Add AutoTokenizer & Sentence Transformers support (#1)

- Add bert-base-uncased tokenizer (40ebbcabab30ca13000beac9811ee5ddfc0c531c)
- Add return_dict to NomicBertModel forward (20e35527c9a0212fe919ef1d3b3813b532a79612)
- Add files for Sentence Transformers integration (d43334a0322989a2b52e9ba6a2ee15c72d94ab6d)
- Also add a Normalize module (830f745af02d088a59d01244714f1acc7fa9735a)
- Update README metadata; allows for widget & "Use in Sentence Transformers" (45757a5edbda59f2eec844e0edced22bd4803537)
- Merge branch 'main' into integration/sentence_transformers (790cf31f166f3a5ba05760ea463260f8910f9067)
- Remove accidental .vscode push (d46c50ab01192d2570f274c66076f66e5306ab43)

1_Pooling/config.json

@@ -1,10 +1,9 @@
 {
-    "word_embedding_dimension": 768,
-    "pooling_mode_cls_token": false,
-    "pooling_mode_mean_tokens": true,
-    "pooling_mode_max_tokens": false,
-    "pooling_mode_mean_sqrt_len_tokens": false,
-    "pooling_mode_weightedmean_tokens": false,
-    "pooling_mode_lasttoken": false
-  }
-  
+  "word_embedding_dimension": 768,
+  "pooling_mode_cls_token": false,
+  "pooling_mode_mean_tokens": true,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false,
+  "pooling_mode_weightedmean_tokens": false,
+  "pooling_mode_lasttoken": false
+}

README.md

@@ -1,6 +1,10 @@
 ---
+library_name: sentence-transformers
+pipeline_tag: sentence-similarity
 tags:
-- mteb
+  - feature-extraction
+  - sentence-similarity
+  - mteb
 model-index:
 - name: epoch_0_model
   results:
@@ -2655,6 +2659,17 @@ Training data to train the models is released in its entirety. For more details,
 
 ## Usage
 
+### Sentence Transformers
+```python
+from sentence_transformers import SentenceTransformer
+
+model = SentenceTransformer("../nomic-embed-text-v1", trust_remote_code=True)
+sentences = ['What is TSNE?', 'Who is Laurens van der Maaten?']
+embeddings = model.encode(sentences)
+print(embeddings)
+```
+
+### Transformers
 
 ```python
 import torch
@@ -2689,8 +2704,8 @@ The model natively supports scaling of the sequence length past 2048 tokens. To
 + tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased', model_max_length=8192)
 
 
-- model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1-unsupervised', trust_remote_code=True)
-+ model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1-unsupervised', trust_remote_code=True, rotary_scaling_factor=2)
+- model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1', trust_remote_code=True)
++ model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1', trust_remote_code=True, rotary_scaling_factor=2)
 ```
 
 # Join the Nomic Community

config_sentence_transformers.json

@@ -0,0 +1,7 @@
+{
+  "__version__": {
+    "sentence_transformers": "2.4.0.dev0",
+    "transformers": "4.37.2",
+    "pytorch": "2.1.0+cu121"
+  }
+}

modeling_hf_nomic_bert.py

@@ -1069,6 +1069,7 @@ class NomicBertModel(NomicBertPreTrainedModel):
         position_ids=None,
         token_type_ids=None,
         attention_mask=None,
+        return_dict=None,
     ):
         if token_type_ids is None:
             token_type_ids = torch.zeros_like(input_ids)
@@ -1080,7 +1081,7 @@ class NomicBertModel(NomicBertPreTrainedModel):
 
         attention_mask = self.get_extended_attention_mask(attention_mask, input_ids.shape)
         sequence_output = self.encoder(
-            hidden_states, attention_mask=attention_mask
+            hidden_states, attention_mask=attention_mask, return_dict=return_dict,
         )
 
         pooled_output = self.pooler(sequence_output) if self.pooler is not None else None

modules.json

@@ -0,0 +1,20 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Normalize",
+    "type": "sentence_transformers.models.Normalize"
+  }
+]

sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 8192,
+  "do_lower_case": false
+}

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 8192,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}