LorMolf/mnrl-toole-overlap-roberta-base

SentenceTransformer based on FacebookAI/roberta-base

This is a sentence-transformers model finetuned from FacebookAI/roberta-base. 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: FacebookAI/roberta-base
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 dimensions
• 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: RobertaModel 
  (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("LorMolf/mnrl-toole-overlap-roberta-base")
# Run inference
sentences = [
    'What are the top news stories on Sky News today?',
    'def NewsTool:\n\t"""\n\tDescription:\n\tStay connected to global events with our up-to-date news around the world.\n\t"""',
    'def RepoTool:\n\t"""\n\tDescription:\n\tDiscover GitHub projects tailored to your needs, explore their structures with insightful summaries, and get quick coding solutions with curated snippets. Elevate your coding journey with RepoTool, your go-to companion for GitHub project exploration and code mastery.\n\t"""',
]
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

Device Aware Information Retrieval

• Dataset: dev
• Evaluated with src.port.retrieval_evaluator.DeviceAwareInformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.0574
cosine_accuracy@3	0.1588
cosine_accuracy@5	0.2771
cosine_accuracy@10	0.5484
cosine_precision@1	0.0574
cosine_precision@3	0.0529
cosine_precision@5	0.0554
cosine_precision@10	0.0548
cosine_recall@1	0.0574
cosine_recall@3	0.1588
cosine_recall@5	0.2771
cosine_recall@10	0.5484
cosine_ndcg@1	0.0574
cosine_ndcg@3	0.1156
cosine_ndcg@5	0.1636
cosine_ndcg@10	0.2502
cosine_mrr@10	0.162
cosine_map@100	0.1927

Training Details

Training Dataset

Unnamed Dataset

• Size: 30,000 training samples
• Columns: sentence_0, sentence_1, and sentence_2
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	sentence_2
  type	string	string	string
  details	min: 8 tokens mean: 25.83 tokens max: 121 tokens	min: 22 tokens mean: 40.05 tokens max: 80 tokens	min: 22 tokens mean: 37.91 tokens max: 80 tokens

• Samples:

  sentence_0	sentence_1	sentence_2
  Can you please convert this ABC notation, which represents musical notation, into both a MIDI file, which is a digital audio file format, and a PostScript file, which is a page description language file format used for printing?	def abc_to_audio: """ Description: Converts ABC music notation to WAV, MIDI, and PostScript files. """	def heygen: """ Description: Meet HeyGen - The best AI video generation platform for your team. """
  I urgently require a detailed and extensive report that thoroughly analyzes every aspect of my website's SEO performance. Additionally, I need comprehensive suggestions and recommendations for enhancing its overall performance and making necessary improvements.	def seoanalysis: """ Description: Use AI to analyze and improve the SEO of a website. Get advice on websites, keywords and competitors. """	def WebRewind: """ Description: Get the picture of a website at a specific date. """
  I am actively seeking to hire a highly skilled freelance engineer who specializes in civil engineering and possesses expertise in all aspects of the field, specifically for a construction project.	def TalentOrg: """ Description: Find and hire freelance engineering talents from around the world. """	def Agones: """ Description: Agones provides soccer (football) results for matches played all over the world in the past 15 years. """

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 2
• per_device_eval_batch_size: 2
• num_train_epochs: 1
• fp16: True
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 2
• per_device_eval_batch_size: 2
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 5e-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: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• 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: True
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• 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: False
• 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}
• tp_size: 0
• 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
• 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: None
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• 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
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss	dev_cosine_ndcg@10
-1	-1	-	0.4447
0.0333	500	0.4387	-
0.0667	1000	0.4728	-
0.1	1500	0.6096	-
0.1333	2000	1.2803	-
0.1667	2500	1.3866	-
0.2	3000	1.3848	0.1957
0.2333	3500	1.3854	-
0.2667	4000	1.3864	-
0.3	4500	1.3859	-
0.3333	5000	1.3855	-
0.3667	5500	1.3856	-
0.4	6000	1.3863	0.1970
0.4333	6500	1.3852	-
0.4667	7000	1.3858	-
0.5	7500	1.3862	-
0.5333	8000	1.3856	-
0.5667	8500	1.3857	-
0.6	9000	1.3855	0.2291
0.6333	9500	1.3858	-
0.6667	10000	1.3856	-
0.7	10500	1.3859	-
0.7333	11000	1.386	-
0.7667	11500	1.386	-
0.8	12000	1.3868	0.2648
0.8333	12500	1.3858	-
0.8667	13000	1.3861	-
0.9	13500	1.3862	-
0.9333	14000	1.3857	-
0.9667	14500	1.3856	-
1.0	15000	1.3859	0.2502

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 4.0.2
• Transformers: 4.51.2
• PyTorch: 2.6.0+cu124
• Accelerate: 1.6.0
• Datasets: 3.5.0
• Tokenizers: 0.21.1

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",
}

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}
}