Skip to content

Latest commit

 

History

History
37 lines (34 loc) · 7.86 KB

pre-trained_models.md

File metadata and controls

37 lines (34 loc) · 7.86 KB

Pre-trained Checkpoint Files

The table below documents all the deepGEFF models that have been trained to predict the target output variable, FWI-Reanalysis. Checkpoint files for these pre-trained models are hosted on a Google Bucket and can be downloaded by running the command:

gsutil cp gs://deepgeff-models-v0/pre-trained_models/{Input Days}_{Output Days}/{Run ID}/{Checkpoint file name} .
Run ID Checkpoint file name Output Days Input Variables Months Input Days Preprocessing Epochs Accuracy
1ltufud2 1ltufud2_94_100.ckpt 10 T2, TP, RH, WSpeed 12 4 - Box-cox transformation 100 1ltufud2_acc.png
162eu91w 162eu91w_87_100.ckpt 10 T2, TP, RH, WSpeed 12 4 - Undersampling for FWI < 10
- Box-cox transformation
100 162eu91w_acc.png
32jafd0h 32jafd0h_84_100.ckpt 10 T2, TP, RH, WSpeed 12 4 - Class-Balanced Loss
- Box-cox transformation
100 32jafd0h_acc.png
9x14d9sv 9x14d9sv_94_100.ckpt 10 T2, TP, RH, WSpeed 12 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
100 9x14d9sv_acc.png
g43pv0ng g43pv0ng_28_30.ckpt 10 T2, TP, RH, WSpeed, SKT 4 4 - Box-cox transformation 30 g43pv0ng_acc.png
2arincwa 2arincwa_29_100.ckpt 10 T2, TP, RH, WSpeed, LAI 4 4 - Box-cox transformation 30 2arincwa_acc.png
lclnszrv lclnszrv_29_100.ckpt 10 T2, TP, RH, WSpeed, SWVL1 4 4 - Box-cox transformation 30 lclnszrv_acc.png
120ecc0k 120ecc0k_29_30.ckpt 10 T2, TP, RH, WSpeed 4 4 - Box-cox transformation 30 120ecc0k_acc.png
3c5dj0kl 3c5dj0kl_27_30.ckpt 1 T2, TP, RH, WSpeed 4 2 - Box-cox transformation 30 3c5dj0kl_acc.png
2ja63rwe 2ja63rwe_29_30.ckpt 1 T2, TP, RH, WSpeed, SKT 4 2 - Box-cox transformation 30 2ja63rwe_acc.png
3npb5bmq 3npb5bmq_26_30.ckpt 1 T2, TP, RH, WSpeed, SWVL1 4 2 - Box-cox transformation 30 3npb5bmq_acc.png
ij6k1cji ij6k1cji_29_30.ckpt 1 T2, TP, RH, WSpeed, LAI 4 2 - Box-cox transformation 30 ij6k1cji_acc.png
3d3qbavv 3d3qbavv_29_30.ckpt 14 T2, TP, RH, WSpeed 3 4 - Box-cox transformation 30 3d3qbavv_acc.png
2y0w8svj 2y0w8svj_28_30.ckpt 14 T2, TP, RH, WSpeed, SWVL1 3 4 - Box-cox transformation 30 2y0w8svj_acc.png
14c0co74 14c0co74_29_30.ckpt 14 T2, TP, RH, WSpeed, SKT 3 4 - Box-cox transformation 30 14c0co74_acc.png
voh3uyh2 voh3uyh2_29_30.ckpt 14 T2, TP, RH, WSpeed, LAI 3 4 - Box-cox transformation 30 voh3uyh2_acc.png
28ow10nu 28ow10nu_28_30.ckpt 14 T2, TP, RH, WSpeed, SWVL1 3 4 - Undersampling for FWI < 10
- Box-cox transformation
30 28ow10nu_acc.png
1gt9ov43 1gt9ov43_28_30.ckpt 14 T2, TP, RH, WSpeed, SKT 3 4 - Undersampling for FWI < 10
- Box-cox transformation
30 1gt9ov43_acc.png
9jqbn6jt 9jqbn6jt_27_30.ckpt 14 T2, TP, RH, WSpeed, LAI 3 4 - Undersampling for FWI < 10
- Box-cox transformation
30 9jqbn6jt_acc.png
f1nbkzcf f1nbkzcf_28_30.ckpt 14 T2, TP, RH, WSpeed, SWVL1 3 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
30 f1nbkzcf_acc.png
23xhuo2a 23xhuo2a_28_30.ckpt 14 T2, TP, RH, WSpeed, SKT 3 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
30 23xhuo2a_acc.png
7rwrvsic 7rwrvsic_29_30.ckpt 14 T2, TP, RH, WSpeed, LAI 3 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
30 7rwrvsic_acc.png
r44sa1l2 r44sa1l2_27_30.ckpt 14 T2, TP, RH, WSpeed, SKT 4 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
30 r44sa1l2_acc.png
3txa61f4 3txa61f4_58_60.ckpt 14 T2, TP, RH, WSpeed, SKT 3 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
60 3txa61f4_acc.png
3b4w0acj 3b4w0acj_28_30.ckpt 14 T2, TP, RH, WSpeed, SKT 5 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
30 3b4w0acj_acc.png
1ga5loax 1ga5loax_59_60.ckpt 14 T2, TP, RH, WSpeed 3 4 - Class-Balanced Loss
- Undersampling for FWI < 10
- Box-cox transformation
60 1ga5loax_acc.png
2tcenlrp 2tcenlrp_29_30.ckpt 21 T2, TP, RH, WSpeed 6 7 - Box-cox transformation 30 2tcenlrp_acc.png