Hugging Face's logo

Datasets:
Real-TSF
/
TIME-Output

Tasks:
Time Series Forecasting
Modalities:
Tabular
Text
Formats:
csv
Size:
1K - 10K
Tags:
time-series
forecasting
benchmark
Libraries:
Datasets
Dask
Polars
License:
Dataset card Data Studio Files
xet
 Community
54

Submit evaluation results for Timestral-28M

#52
by adrienpetralia - opened
base: refs/heads/main
←
from: refs/pr/52
Discussion Files changed
+4508
-0
This view is limited to 50 files because it contains too many changes. See the raw diff here.
Files changed (50) hide show
  1. results/Timestral-28M/Australia_Solar/H/long/config.json +40 -0
  2. results/Timestral-28M/Australia_Solar/H/long/metrics.npz +3 -0
  3. results/Timestral-28M/Australia_Solar/H/long/predictions.npz +3 -0
  4. results/Timestral-28M/Australia_Solar/H/medium/config.json +40 -0
  5. results/Timestral-28M/Australia_Solar/H/medium/metrics.npz +3 -0
  6. results/Timestral-28M/Australia_Solar/H/medium/predictions.npz +3 -0
  7. results/Timestral-28M/Australia_Solar/H/short/config.json +40 -0
  8. results/Timestral-28M/Australia_Solar/H/short/metrics.npz +3 -0
  9. results/Timestral-28M/Australia_Solar/H/short/predictions.npz +3 -0
  10. results/Timestral-28M/Auto_Production_SF/M/short/config.json +40 -0
  11. results/Timestral-28M/Auto_Production_SF/M/short/metrics.npz +3 -0
  12. results/Timestral-28M/Auto_Production_SF/M/short/predictions.npz +3 -0
  13. results/Timestral-28M/CPHL/15T/long/config.json +40 -0
  14. results/Timestral-28M/CPHL/15T/long/metrics.npz +3 -0
  15. results/Timestral-28M/CPHL/15T/long/predictions.npz +3 -0
  16. results/Timestral-28M/CPHL/15T/medium/config.json +40 -0
  17. results/Timestral-28M/CPHL/15T/medium/metrics.npz +3 -0
  18. results/Timestral-28M/CPHL/15T/medium/predictions.npz +3 -0
  19. results/Timestral-28M/CPHL/15T/short/config.json +40 -0
  20. results/Timestral-28M/CPHL/15T/short/metrics.npz +3 -0
  21. results/Timestral-28M/CPHL/15T/short/predictions.npz +3 -0
  22. results/Timestral-28M/CPHL/30T/long/config.json +40 -0
  23. results/Timestral-28M/CPHL/30T/long/metrics.npz +3 -0
  24. results/Timestral-28M/CPHL/30T/long/predictions.npz +3 -0
  25. results/Timestral-28M/CPHL/30T/medium/config.json +40 -0
  26. results/Timestral-28M/CPHL/30T/medium/metrics.npz +3 -0
  27. results/Timestral-28M/CPHL/30T/medium/predictions.npz +3 -0
  28. results/Timestral-28M/CPHL/30T/short/config.json +40 -0
  29. results/Timestral-28M/CPHL/30T/short/metrics.npz +3 -0
  30. results/Timestral-28M/CPHL/30T/short/predictions.npz +3 -0
  31. results/Timestral-28M/CPHL/H/long/config.json +40 -0
  32. results/Timestral-28M/CPHL/H/long/metrics.npz +3 -0
  33. results/Timestral-28M/CPHL/H/long/predictions.npz +3 -0
  34. results/Timestral-28M/CPHL/H/medium/config.json +40 -0
  35. results/Timestral-28M/CPHL/H/medium/metrics.npz +3 -0
  36. results/Timestral-28M/CPHL/H/medium/predictions.npz +3 -0
  37. results/Timestral-28M/CPHL/H/short/config.json +40 -0
  38. results/Timestral-28M/CPHL/H/short/metrics.npz +3 -0
  39. results/Timestral-28M/CPHL/H/short/predictions.npz +3 -0
  40. results/Timestral-28M/Coastal_T_S/15T/long/config.json +40 -0
  41. results/Timestral-28M/Coastal_T_S/15T/long/metrics.npz +3 -0
  42. results/Timestral-28M/Coastal_T_S/15T/long/predictions.npz +3 -0
  43. results/Timestral-28M/Coastal_T_S/15T/medium/config.json +40 -0
  44. results/Timestral-28M/Coastal_T_S/15T/medium/metrics.npz +3 -0
  45. results/Timestral-28M/Coastal_T_S/15T/medium/predictions.npz +3 -0
  46. results/Timestral-28M/Coastal_T_S/15T/short/config.json +40 -0
  47. results/Timestral-28M/Coastal_T_S/15T/short/metrics.npz +3 -0
  48. results/Timestral-28M/Coastal_T_S/15T/short/predictions.npz +3 -0
  49. results/Timestral-28M/Coastal_T_S/20T/long/config.json +40 -0
  50. results/Timestral-28M/Coastal_T_S/20T/long/metrics.npz +3 -0
