Upload model - xpynr46w

README.md

@@ -17,7 +17,7 @@ Do not remove elements like the above surrounded by two curly braces and do not
 This model class uses satellite data, and numerical weather predictions to forecast the near-term (up to 8 hours ahead) PV power output at all Grid Service Points (GSPs) in Great Britain. More information can be found in the model repo [1]. The model repo also includes links to our workshop paper on this model and some experimental notes.
 -->
 
-This model uses NWP and Satellite data to predict regional production of solar power in Netherlands. Predictions are done with 15 minutely resolution from 15 minutes to 36 hours ahead.
+This is a summation model that uses NWP and Satellite data to predict national production of solar power in Netherlands. Predictions are done with 15 minutely resolution from 15 minutes to 36 hours ahead.
 
 - **Developed by:** openclimatefix
 - **Model type:** Fusion model
@@ -48,7 +48,6 @@ Data is prepared with the `ocf_data_sampler/torch_datasets/datasets/pvnet_uk` Da
 
 <!-- Do not remove the lines below -->
 The training logs for this model commit can be found here:
- - [https://wandb.ai/openclimatefix/NL-Solar/runs/7s0axgkr](https://wandb.ai/openclimatefix/NL-Solar/runs/7s0axgkr)
 
 
 <!-- The hardware section is also just nice to have -->
@@ -60,10 +59,10 @@ Trained on a single NVIDIA Tesla T4
 
 This model was trained using the following Open Climate Fix packages:
 
-- [1] https://github.com/openclimatefix/PVNet
+- [1] https://github.com/openclimatefix/pvnet-summation
 - [2] https://github.com/openclimatefix/ocf-data-sampler
 
 <!-- Especially do not change the two lines below -->
 The versions of these packages can be found below:
- - pvnet==5.3.0.post0+git.489723d6.dirty
- - ocf-data-sampler==1.0.6
+ - pvnet_summation==1.2.0.post1+git.f4ac6580
+ - ocf-data-sampler==1.0.1.post1+git.b7c40f80.dirty

datamodule_config.yaml

@@ -1,3 +1,3 @@
 pvnet_model:
   model_id: openclimatefix-models/pvnet_nl
-  revision: de5ad0748be8952a4c0600a1ae6455bf802f5896
+  revision: 8379a1f0efc213e6bafc583ccdf3c9ea7713ddd5

full_experiment_config.yaml

@@ -32,7 +32,7 @@ model:
     use_horizon_encoding: true
     use_solar_position: true
     force_non_crossing: true
-    history_minutes: 2880
+    history_minutes: 60
     forecast_minutes: 2160
     interval_minutes: 15
     num_input_locations: 12
@@ -54,9 +54,7 @@ model:
     threshold: 0.0
 datamodule:
   data_source_paths:
-    site:
-      file_path: /home/alex/NL/NL_regional_generation_kw.nc
-      metadata_file_path: /home/alex/NL/NL_regional_metadata.csv
+    generation: /home/alex/NL/NL_regional.zarr
     satellite:
     - /mnt/storage_u2_30tb_a/ml_training_zarrs/sat/uk_sat_crops/v1/2019_nonhrv.zarr
     - /mnt/storage_u2_30tb_a/ml_training_zarrs/sat/uk_sat_crops/v1/2020_nonhrv.zarr
@@ -65,7 +63,7 @@ datamodule:
     - /mnt/storage_u2_30tb_a/ml_training_zarrs/sat/uk_sat_crops/v3/2023_nonhrv.zarr
     - /mnt/storage_u2_30tb_a/ml_training_zarrs/sat/uk_sat_crops/v3/2024_nonhrv.zarr
     nwp:
-      ecmwf: /mnt/storage_u2_4tb_a/ecmwf_nl_crop_rechunked/*_float32.zarr
+      ecmwf: /mnt/storage_u2_30tb_a/nl_data_rechunked/ecmwf_nl_crop_rechunked/*_float32.zarr
   batch_size: 32
   num_workers: 8
   prefetch_factor: 2
@@ -74,7 +72,7 @@ datamodule:
   max_num_val_samples: null
   pvnet_model:
     model_id: openclimatefix-models/pvnet_nl
-    revision: de5ad0748be8952a4c0600a1ae6455bf802f5896
+    revision: 8379a1f0efc213e6bafc583ccdf3c9ea7713ddd5
   train_period:
   - null
   - '2023-12-31'
@@ -99,7 +97,7 @@ callbacks:
     every_n_epochs: 1
     verbose: false
     filename: epoch={epoch}-step={step}
-    dirpath: /home/alex/NL/callbakcs/${model_name}
+    dirpath: /home/alex/NL/summation_callbacks/${model_name}
     auto_insert_metric_name: false
     save_on_train_epoch_end: false
 logger:
@@ -108,14 +106,14 @@ logger:
     project: NL-Solar
     name: ${model_name}
     save_dir: /home/alex/NL/wandb
-    offline: false
+    offline: true
     id: null
     log_model: false
     prefix: ''
     job_type: train
     group: ''
     tags: []
-sample_save_dir: /mnt/storage_u2_30tb_a/NL_summation_samples
+sample_save_dir: /mnt/storage_u2_30tb_a/NL_summation_samples_no_pv
 work_dir: ${hydra:runtime.cwd}
-model_name: v7_summation_horizon
+model_name: v7_summation_horizon_no_pv
 seed: 2727831

