This is a sentence-transformers model aimed to be used with bert.cpp by Gerganov's GGML Library: It maps sentences & paragraphs to a 384 dimensional dense vector space and was designed for semantic search. It has been trained on 215M (question, answer) pairs from diverse sources. For an introduction to semantic search, have a look at: SBERT.net - Semantic Search
Usage (Start Server)
Using this model becomes easy when you have bert.cpp installed:
./build/bin/server -m models/all-MiniLM-L6-v2/ggml-model-q4_0.bin --port 8085
# bert_model_load: loading model from 'models/all-MiniLM-L6-v2/ggml-model-q4_0.bin' - please wait ...
# bert_model_load: n_vocab = 30522
# bert_model_load: n_ctx = 512
# bert_model_load: n_embd = 384
# bert_model_load: n_intermediate = 1536
# bert_model_load: n_head = 12
# bert_model_load: n_layer = 6
# bert_model_load: f16 = 2
# bert_model_load: ggml ctx size = 13.57 MB
# bert_model_load: ............ done
# bert_model_load: model size = 13.55 MB / num tensors = 101
# Server running on port 8085 with 4 threads
# Waiting for a client
Usage (Start Client)
Then you can use the model like this:
python3 examples/sample_client.py 8085
# Loading texts from sample_client_texts.txt...
# Loaded 1738 lines.
# Starting with a test query "Should I get health insurance?"
# Closest texts:
# 1. Will my Medicare premiums be higher because of my higher income?
# (similarity score: 0.4844)
# 2. Can I sign up for Medicare Part B if I am working and have health insurance through an employer?
# (similarity score: 0.4575)
# 3. Should I sign up for Medicare Part B if I have Veterans' Benefits?
# (similarity score: 0.4052)
# Enter a text to find similar texts (enter 'q' to quit): expensive
# Closest texts:
# 1. It is priced at $ 5,995 for an unlimited number of users tapping into the single processor , or $ 195 per user with a minimum of five users .
# (similarity score: 0.4597)
# 2. The new system costs between $ 1.1 million and $ 22 million , depending on configuration .
# (similarity score: 0.4547)
# 3. Each hull will cost about $ 1.4 billion , with each fully outfitted submarine costing about $ 2.2 billion , Young said .
# (similarity score: 0.4078)
Converting models to ggml format
Converting models is similar to llama.cpp. Use models/convert-to-ggml.py to make hf models into either f32 or f16 ggml models. Then use ./build/bin/quantize to turn those into Q4_0, 4bit per weight models.
There is also models/run_conversions.sh which creates all 4 versions (f32, f16, Q4_0, Q4_1) at once.
sh
1cd models
2# Clone a model from hf
3git clone https://huggingface.co/sentence-transformers/multi-qa-MiniLM-L6-cos-v1
4# Run conversions to 4 ggml formats (f32, f16, Q4_0, Q4_1)
5sh run_conversions.sh multi-qa-MiniLM-L6-cos-v1
Technical Details
In the following some technical details how this model must be used:
Setting
Value
Dimensions
384
Produces normalized embeddings
Yes
Pooling-Method
Mean pooling
Suitable score functions
dot-product (util.dot_score), cosine-similarity (util.cos_sim), or euclidean distance
Note: When loaded with sentence-transformers, this model produces normalized embeddings with length 1. In that case, dot-product and cosine-similarity are equivalent. dot-product is preferred as it is faster. Euclidean distance is proportional to dot-product and can also be used.
Benchmarks
Running MTEB (Massive Text Embedding Benchmark) with bert.cpp vs. sbert(cpu mode) gives comparable results between the two, with quantization having minimal effect on accuracy and eval time being similar or better than sbert with batch_size=1 (bert.cpp doesn't support batching).
bert-base-uncased is not a very good sentence embeddings model, but it's here to show that bert.cpp correctly runs models that are not from SentenceTransformers. Technically any hf model with architecture BertModel or BertForMaskedLM should work.
Data Type
STSBenchmark
eval time
EmotionClassification
eval time
f32
0.4738
52.38
0.3361
88.56
f16
0.4739
33.24
0.3361
55.86
q4_0
0.4940
33.93
0.3375
57.82
q4_1
0.4612
36.86
0.3318
59.63
sbert
0.4729
16.97
0.3527
28.77
sbert-batchless
0.4729
69.97
0.3526
79.02
Background
The project aims to train sentence embedding models on very large sentence level datasets using a self-supervised
contrastive learning objective. We use a contrastive learning objective: given a sentence from the pair, the model should predict which out of a set of randomly sampled other sentences, was actually paired with it in our dataset.
Our model is intented to be used for semantic search: It encodes queries / questions and text paragraphs in a dense vector space. It finds relevant documents for the given passages.
Note that there is a limit of 512 word pieces: Text longer than that will be truncated. Further note that the model was just trained on input text up to 250 word pieces. It might not work well for longer text.
Training procedure
The full training script is accessible in this current repository: train_script.py.
Pre-training
We use the pretrained nreimers/MiniLM-L6-H384-uncased model. Please refer to the model card for more detailed information about the pre-training procedure.
Training
We use the concatenation from multiple datasets to fine-tune our model. In total we have about 215M (question, answer) pairs.
We sampled each dataset given a weighted probability which configuration is detailed in the data_config.json file.
The model was trained with MultipleNegativesRankingLoss using Mean-pooling, cosine-similarity as similarity function, and a scale of 20.
Dataset
Number of training tuples
WikiAnswers Duplicate question pairs from WikiAnswers
77,427,422
PAQ Automatically generated (Question, Paragraph) pairs for each paragraph in Wikipedia
64,371,441
Stack Exchange (Title, Body) pairs from all StackExchanges
25,316,456
Stack Exchange (Title, Answer) pairs from all StackExchanges
21,396,559
MS MARCO Triplets (query, answer, hard_negative) for 500k queries from Bing search engine