results/Timestral-28M/Australia_Solar/H/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Australia_Solar/H/long",
3
+ "num_series": 1,
4
+ "num_windows": 15,
5
+ "num_variates": 3,
6
+ "prediction_length": 168,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 34896,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 100.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Australia_Solar/H/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88fa5c4f59b72ccbb1ff9417e8cf5982261b8902400d56687b05c41469419a
3
+ size 4531
results/Timestral-28M/Australia_Solar/H/long/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d552b280433b1a35c92715214a791d804b5a0eaff69493818c2039442b0d74a
3
+ size 120827
results/Timestral-28M/Australia_Solar/H/medium/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Australia_Solar/H/medium",
3
+ "num_series": 1,
4
+ "num_windows": 35,
5
+ "num_variates": 3,
6
+ "prediction_length": 72,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 34992,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 100.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Australia_Solar/H/medium/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63d159acbb29566136702b8a84481af642cdf787c89fc0a667b96e5e97b6c5cd
3
+ size 8357
results/Timestral-28M/Australia_Solar/H/medium/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a63d7921650d325c5d017c040a1b8d42d0fa89cabe9b4f3b385ce736758b8901
3
+ size 121805
results/Timestral-28M/Australia_Solar/H/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Australia_Solar/H/short",
3
+ "num_series": 1,
4
+ "num_windows": 105,
5
+ "num_variates": 3,
6
+ "prediction_length": 24,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 35040,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 100.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Australia_Solar/H/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb69a2b2e634dd8c4d033d4524a1367dd458b65c88f4f843008e844037582149
3
+ size 20762
results/Timestral-28M/Australia_Solar/H/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8104a136d61c986295f9ff454ebf5be26544a347e7e5e9e8c29e7d77fe295c68
3
+ size 122600
results/Timestral-28M/Auto_Production_SF/M/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Auto_Production_SF/M/short",
3
+ "num_series": 1,
4
+ "num_windows": 5,
5
+ "num_variates": 1,
6
+ "prediction_length": 12,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "M",
20
+ "seasonality": 12,
21
+ "context_length": 355,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Auto_Production_SF/M/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e7312801ff5deedb9421bbcbf4ee619fa0a87835d74c31c7b64b1129e1b204a
3
+ size 1820
results/Timestral-28M/Auto_Production_SF/M/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71787e400a498b5b98b75a30d59b430288a539decc044626bf2ffda9474858a9
3
+ size 1317
results/Timestral-28M/CPHL/15T/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/15T/long",
3
+ "num_series": 2,
4
+ "num_windows": 5,
5
+ "num_variates": 1,
6
+ "prediction_length": 288,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 10543,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/15T/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47d314e514e087d1880c1b302333cf143d44960a1cf12f09d56f81af270f1a10
3
+ size 2173
results/Timestral-28M/CPHL/15T/long/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2eb3b6089b589a6cbbd246f0b50217d44e2d524a5b4820578a75943844541bf
3
+ size 38142
results/Timestral-28M/CPHL/15T/medium/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/15T/medium",
3
+ "num_series": 2,
4
+ "num_windows": 15,
5
+ "num_variates": 1,
6
+ "prediction_length": 96,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 10735,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/15T/medium/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f68104efe4fe66900e28e022d8898280f5d5485ea5beff2c3739121389cfe9a7
