IndicSBERT-Matryoshka

This is a sentence-transformers model: Sentence Tranformers is a commonly used framework to train embedding models, and it recently implemented support for Matryoshka models. Training a Matryoshka embedding model using Sentence Transformers is quite elementary: rather than applying some loss function on only the full-size embeddings, we also apply that same loss function on truncated portions of the embeddings.

For example, if a model has an original embedding dimension of 768, it can now be trained on 768, 512, 256, 128 and 64. Each of these losses will be added together, optionally with some weight. this model is specifically finetuned on 11 major Indian languages.

This model is finetuned using: https://huggingface.co./l3cube-pune/indic-sentence-bert-nli

Usage (Sentence-Transformers)

Using this model becomes easy when you have sentence-transformers installed:

pip install -U sentence-transformers

Then you can use the model like this:

from sentence_transformers import SentenceTransformer
from sentence_transformers.util import cos_sim

matryoshka_dim = 64 # Specify the embedding shape here

sentences =
      [
        "मौसम बहुत अच्छा है!",
        "बाहर बहुत धूप है!",
        "वह गाड़ी चलाकर स्टेडियम गया।",
      ]

model = SentenceTransformer("hrusheekeshsawarkar/indic-sentence-bert-nli-matryoshka",truncate_dim=matryoshka_dim)
embeddings = model.encode(sentences)
print(embeddings)

# Similarity of the first sentence to the other two:
similarities = cos_sim(embeddings[0], embeddings[1:])
print(similarities)

Usage (HuggingFace Transformers)

Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.

from transformers import AutoTokenizer, AutoModel
import torch


#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)


# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('{MODEL_NAME}')
model = AutoModel.from_pretrained('{MODEL_NAME}')

# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

# Compute token embeddings
with torch.no_grad():
    model_output = model(**encoded_input)

# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])

print("Sentence embeddings:")
print(sentence_embeddings)

Evaluation Results

For an automated evaluation of this model, see the Sentence Embeddings Benchmark: https://seb.sbert.net

Training

The model was trained with the parameters:

DataLoader:

torch.utils.data.dataloader.DataLoader of length 3850 with parameters:

{'batch_size': 16, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}

Loss:

sentence_transformers.losses.MatryoshkaLoss.MatryoshkaLoss with parameters:

{'loss': 'CoSENTLoss', 'matryoshka_dims': [768, 512, 256, 128, 64], 'matryoshka_weights': [1, 1, 1, 1, 1], 'n_dims_per_step': -1}

Parameters of the fit()-Method:

{
    "epochs": 4,
    "evaluation_steps": 1000,
    "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
    "max_grad_norm": 1,
    "optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
    "optimizer_params": {
        "lr": 2e-05
    },
    "scheduler": "WarmupLinear",
    "steps_per_epoch": null,
    "warmup_steps": 1540,
    "weight_decay": 0.01
}

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'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, 'include_prompt': True})
)