jonsaadfalcon
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -16,8 +16,9 @@ Check out our [GitHub](https://github.com/HazyResearch/m2/tree/main) for instruc
|
|
16 |
|
17 |
You can load this model using Hugging Face `AutoModel`:
|
18 |
```python
|
19 |
-
from transformers import AutoModelForMaskedLM
|
20 |
-
|
|
|
21 |
```
|
22 |
|
23 |
This model uses the Hugging Face `bert-base-uncased tokenizer`:
|
@@ -30,11 +31,12 @@ tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
|
|
30 |
|
31 |
This model generates embeddings for retrieval. The embeddings have a dimensionality of 768:
|
32 |
```
|
33 |
-
from transformers import AutoTokenizer, AutoModelForMaskedLM
|
34 |
|
35 |
max_seq_length = 32768
|
36 |
testing_string = "Every morning, I make a cup of coffee to start my day."
|
37 |
-
|
|
|
38 |
|
39 |
tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased", model_max_length=max_seq_length)
|
40 |
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
|
|
50 |
```python
|
51 |
mlm = AutoModelForMaskedLM.from_pretrained(
|
52 |
"hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1",
|
|
|
53 |
trust_remote_code=True,
|
54 |
)
|
55 |
```
|
|
|
16 |
|
17 |
You can load this model using Hugging Face `AutoModel`:
|
18 |
```python
|
19 |
+
from transformers import AutoModelForMaskedLM, BertConfig
|
20 |
+
config = BertConfig.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1")
|
21 |
+
model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", config=config, trust_remote_code=True)
|
22 |
```
|
23 |
|
24 |
This model uses the Hugging Face `bert-base-uncased tokenizer`:
|
|
|
31 |
|
32 |
This model generates embeddings for retrieval. The embeddings have a dimensionality of 768:
|
33 |
```
|
34 |
+
from transformers import AutoTokenizer, AutoModelForMaskedLM, BertConfig
|
35 |
|
36 |
max_seq_length = 32768
|
37 |
testing_string = "Every morning, I make a cup of coffee to start my day."
|
38 |
+
config = BertConfig.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1")
|
39 |
+
model = AutoModelForMaskedLM.from_pretrained("hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1", config=config, trust_remote_code=True)
|
40 |
|
41 |
tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased", model_max_length=max_seq_length)
|
42 |
input_ids = tokenizer([testing_string], return_tensors="pt", padding="max_length", return_token_type_ids=False, truncation=True, max_length=max_seq_length)
|
|
|
52 |
```python
|
53 |
mlm = AutoModelForMaskedLM.from_pretrained(
|
54 |
"hazyresearch/M2-BERT-32K-Retrieval-Encoder-V1",
|
55 |
+
config=config,
|
56 |
trust_remote_code=True,
|
57 |
)
|
58 |
```
|