3
+ size 3472
results/Timestral-28M/CPHL/15T/medium/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af745f75b3da526d52412ac7a44506a1aa53168310e208f0c4a6c09585fe142e
3
+ size 38945
results/Timestral-28M/CPHL/15T/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/15T/short",
3
+ "num_series": 2,
4
+ "num_windows": 120,
5
+ "num_variates": 1,
6
+ "prediction_length": 12,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 10819,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/15T/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d03d32cc0ac24e38a09ba69a18d2dbaec95bf0dddee50ec7c218ab0d470f43c7
3
+ size 15737
results/Timestral-28M/CPHL/15T/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9691041528ed861da36f6318dd14f48d99d680a92bb99d90e66494d268a62082
3
+ size 40013
results/Timestral-28M/CPHL/30T/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/30T/long",
3
+ "num_series": 2,
4
+ "num_windows": 10,
5
+ "num_variates": 1,
6
+ "prediction_length": 144,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "30T",
20
+ "seasonality": 48,
21
+ "context_length": 17422,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/30T/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:796b42acc2245498f42f7558f89975707d336e96949b7d59b08f9997407e42b2
3
+ size 2831
results/Timestral-28M/CPHL/30T/long/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:868e36c3080c7858f53842057304bc1cc0a6fa1d8cfcad0c0a3192daf8ae23dd
3
+ size 39941
results/Timestral-28M/CPHL/30T/medium/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/30T/medium",
3
+ "num_series": 2,
4
+ "num_windows": 30,
5
+ "num_variates": 1,
6
+ "prediction_length": 48,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "30T",
20
+ "seasonality": 48,
21
+ "context_length": 17518,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/30T/medium/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a30ab08a940d4ba48ffa8c7bd96666173628ba06916edcb471a20dd87d425ed
3
+ size 5466
results/Timestral-28M/CPHL/30T/medium/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f090e0b69dc52c75edab1ca23f43a6133b135205cc0ab8bf7701b0ecef120b48
3
+ size 40376
results/Timestral-28M/CPHL/30T/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/30T/short",
3
+ "num_series": 2,
4
+ "num_windows": 120,
5
+ "num_variates": 1,
6
+ "prediction_length": 12,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "30T",
20
+ "seasonality": 48,
21
+ "context_length": 17554,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/30T/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a85aa2959be76a0d6fe291138eab041192397c10c02e0a41099325f92e9f30c
3
+ size 15880
results/Timestral-28M/CPHL/30T/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:421268fc8dca0bf8073f7dbe5d51f035e7d826c99231d71be97137e10119378e
3
+ size 41381
results/Timestral-28M/CPHL/H/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/H/long",
3
+ "num_series": 4,
4
+ "num_windows": 2,
5
+ "num_variates": 1,
6
+ "prediction_length": 336,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 8447,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/H/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecf9a5b5ef52bcb10fb62af21484c2acb7f82442606e8b8d94f3debb90057401
3
+ size 2035
results/Timestral-28M/CPHL/H/long/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c670ea85b2a44cab17f54dcef5a94437a6fe3e07a2ae951f557fa0b90de03980
3
+ size 37641
results/Timestral-28M/CPHL/H/medium/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/H/medium",
3
+ "num_series": 4,
4
+ "num_windows": 4,
5
+ "num_variates": 1,
6
+ "prediction_length": 168,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 8615,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/H/medium/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c486da9f63792636c630c230e9effd35518bd821e4c96036c5ba336bd64bdee0
3
+ size 2556
