ft_8_4e6_base_x4 / README.md
damgomz's picture
Upload README.md with huggingface_hub
b8629f2 verified
metadata
language: en
tags:
  - text-classification
pipeline_tag: text-classification
widget:
  - text: >-
      GEPS Techno is the pioneer of hybridization of renewable energies at sea.
      We imagine, design and commercialize innovative off-grid systems that aim
      to generate power at sea, stabilize and collect data. The success of our
      low power platforms WAVEPEAL enabled us to scale-up the device up to
      WAVEGEM, the 150-kW capacity platform.

Environmental Impact (CODE CARBON DEFAULT)

Metric Value
Duration (in seconds) 78589.64965057373
Emissions (Co2eq in kg) 0.0475558110146995
CPU power (W) 42.5
GPU power (W) [No GPU]
RAM power (W) 3.75
CPU energy (kWh) 0.927792636016177
GPU energy (kWh) [No GPU]
RAM energy (kWh) 0.081863443784664
Consumed energy (kWh) 1.0096560798008434
Country name Switzerland
Cloud provider nan
Cloud region nan
CPU count 2
CPU model Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz
GPU count nan
GPU model nan

Environmental Impact (for one core)

Metric Value
CPU energy (kWh) 0.15128507557735443
Emissions (Co2eq in kg) 0.030780946113141375

Note

14 juin 2024

My Config

Config Value
checkpoint albert-base-v2
model_name ft_8_4e6_base_x4
sequence_length 400
num_epoch 6
learning_rate 4e-06
batch_size 8
weight_decay 0.0
warm_up_prop 0.0
drop_out_prob 0.1
packing_length 100
train_test_split 0.2
num_steps 29328

Training and Testing steps

Epoch Train Loss Test Loss F-beta Score
0 0.000000 0.703717 0.702334
1 0.316409 0.233850 0.909341
2 0.201761 0.214236 0.903653
3 0.154353 0.219439 0.932031
4 0.109899 0.254901 0.907794
5 0.071217 0.314869 0.918212
6 0.044369 0.352658 0.918399