metadata

base_model: nomic-ai/nomic-embed-text-v1.5
datasets: []
language:
  - en
library_name: sentence-transformers
license: apache-2.0
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:530
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
widget:
  - source_sentence: >-
      If you receive a BharatPe speaker that you didn't order, please contact
      BharatPe support immediately. They will assist in resolving the issue and
      advise on the next steps.
    sentences:
      - Can I control multiple BharatPe speakers from one app?
      - >-
        What to do if the BharatPe speaker's transaction announcements are
        intermittently silent?
      - What should I do if I receive a BharatPe speaker without ordering it?
  - source_sentence: >-
      Remote control capabilities depend on the model of the BharatPe speaker.
      Check if your model supports remote control through the BharatPe app or a
      connected device.
    sentences:
      - How do I update my personal details in my Bharatpe account?
      - What are the benefits of the BharatPe speaker?
      - Can I control the BharatPe speaker remotely?
  - source_sentence: >-
      If the announcements are not clear, check the speaker's volume settings
      and ensure it's not placed near noisy equipment. If clarity doesn't
      improve, the speaker may need servicing.
    sentences:
      - >-
        What to do if my BharatPe speaker is not syncing with the transaction
        history in the app?
      - What should I do if the speaker is not announcing payments clearly?
      - The speaker doesn't produce any sound, what can be done?
  - source_sentence: >-
      If the speaker is causing interference, try relocating it or other devices
      to reduce the interference. Ensure there's a reasonable distance between
      the speaker and other wireless equipment.
    sentences:
      - Can I use my Bharatpe device for international transactions?
      - How do I know if my BharatPe speaker is under warranty?
      - >-
        What should I do if the BharatPe speaker is causing interference with
        other wireless devices?
  - source_sentence: >-
      I can understand and respond in multiple Indian regional languages. Feel
      free to communicate with me in the language you're most comfortable with.
    sentences:
      - How can I check if the BharatPe speaker is receiving a network signal?
      - Bharti, can you provide tips for effective online communication?
      - Bharti, what languages can you understand and respond to?
model-index:
  - name: Nomic v1.5 Chatbot Matryoshka
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 768
          type: dim_768
        metrics:
          - type: cosine_accuracy@1
            value: 0.9069767441860465
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9767441860465116
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.9767441860465116
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9767441860465116
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.9069767441860465
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.32558139534883723
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1953488372093023
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09767441860465115
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.9069767441860465
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9767441860465116
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.9767441860465116
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9767441860465116
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9509950990863808
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9418604651162791
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.942829457364341
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 512
          type: dim_512
        metrics:
          - type: cosine_accuracy@1
            value: 0.9069767441860465
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9767441860465116
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.9767441860465116
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9767441860465116
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.9069767441860465
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.32558139534883723
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1953488372093023
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09767441860465115
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.9069767441860465
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9767441860465116
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.9767441860465116
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9767441860465116
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9509950990863808
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9418604651162791
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.9426356589147287
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 256
          type: dim_256
        metrics:
          - type: cosine_accuracy@1
            value: 0.8837209302325582
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9534883720930233
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.9767441860465116
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9767441860465116
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.8837209302325582
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3178294573643411
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1953488372093023
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09767441860465115
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.8837209302325582
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9534883720930233
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.9767441860465116
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9767441860465116
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.937755019041576
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9244186046511628
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.9246686671667917
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 128
          type: dim_128
        metrics:
          - type: cosine_accuracy@1
            value: 0.8837209302325582
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9767441860465116
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.9767441860465116
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9767441860465116
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.8837209302325582
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.32558139534883723
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1953488372093023
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09767441860465115
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.8837209302325582
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9767441860465116
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.9767441860465116
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9767441860465116
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9393671921096366
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9263565891472867
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.9263565891472867
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 64
          type: dim_64
        metrics:
          - type: cosine_accuracy@1
            value: 0.9302325581395349
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9767441860465116
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.9767441860465116
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9767441860465116
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.9302325581395349
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.32558139534883723
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1953488372093023
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09767441860465115
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.9302325581395349
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9767441860465116
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.9767441860465116
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9767441860465116
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9595781280730911
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9534883720930233
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.9537827494848395
            name: Cosine Map@100