results/Timestral-28M/CPHL/H/medium/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b69f9157f982f62a689cd4f98fb233c200886fae710c0a435bf0c447d07fe5e
3
+ size 38277
results/Timestral-28M/CPHL/H/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "CPHL/H/short",
3
+ "num_series": 4,
4
+ "num_windows": 28,
5
+ "num_variates": 1,
6
+ "prediction_length": 24,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "H",
20
+ "seasonality": 24,
21
+ "context_length": 8759,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/CPHL/H/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1c828aba337422d4ce2a0d02704f485e4729fb3222a5f97398a1ed44bc0ef53
3
+ size 8500
results/Timestral-28M/CPHL/H/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5a6a85a6b8ebfd4cc0ac531e9cf83f7aa1586d8465faaf191c3ef959f291385
3
+ size 39198
results/Timestral-28M/Coastal_T_S/15T/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Coastal_T_S/15T/long",
3
+ "num_series": 5,
4
+ "num_windows": 5,
5
+ "num_variates": 3,
6
+ "prediction_length": 288,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 22846,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Coastal_T_S/15T/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff66d2d89e511f7c640cefd3cc46945d34d2528a001000280fa064ccd506fb3d
3
+ size 6279
results/Timestral-28M/Coastal_T_S/15T/long/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cda94fa7997b8536b792ccdec80c14bfd297a89fcb82a3aad9e10135422b5bee
3
+ size 88305
results/Timestral-28M/Coastal_T_S/15T/medium/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Coastal_T_S/15T/medium",
3
+ "num_series": 5,
4
+ "num_windows": 15,
5
+ "num_variates": 3,
6
+ "prediction_length": 96,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 23038,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Coastal_T_S/15T/medium/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9db6263dbc87dd7c4699a131027cb2e5d885faacd123d0754b278bb34db0f1f0
3
+ size 14839
results/Timestral-28M/Coastal_T_S/15T/medium/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42aeddcdfaf1ce44c73ed25dc024141f8837bc639b1a3cb1dc25d0dee3ad829a
3
+ size 96994
results/Timestral-28M/Coastal_T_S/15T/short/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Coastal_T_S/15T/short",
3
+ "num_series": 5,
4
+ "num_windows": 120,
5
+ "num_variates": 3,
6
+ "prediction_length": 12,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "15T",
20
+ "seasonality": 96,
21
+ "context_length": 23122,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Coastal_T_S/15T/short/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c8bd208a3bb0189875a8ceecf4acf8cfe086411cf13978bc805ed3a1987ea05
3
+ size 102964
results/Timestral-28M/Coastal_T_S/15T/short/predictions.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37da86b2f98e1885a29d192ecbc14a469f25a29573d1fc0e56c1950b6262aaea
3
+ size 125042
results/Timestral-28M/Coastal_T_S/20T/long/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset_config": "Coastal_T_S/20T/long",
3
+ "num_series": 1,
4
+ "num_windows": 5,
5
+ "num_variates": 3,
6
+ "prediction_length": 216,
7
+ "num_quantiles": 9,
8
+ "quantile_levels": [
9
+ 0.1,
10
+ 0.2,
11
+ 0.3,
12
+ 0.4,
13
+ 0.5,
14
+ 0.6,
15
+ 0.7,
16
+ 0.8,
17
+ 0.9
18
+ ],
19
+ "freq": "20T",
20
+ "seasonality": 72,
21
+ "context_length": 7982,
22
+ "metric_names": [
23
+ "MSE",
24
+ "MAE",
25
+ "RMSE",
26
+ "MAPE",
27
+ "sMAPE",
28
+ "MASE",
29
+ "ND",
30
+ "CRPS"
31
+ ],
32
+ "prediction_scale_factor": 1.0,
33
+ "model": "Timestral-28M",
34
+ "model_id": "mistralai/Timestral-28M",
35
+ "max_context": 32768,
36
+ "per_shot_horizon_patches": 32,
37
+ "symmetric_averaging": true,
38
+ "clip_nonnegative": true,
39
+ "to_univariate": false
40
+ }
results/Timestral-28M/Coastal_T_S/20T/long/metrics.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9f9a7dc39d9f744e40b0a0e41b107dc4dc109afc40f7af351995c3e25d60725
3
+ size 2489