Update README.md

README.md

@@ -4,7 +4,7 @@ tags:
 - sentence-transformers
 - feature-extraction
 - sentence-similarity
-
+license: mit
 ---
 
 For more details please refer to our github repo: https://github.com/FlagOpen/FlagEmbedding
@@ -25,6 +25,17 @@ This allows you to obtain token weights (similar to the BM25) without any additi
 Utilizing the re-ranking model (e.g., [bge-reranker](https://github.com/FlagOpen/FlagEmbedding/tree/master/FlagEmbedding/reranker), [cohere-reranker](https://txt.cohere.com/rerank/)) after retrieval can further filter the selected text.
 
 
+## Model Specs
+
+| Model Name |  Dimension | Sequence Length |
+|:----:|:---:|:---:|
+| [BAAI/bge-m3](https://huggingface.co/BAAI/bge-m3) | 1024 | 8192 |
+| [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5) | 1024 | 512 |
+| [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) |  768 | 512 |
+| [BAAI/bge-small-en-v1.5](https://huggingface.co/BAAI/bge-small-en-v1.5) |  384 | 512 |
+
+
+
 ## FAQ
 
 **1. Introduction for different retrieval methods**
@@ -72,13 +83,17 @@ pip install -U FlagEmbedding
 ```python
 from FlagEmbedding import BGEM3FlagModel
 
-model = BGEM3FlagModel('BAAI/bge-m3',  use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation
+model = BGEM3FlagModel('BAAI/bge-m3',  
+                       use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation
 
 sentences_1 = ["What is BGE M3?", "Defination of BM25"]
 sentences_2 = ["BGE M3 is an embedding model supporting dense retrieval, lexical matching and multi-vector interaction.", 
                "BM25 is a bag-of-words retrieval function that ranks a set of documents based on the query terms appearing in each document"]
 
-embeddings_1 = model.encode(sentences_1)['dense_vecs']
+embeddings_1 = model.encode(sentences_1, 
+                            batch_size=12, 
+                            max_length=8192, # If you don't need such a long length, you can set a smaller value to speed up the encoding process.
+                            )['dense_vecs']
 embeddings_2 = model.encode(sentences_2)['dense_vecs']
 similarity = embeddings_1 @ embeddings_2.T
 print(similarity)
@@ -148,13 +163,17 @@ sentences_2 = ["BGE M3 is an embedding model supporting dense retrieval, lexical
                "BM25 is a bag-of-words retrieval function that ranks a set of documents based on the query terms appearing in each document"]
 
 sentence_pairs = [[i,j] for i in sentences_1 for j in sentences_2]
-print(model.compute_score(sentence_pairs))
+
+print(model.compute_score(sentence_pairs, 
+                          max_passage_length=128, # a smaller max length leads to a lower latency
+                          weights_for_different_modes=[0.4, 0.2, 0.4])) # weights_for_different_modes(w) is used to do weighted sum: w[0]*dense_score + w[1]*sparse_score + w[2]*colbert_score
+
 # {
-#     'colbert': [0.7796499729156494, 0.4621465802192688, 0.4523794651031494, 0.7898575067520142],
-#     'sparse': [0.05865478515625, 0.0026397705078125, 0.0, 0.0540771484375],
-#     'dense': [0.6259765625, 0.347412109375, 0.349853515625, 0.67822265625],
-#     'sparse+dense': [0.5266395211219788, 0.2692706882953644, 0.2691181004047394, 0.563307523727417],
-#     'colbert+sparse+dense': [0.6366440653800964, 0.3531297743320465, 0.3487969636917114, 0.6618075370788574]
+#   'colbert': [0.7796499729156494, 0.4621465802192688, 0.4523794651031494, 0.7898575067520142], 
+#   'sparse': [0.195556640625, 0.00879669189453125, 0.0, 0.1802978515625], 
+#   'dense': [0.6259765625, 0.347412109375, 0.349853515625, 0.67822265625], 
+#   'sparse+dense': [0.482503205537796, 0.23454029858112335, 0.2332356721162796, 0.5122477412223816], 
+#   'colbert+sparse+dense': [0.6013619303703308, 0.3255828022956848, 0.32089319825172424, 0.6232916116714478]
 # }
 ```
 
@@ -172,8 +191,10 @@ print(model.compute_score(sentence_pairs))
 ![avatar](./imgs/mkqa.jpg)
 
 - Long Document Retrieval
-
-![avatar](./imgs/long.jpg)
+  - MLDR:   
+  ![avatar](./imgs/long.jpg)
+  - NarritiveQA:  
+  ![avatar](./imgs/nqa.jpg)
 
 
 ## Training
@@ -191,8 +212,8 @@ Refer to our [report](https://github.com/FlagOpen/FlagEmbedding/blob/master/Flag
 ## Models
 
 We release two versions:
-- [BAAI/bge-m3-unsupervised](https://huggingface.co/BAAI/bge-m3-unsupervised): the model after contrastive learning in a large-scale dataset
-- [BAAI/bge-m3](https://huggingface.co/BAAI/bge-m3): the final model fine-tuned from BAAI/bge-m3-unsupervised
+- BAAI/bge-m3-unsupervised: the model after contrastive learning in a large-scale dataset
+- BAAI/bge-m3: the final model fine-tuned from BAAI/bge-m3-unsupervised
 
 ## Acknowledgement
 
@@ -204,7 +225,4 @@ If you find this repository useful, please consider giving a star :star: and cit
 
 ```
 
-```
-
-
-
+```