Upload sbert-hr-v2 fine-tuned model

86d4444 verified 3 months ago

16 kB

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:25128
  - loss:MultipleNegativesRankingLoss
base_model: sentence-transformers/distiluse-base-multilingual-cased-v1
widget:
  - source_sentence: >-
      Do we have your updated personal information on file? (Answer with Yes or
      No)
    sentences:
      - >-
        Employee: Offensive tweets were sent which caused me to feel anger and
        frustration.
      - >-
        Employee: Yes, I'll need help arranging for my vehicle to be
        transported.
      - >-
        Employee: The training will last for two days. The contact is Sofia
        Alvarez, her email is salvarez@lawfirm.com.
  - source_sentence: What level of access do you require? (e.g., Full, Read-Only, Limited)
    sentences:
      - 'Employee: The dates will be from 2023-06-15 to 2023-06-30.'
      - >-
        Employee: I'd say a 5, I really tried to fully invest myself in any
        collaborative work.
      - 'Employee: No other notes. I plan to return on June 15th, 2023.'
  - source_sentence: >-
      What type of time off are you requesting? (e.g., Vacation, Sick Leave,
      Personal Day)
    sentences:
      - >-
        Employee: I would like to select Plan A, and yes you should have my
        current information.
      - >-
        Employee: An apology from John and some workplace training would help.
        That's all I need to add.
      - >-
        Employee: The job transfer will take care of my employment, so no
        additional assistance is needed.
  - source_sentence: Describe the skill development or learning growth shown by the employee.
    sentences:
      - >-
        Employee: I would like my coverage to begin on 2023-03-01. I am looking
        to enroll in health insurance.
      - >-
        Employee: The date range for this review is January 1st, 2023 to
        December 31st, 2023. Alex has improved their digital art skills over the
        past year.
      - 'Employee: The incident was moderate and only affected me.'
  - source_sentence: Where did the incident occur? (Please provide the specific location)
    sentences:
      - 'Employee: I received first aid treatment in the office.'
      - 'Employee: My supervisor Principal Jones approved the request.'
      - 'Employee: Yes, I sprained my ankle. The incident occurred at 10:30.'
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - pearson_cosine
  - spearman_cosine
model-index:
  - name: >-
      SentenceTransformer based on
      sentence-transformers/distiluse-base-multilingual-cased-v1
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: hr eval
          type: hr_eval
        metrics:
          - type: pearson_cosine
            value: 0.40439726027502476
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.38135944113883574
            name: Spearman Cosine

SentenceTransformer based on sentence-transformers/distiluse-base-multilingual-cased-v1

This is a sentence-transformers model finetuned from sentence-transformers/distiluse-base-multilingual-cased-v1. It maps sentences & paragraphs to a 512-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: sentence-transformers/distiluse-base-multilingual-cased-v1
Maximum Sequence Length: 128 tokens
Output Dimensionality: 512 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': 128, 'do_lower_case': False, 'architecture': 'DistilBertModel'})
  (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})
  (2): Dense({'in_features': 768, 'out_features': 512, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
)

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 = [
    'Where did the incident occur? (Please provide the specific location)',
    'Employee: I received first aid treatment in the office.',
    'Employee: Yes, I sprained my ankle. The incident occurred at 10:30.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 512]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.5309, 0.4969],
#         [0.5309, 1.0000, 0.9365],
#         [0.4969, 0.9365, 1.0000]])

Evaluation

Metrics

Semantic Similarity

Dataset: hr_eval
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.4044
spearman_cosine	0.3814

Training Details

Training Dataset

Unnamed Dataset

Size: 25,128 training samples
Columns: sentence_0, sentence_1, and label
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1 label
type string string float
details
min: 10 tokens
mean: 19.45 tokens
max: 32 tokens

min: 8 tokens
mean: 24.39 tokens
max: 56 tokens

min: 0.0
mean: 0.18
max: 1.0

	sentence_0	sentence_1	label
type	string	string	float
details	min: 10 tokens mean: 19.45 tokens max: 32 tokens	min: 8 tokens mean: 24.39 tokens max: 56 tokens	min: 0.0 mean: 0.18 max: 1.0

Samples:

sentence_0	sentence_1	label
`What format do you prefer for the training? (e.g., Online, In-person, Workshop, Seminar)`	`Employee: It's just a planned vacation. There's nothing else to note.`	`0.0`
`Who was involved in the incident? (Names or descriptions of individuals)`	`Employee: It was verbal harassment that happened at work. The person involved was John Smith.`	`1.0`
`What immediate actions were taken following the incident?`	`Employee: My vacation will end on June 22nd, 2023. I have not taken any other time off lately.`	`0.0`

Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 16
per_device_eval_batch_size: 16
num_train_epochs: 1
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: 16
per_device_eval_batch_size: 16
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
num_train_epochs: 1
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: None
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
bf16: False
fp16: False
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}
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}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
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
hub_revision: None
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: no
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
liger_kernel_config: None
eval_use_gather_object: False
average_tokens_across_devices: True
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	hr_eval_spearman_cosine
0.0637	100	-	0.2010
0.1273	200	-	0.1586
0.1910	300	-	0.2509
0.2546	400	-	0.2886
0.3183	500	2.7402	0.3120
0.3819	600	-	0.2132
0.4456	700	-	0.2521
0.5092	800	-	0.2559
0.5729	900	-	0.3063
0.6365	1000	2.668	0.3009
0.7002	1100	-	0.3297
0.7638	1200	-	0.3562
0.8275	1300	-	0.3723
0.8912	1400	-	0.3913
0.9548	1500	2.6426	0.3831
1.0	1571	-	0.3814

Framework Versions

Python: 3.12.8
Sentence Transformers: 5.2.0
Transformers: 4.57.6
PyTorch: 2.5.0
Accelerate: 1.12.0
Datasets: 4.5.0
Tokenizers: 0.22.2

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}
}