SentenceTransformer based on dunzhang/stella_en_400M_v5
This is a sentence-transformers model finetuned from dunzhang/stella_en_400M_v5. It maps sentences & paragraphs to a 1024-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: dunzhang/stella_en_400M_v5
- Maximum Sequence Length: 512 tokens
- Output Dimensionality: 1024 tokens
- Similarity Function: Cosine Similarity
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': 512, 'do_lower_case': False}) with Transformer model: NewModel
(1): Pooling({'word_embedding_dimension': 1024, '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})
(2): Dense({'in_features': 1024, 'out_features': 1024, 'bias': True, 'activation_function': 'torch.nn.modules.linear.Identity'})
)
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("sentence_transformers_model_id")
# Run inference
sentences = [
'GranularKnowledge: Select the correct answer to an iteration question to the required degree of accuracy\nQuestion: \n\nUsing a starting value of , what is the solution to to decimal places?\nAnswer: ',
'Stops iteration too early when solving equations',
'Does not understand the term quadratic',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
Evaluation
Metrics
Information Retrieval
- Dataset:
val
- Evaluated with
InformationRetrievalEvaluator
Metric | Value |
---|---|
cosine_accuracy@25 | 0.7095 |
cosine_precision@25 | 0.0285 |
cosine_recall@25 | 0.7095 |
cosine_ndcg@25 | 0.3585 |
cosine_mrr@25 | 0.2591 |
cosine_map@25 | 0.2591 |
dot_accuracy@25 | 0.6969 |
dot_precision@25 | 0.028 |
dot_recall@25 | 0.6969 |
dot_ndcg@25 | 0.3429 |
dot_mrr@25 | 0.2438 |
dot_map@25 | 0.2438 |
Training Details
Training Dataset
Unnamed Dataset
- Size: 3,496 training samples
- Columns:
anchor
andpositive
- Approximate statistics based on the first 1000 samples:
anchor positive type string string details - min: 29 tokens
- mean: 85.42 tokens
- max: 403 tokens
- min: 6 tokens
- mean: 14.81 tokens
- max: 39 tokens
- Samples:
- Loss:
CachedMultipleNegativesSymmetricRankingLoss
with these parameters:{ "scale": 20.0, "similarity_fct": "cos_sim", "mini_batch_size": 32 }
Training Hyperparameters
Non-Default Hyperparameters
eval_strategy
: stepsper_device_train_batch_size
: 1394per_device_eval_batch_size
: 1394learning_rate
: 0.0001num_train_epochs
: 5warmup_ratio
: 0.1save_only_model
: Truebf16
: Trueload_best_model_at_end
: Truebatch_sampler
: no_duplicates
All Hyperparameters
Click to expand
overwrite_output_dir
: Falsedo_predict
: Falseeval_strategy
: stepsprediction_loss_only
: Trueper_device_train_batch_size
: 1394per_device_eval_batch_size
: 1394per_gpu_train_batch_size
: Noneper_gpu_eval_batch_size
: Nonegradient_accumulation_steps
: 1eval_accumulation_steps
: Nonetorch_empty_cache_steps
: Nonelearning_rate
: 0.0001weight_decay
: 0.0adam_beta1
: 0.9adam_beta2
: 0.999adam_epsilon
: 1e-08max_grad_norm
: 1.0num_train_epochs
: 5max_steps
: -1lr_scheduler_type
: linearlr_scheduler_kwargs
: {}warmup_ratio
: 0.1warmup_steps
: 0log_level
: passivelog_level_replica
: warninglog_on_each_node
: Truelogging_nan_inf_filter
: Truesave_safetensors
: Truesave_on_each_node
: Falsesave_only_model
: Truerestore_callback_states_from_checkpoint
: Falseno_cuda
: Falseuse_cpu
: Falseuse_mps_device
: Falseseed
: 42data_seed
: Nonejit_mode_eval
: Falseuse_ipex