model_config.yaml

@@ -1,4 +1,4 @@
-_target_: pvnet.models.LateFusionModel
+_target_: pvnet_summation.models.horizon_dense_model.HorizonDenseModel
 output_quantiles:
 - 0.02
 - 0.1
@@ -7,56 +7,26 @@ output_quantiles:
 - 0.75
 - 0.9
 - 0.98
-nwp_encoders_dict:
-  ecmwf:
-    _target_: pvnet.models.late_fusion.encoders.encoders3d.DefaultPVNet
-    _partial_: true
-    in_channels: 13
-    out_features: 64
-    fc_features: 32
-    number_of_conv3d_layers: 4
-    conv3d_channels: 32
-    image_size_pixels: 10
-sat_encoder:
-  _target_: pvnet.models.late_fusion.encoders.encoders3d.DefaultPVNet
-  _partial_: true
-  in_channels: 11
-  out_features: 256
-  number_of_conv3d_layers: 6
-  conv3d_channels: 32
-  image_size_pixels: 20
-add_image_embedding_channel: false
 output_network:
   _target_: pvnet.models.late_fusion.linear_networks.networks.ResFCNet
   _partial_: true
-  fc_hidden_features: 128
-  n_res_blocks: 6
+  fc_hidden_features: 32
+  n_res_blocks: 1
   res_block_layers: 2
   dropout_frac: 0.2
-include_sun: true
-include_generation_history: false
-location_id_mapping:
-  1: 1
-  2: 2
-  3: 3
-  4: 4
-  5: 5
-  6: 6
-  7: 7
-  8: 8
-  9: 9
-  10: 10
-  11: 11
-  12: 12
-forecast_minutes: 2160
+predict_difference_from_sum: true
+use_horizon_encoding: true
+use_solar_position: true
+force_non_crossing: true
 history_minutes: 60
+forecast_minutes: 2160
 interval_minutes: 15
-min_sat_delay_minutes: 5
-sat_history_minutes: 60
-pv_history_minutes: 60
-nwp_history_minutes:
-  ecmwf: 120
-nwp_forecast_minutes:
-  ecmwf: 2220
-nwp_interval_minutes:
-  ecmwf: 60
+num_input_locations: 12
+input_quantiles:
+- 0.02
+- 0.1
+- 0.25
+- 0.5
+- 0.75
+- 0.9
+- 0.98

model_weights.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b5540fd5e150b356103f603525426622f01603b42fd4b0b038a293a00424723
-size 15898832
+oid sha256:b04709d3a3036e93bf6e9dc17f4cfacd8e8a0b94510a61dafff16a982e555b5f
+size 23684