I make a cup of coffee to start my day."
-model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", trust_remote_code=True)
+config = BertConfig.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1")
+model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", config=config, trust_remote_code=True)
 
 tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased", model_max_length=max_seq_length)
 input_ids = tokenizer([testing_string], return_tensors="pt", padding="max_length", return_token_type_ids=False, truncation=True, max_length=max_seq_length)
@@ -50,6 +52,7 @@ This model requires `trust_remote_code=True` to be passed to the `from_pretraine
 ```python
 mlm = AutoModelForMaskedLM.from_pretrained(
    "hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1",
+   config=config,
    trust_remote_code=True,
 )
 ```