: Falsebf16
: Truefp16
: Falsefp16_opt_level
: O1half_precision_backend
: autobf16_full_eval
: Falsefp16_full_eval
: Falsetf32
: Nonelocal_rank
: 0ddp_backend
: Nonetpu_num_cores
: Nonetpu_metrics_debug
: Falsedebug
: []dataloader_drop_last
: Falsedataloader_num_workers
: 0dataloader_prefetch_factor
: Nonepast_index
: -1disable_tqdm
: Falseremove_unused_columns
: Truelabel_names
: Noneload_best_model_at_end
: Trueignore_data_skip
: Falsefsdp
: []fsdp_min_num_params
: 0fsdp_config
: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}fsdp_transformer_layer_cls_to_wrap
: Noneaccelerator_config
: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}deepspeed
: Nonelabel_smoothing_factor
: 0.0optim
: adamw_torchoptim_args
: Noneadafactor
: Falsegroup_by_length
: Falselength_column_name
: lengthddp_find_unused_parameters
: Noneddp_bucket_cap_mb
: Noneddp_broadcast_buffers
: Falsedataloader_pin_memory
: Truedataloader_persistent_workers
: Falseskip_memory_metrics
: Trueuse_legacy_prediction_loop
: Falsepush_to_hub
: Falseresume_from_checkpoint
: Nonehub_model_id
: Nonehub_strategy
: every_savehub_private_repo
: Falsehub_always_push
: Falsegradient_checkpointing
: Falsegradient_checkpointing_kwargs
: Noneinclude_inputs_for_metrics
: Falseeval_do_concat_batches
: Truefp16_backend
: autopush_to_hub_model_id
: Nonepush_to_hub_organization
: Nonemp_parameters
:auto_find_batch_size
: Falsefull_determinism
: Falsetorchdynamo
: Noneray_scope
: lastddp_timeout
: 1800torch_compile
: Falsetorch_compile_backend
: Nonetorch_compile_mode
: Nonedispatch_batches
: Nonesplit_batches
: Noneinclude_tokens_per_second
: Falseinclude_num_input_tokens_seen
: Falseneftune_noise_alpha
: Noneoptim_target_modules
: Nonebatch_eval_metrics
: Falseeval_on_start
: Falseuse_liger_kernel
: Falseeval_use_gather_object
: Falsebatch_sampler
: no_duplicatesmulti_dataset_batch_sampler
: proportional
Training Logs
Epoch | Step | Training Loss | val_cosine_map@25 |
---|---|---|---|
0.3333 | 1 | 2.9091 | 0.1645 |
0.6667 | 2 | 2.6755 | 0.2077 |
1.0 | 3 | 2.0208 | 0.2260 |
1.3333 | 4 | 1.6665 | 0.2413 |
1.6667 | 5 | 1.4121 | 0.2556 |
2.0 | 6 | 1.0487 | 0.2505 |
2.3333 | 7 | 1.0154 | 0.2556 |
2.6667 | 8 | 1.1214 | 0.2588 |
3.0 | 9 | 0.9012 | 0.2636 |
3.3333 | 10 | 1.4379 | 0.2624 |
3.6667 | 11 | 1.0769 | 0.2613 |
4.0 | 12 | 1.0226 | 0.2560 |
4.3333 | 13 | 0.9546 | 0.2565 |
4.6667 | 14 | 0.9177 | 0.2587 |
5.0 | 15 | 1.4128 | 0.2591 |
- The bold row denotes the saved checkpoint.
Framework Versions
- Python: 3.11.6
- Sentence Transformers: 3.1.1
- Transformers: 4.45.1
- PyTorch: 2.4.1+cu121
- Accelerate: 0.34.2
- Datasets: 2.18.0
- Tokenizers: 0.20.0
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",
}
- Downloads last month
- 27
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for ebinan92/dunzhang_stella_en_400M_v5
Base model
dunzhang/stella_en_400M_v5Evaluation results
- Cosine Accuracy@25 on valself-reported0.710
- Cosine Precision@25 on valself-reported0.028
- Cosine Recall@25 on valself-reported0.710
- Cosine Ndcg@25 on valself-reported0.359
- Cosine Mrr@25 on valself-reported0.259
- Cosine Map@25 on valself-reported0.259
- Dot Accuracy@25 on valself-reported0.697
- Dot Precision@25 on valself-reported0.028
- Dot Recall@25 on valself-reported0.697
- Dot Ndcg@25 on valself-reported0.343