license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: IKT_classifier_economywide_best
results: []
widget:
- text: >-
One million trees have been planted in the embankments, river/ canal banks
to mitigate carbon emission and 2725.1 ha marsh lands were rehabilitated
and included in fisheries culture to enhance livelihood activities by the
Ministry of Livestock and fisheries. Surface Water Use and Rainwater
Harvesting Several city water supply authorities are implementing projects
to increase surface water use and reducing ground water use. These
projects will reduce energy consumption for pumping groundwater and
contribute to GHG emission reduction.
example_title: NEGATIVE
- text: >-
CA global solution is needed to address a global problem. Along with the
rest of the global community, Singapore will play our part to reduce
emissions in support of the long-term temperature goal of the Paris
Agreement. We have put forth a long-term low- emissions development
strategy (LEDS) that aspires to halve emissions from its peak to 33 MtCO2e
by 2050, with a view to achieving net-zero emissions as soon as viable in
the second half of the century.
example_title: ECONOMY-WIDE
IKT_classifier_economywide_best
This model is a fine-tuned version of sentence-transformers/all-mpnet-base-v2 on the GIZ/policy_qa_v0_1 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1642
- Precision Weighted: 0.9530
- Precision Macro: 0.9524
- Recall Weighted: 0.9528
- Recall Samples: 0.9532
- F1-score: 0.9527
- Accuracy: 0.9528
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 9.375102561418467e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100.0
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision Weighted | Precision Macro | Recall Weighted | Recall Samples | F1-score | Accuracy |
|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 30 | 0.3847 | 0.9356 | 0.9340 | 0.9340 | 0.9354 | 0.9339 | 0.9340 |
| No log | 2.0 | 60 | 0.3545 | 0.8911 | 0.8933 | 0.8868 | 0.8832 | 0.8853 | 0.8868 |
| No log | 3.0 | 90 | 0.1387 | 0.9623 | 0.9621 | 0.9623 | 0.9621 | 0.9621 | 0.9623 |
| No log | 4.0 | 120 | 0.1840 | 0.9541 | 0.9555 | 0.9528 | 0.9511 | 0.9525 | 0.9528 |
| No log | 5.0 | 150 | 0.1642 | 0.9530 | 0.9524 | 0.9528 | 0.9532 | 0.9527 | 0.9528 |
Environmental Impact
Carbon emissions were estimated using the codecarbon. The carbon emission reported are incluidng the hyperparamter search performed on subset of training data.
- Hardware Type: 16GB T4
- Hours used: 1
- Cloud Provider: Google Colab
- Carbon Emitted : 0.03666153971974974
[codecarbon INFO @ 20:45:15] Energy consumed for RAM : 0.005929 kWh. RAM Power : 9.54426097869873 W [codecarbon INFO @ 20:45:15] Energy consumed for all GPUs : 0.042682 kWh. Total GPU Power : 36.884 W [codecarbon INFO @ 20:45:15] Energy consumed for all CPUs : 0.026424 kWh. Total CPU Power : 42.5 W [codecarbon INFO @ 20:45:15] 0.075035 kWh of electricity used since the beginning.
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3