Nomic v1.5 Chatbot Matryoshka

This is a sentence-transformers model finetuned from nomic-ai/nomic-embed-text-v1.5. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: nomic-ai/nomic-embed-text-v1.5
Maximum Sequence Length: 8192 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Language: en
License: apache-2.0

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 8192, 'do_lower_case': False}) with Transformer model: NomicBertModel 
  (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})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("MANMEET75/nomic-embed-text-v1.5-Chatbot-matryoshka")
# Run inference
sentences = [
    "I can understand and respond in multiple Indian regional languages. Feel free to communicate with me in the language you're most comfortable with.",
    'Bharti, what languages can you understand and respond to?',
    'Bharti, can you provide tips for effective online communication?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

Dataset: dim_768
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.907
cosine_accuracy@3	0.9767
cosine_accuracy@5	0.9767
cosine_accuracy@10	0.9767
cosine_precision@1	0.907
cosine_precision@3	0.3256
cosine_precision@5	0.1953
cosine_precision@10	0.0977
cosine_recall@1	0.907
cosine_recall@3	0.9767
cosine_recall@5	0.9767
cosine_recall@10	0.9767
cosine_ndcg@10	0.951
cosine_mrr@10	0.9419
cosine_map@100	0.9428

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.907
cosine_accuracy@3	0.9767
cosine_accuracy@5	0.9767
cosine_accuracy@10	0.9767
cosine_precision@1	0.907
cosine_precision@3	0.3256
cosine_precision@5	0.1953
cosine_precision@10	0.0977
cosine_recall@1	0.907
cosine_recall@3	0.9767
cosine_recall@5	0.9767
cosine_recall@10	0.9767
cosine_ndcg@10	0.951
cosine_mrr@10	0.9419
cosine_map@100	0.9426

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.8837
cosine_accuracy@3	0.9535
cosine_accuracy@5	0.9767
cosine_accuracy@10	0.9767
cosine_precision@1	0.8837
cosine_precision@3	0.3178
cosine_precision@5	0.1953
cosine_precision@10	0.0977
cosine_recall@1	0.8837
cosine_recall@3	0.9535
cosine_recall@5	0.9767
cosine_recall@10	0.9767
cosine_ndcg@10	0.9378
cosine_mrr@10	0.9244
cosine_map@100	0.9247

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.8837
cosine_accuracy@3	0.9767
cosine_accuracy@5	0.9767
cosine_accuracy@10	0.9767
cosine_precision@1	0.8837
cosine_precision@3	0.3256
cosine_precision@5	0.1953
cosine_precision@10	0.0977
cosine_recall@1	0.8837
cosine_recall@3	0.9767
cosine_recall@5	0.9767
cosine_recall@10	0.9767
cosine_ndcg@10	0.9394
cosine_mrr@10	0.9264
cosine_map@100	0.9264

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.9302
cosine_accuracy@3	0.9767
cosine_accuracy@5	0.9767
cosine_accuracy@10	0.9767
cosine_precision@1	0.9302
cosine_precision@3	0.3256
cosine_precision@5	0.1953
cosine_precision@10	0.0977
cosine_recall@1	0.9302
cosine_recall@3	0.9767
cosine_recall@5	0.9767
cosine_recall@10	0.9767
cosine_ndcg@10	0.9596
cosine_mrr@10	0.9535
cosine_map@100	0.9538

Training Details

Training Dataset

Unnamed Dataset

Size: 530 training samples
Columns: positive and anchor
Approximate statistics based on the first 1000 samples:
positive anchor
type string string
details
min: 11 tokens
mean: 35.33 tokens
max: 99 tokens

min: 7 tokens
mean: 17.3 tokens
max: 29 tokens

	positive	anchor
type	string	string
details	min: 11 tokens mean: 35.33 tokens max: 99 tokens	min: 7 tokens mean: 17.3 tokens max: 29 tokens

