zhang19991111's picture
Upload 9 files
92f556b verified
metadata
language: en
license: cc-by-sa-4.0
library_name: span-marker
tags:
  - span-marker
  - token-classification
  - ner
  - named-entity-recognition
  - generated_from_span_marker_trainer
metrics:
  - precision
  - recall
  - f1
widget:
  - text: >-
      Altitude measurements based on near - IR imaging in H and Hcont filters
      showed that the deeper BS2 clouds were located near the methane
      condensation level ( ≈1.2bars ) , while BS1 was generally ∼500 mb above
      that level ( at lower pressures ) .
  - text: >-
      However , our model predicts different performance for large enough memory
      - access latency and validates the intuition that the dynamic programming
      algorithm performs better on these machines .
  - text: >-
      We established a P fertilizer need map based on integrating results from
      the two systems .
  - text: >-
      Here , we have addressed this limitation for the endodermal lineage by
      developing a defined culture system to expand and differentiate human
      foregut stem cells ( hFSCs ) derived from hPSCs . hFSCs can self - renew
      while maintaining their capacity to differentiate into pancreatic and
      hepatic cells .
  - text: >-
      The accumulated percentage gain from selection amounted to 51%/1 % lower
      Striga infestation ( measured by area under Striga number progress curve ,
      ASNPC ) , 46%/62 % lower downy mildew incidence , and 49%/31 % higher
      panicle yield of the C5 - FS compared to the mean of the genepool parents
      at Sadoré / Cinzana , respectively .
pipeline_tag: token-classification
base_model: allenai/specter
model-index:
  - name: SpanMarker with allenai/specter on my-data
    results:
      - task:
          type: token-classification
          name: Named Entity Recognition
        dataset:
          name: my-data
          type: unknown
          split: test
        metrics:
          - type: f1
            value: 0.6710634789777411
            name: F1
          - type: precision
            value: 0.6806020066889632
            name: Precision
          - type: recall
            value: 0.6617886178861788
            name: Recall

SpanMarker with allenai/specter on my-data

This is a SpanMarker model that can be used for Named Entity Recognition. This SpanMarker model uses allenai/specter as the underlying encoder.

Model Details

Model Description

  • Model Type: SpanMarker
  • Encoder: allenai/specter
  • Maximum Sequence Length: 256 tokens
  • Maximum Entity Length: 8 words
  • Language: en
  • License: cc-by-sa-4.0

Model Sources

Model Labels

Label Examples
Data "an overall mitochondrial", "Depth time - series", "defect"
Material "the subject 's fibroblasts", "COXI , COXII and COXIII subunits", "cross - shore measurement locations"
Method "an approximation", "EFSA", "in vitro"
Process "intake", "a significant reduction of synthesis", "translation"

Evaluation

Metrics

Label Precision Recall F1
all 0.6806 0.6618 0.6711
Data 0.5939 0.6190 0.6062
Material 0.765 0.7612 0.7631
Method 0.4667 0.35 0.4
Process 0.6989 0.6341 0.6650

Uses

Direct Use for Inference

from span_marker import SpanMarkerModel

# Download from the 🤗 Hub
model = SpanMarkerModel.from_pretrained("span-marker-allenai/specter-me")
# Run inference
entities = model.predict("We established a P fertilizer need map based on integrating results from the two systems .")

Downstream Use

You can finetune this model on your own dataset.

Click to expand
from span_marker import SpanMarkerModel, Trainer

# Download from the 🤗 Hub
model = SpanMarkerModel.from_pretrained("span-marker-allenai/specter-me")

# Specify a Dataset with "tokens" and "ner_tag" columns
dataset = load_dataset("conll2003") # For example CoNLL2003

# Initialize a Trainer using the pretrained model & dataset
trainer = Trainer(
    model=model,
    train_dataset=dataset["train"],
    eval_dataset=dataset["validation"],
)
trainer.train()
trainer.save_model("span-marker-allenai/specter-me-finetuned")

Training Details

Training Set Metrics

Training set Min Median Max
Sentence length 3 25.6049 106
Entities per sentence 0 5.2439 22

Training Hyperparameters

  • learning_rate: 5e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 10

Framework Versions

  • Python: 3.10.12
  • SpanMarker: 1.5.0
  • Transformers: 4.36.2
  • PyTorch: 2.0.1+cu118
  • Datasets: 2.16.1
  • Tokenizers: 0.15.0

Citation

BibTeX

@software{Aarsen_SpanMarker,
    author = {Aarsen, Tom},
    license = {Apache-2.0},
    title = {{SpanMarker for Named Entity Recognition}},
    url = {https://github.com/tomaarsen/SpanMarkerNER}
}