Samples:

positive	anchor
`BharatPe Speaker comes with the following benefits: - Helps you avoid payment fraud - Lightweight & Easy installation process - Compatible with SIM & GPRS connectivity - Comes with a battery, no hassle of constant charging - Available in 10 Languages - Cashback Offers - Free replacement To Know more and place an order, tap below http://bharatpe.in/speaker.`	`What are the benefits of the BharatPe speaker?`
`BharatPe Speaker comes with the following benefits: - Helps you avoid payment fraud - Lightweight & Easy installation process - Compatible with SIM & GPRS connectivity - Comes with a battery, no hassle of constant charging - Available in 10 Languages - Cashback Offers - Free replacement To Know more and place an order, tap below http://bharatpe.in/speaker.`	`What advantages does the BharatPe speaker offer?`
`BharatPe Speaker comes with the following benefits: - Helps you avoid payment fraud - Lightweight & Easy installation process - Compatible with SIM & GPRS connectivity - Comes with a battery, no hassle of constant charging - Available in 10 Languages - Cashback Offers - Free replacement To Know more and place an order, tap below http://bharatpe.in/speaker.`	`Can you outline the benefits of using the BharatPe speaker?`

Loss: MatryoshkaLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: epoch
per_device_train_batch_size: 32
per_device_eval_batch_size: 16
gradient_accumulation_steps: 16
learning_rate: 2e-05
num_train_epochs: 10
lr_scheduler_type: cosine
warmup_ratio: 0.1
tf32: False
load_best_model_at_end: True
optim: adamw_torch_fused
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 32
per_device_eval_batch_size: 16
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 16
eval_accumulation_steps: None
learning_rate: 2e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 10
max_steps: -1
lr_scheduler_type: cosine
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: True
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
no_cuda: False
use_cpu: False
use_mps_device: False
seed: 42
data_seed: None
jit_mode_eval: False
use_ipex: False
bf16: False
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: False
local_rank: 0
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
past_index: -1
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: True
ignore_data_skip: False
fsdp: []
fsdp_min_num_params: 0
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
fsdp_transformer_layer_cls_to_wrap: None
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch_fused
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
dataloader_pin_memory: True
dataloader_persistent_workers: False
skip_memory_metrics: True
use_legacy_prediction_loop: False
push_to_hub: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
eval_do_concat_batches: True
fp16_backend: auto
push_to_hub_model_id: None
push_to_hub_organization: None
mp_parameters:
auto_find_batch_size: False
full_determinism: False
torchdynamo: None
ray_scope: last
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
dispatch_batches: None
split_batches: None
include_tokens_per_second: False
include_num_input_tokens_seen: False
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	dim_128_cosine_map@100	dim_256_cosine_map@100	dim_512_cosine_map@100	dim_64_cosine_map@100	dim_768_cosine_map@100
0.9412	1	-	0.7883	0.8148	0.8134	0.7657	0.8234
1.8824	2	-	0.8953	0.8956	0.8859	0.8273	0.8855
2.8235	3	-	0.9167	0.9150	0.9310	0.8926	0.9292
3.7647	4	-	0.9205	0.9208	0.9348	0.9073	0.9349
4.7059	5	-	0.9244	0.9247	0.9348	0.9151	0.9388
5.6471	6	-	0.9244	0.9247	0.9387	0.9189	0.9389
6.5882	7	-	0.9244	0.9247	0.9387	0.9189	0.9389
7.5294	8	-	0.9244	0.9247	0.9388	0.9538	0.9428
8.4706	9	-	0.9264	0.9247	0.9426	0.9538	0.9428
9.4118	10	1.9538	0.9264	0.9247	0.9426	0.9538	0.9428

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.1.2+cu121
Accelerate: 0.32.1
Datasets: 2.19.1
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning}, 
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply}, 
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}