Update TS-Bench live eval results with domain breakdown

app.py

@@ -15,6 +15,8 @@ from src.display.css_html_js import custom_css
 from src.utils import (
     get_grouped_dfs,
     load_baseline_name,
+    prepare_domain_rank_summary,
+    prepare_domain_summary,
     prepare_ranks_table,
     prepare_values_table,
 )
@@ -37,18 +39,22 @@ def refresh_leaderboard():
 
     values_table = prepare_values_table(grouped["overall_values"])
     ranks_table = prepare_ranks_table(grouped["overall_ranks"])
+    domain_values_table = prepare_domain_summary(grouped["domain_values"])
+    domain_ranks_table = prepare_domain_rank_summary(grouped["domain_ranks"])
 
     finished = status.get("finished_at", "n/a")
     data_source = status.get("data_source", "n/a")
+    ts_bench_root = status.get("ts_bench_root", "")
+    root_hint = f" · TS-Bench: `{ts_bench_root}`" if ts_bench_root else ""
     status_line = (
-        f"**Live benchmark** · Data: `{data_source}` (Open-Meteo) · "
+        f"**Live benchmark** · Data: `{data_source}`{root_hint} · "
         f"Baseline: `{baseline}` · Last eval: `{finished}` · "
         f"Auto-refresh every {REFRESH_SECONDS}s"
     )
-    return values_table, ranks_table, status_line
+    return values_table, ranks_table, domain_values_table, domain_ranks_table, status_line
 
 
-values_table, ranks_table, status_md = refresh_leaderboard()
+values_table, ranks_table, domain_values_table, domain_ranks_table, status_md = refresh_leaderboard()
 
 demo = gr.Blocks(css=custom_css)
 with demo:
@@ -59,19 +65,34 @@ with demo:
     with gr.Row():
         refresh_btn = gr.Button("Refresh now", variant="secondary")
 
-    values_df = gr.Dataframe(
-        value=values_table,
-        interactive=False,
-        wrap=True,
-        label="Metric values (normalized vs baseline, lower is better)",
-    )
-
-    ranks_df = gr.Dataframe(
-        value=ranks_table,
-        interactive=False,
-        wrap=True,
-        label="Ranks (average per-dataset rank, lower is better)",
-    )
+    with gr.Tabs():
+        with gr.Tab("Overall"):
+            values_df = gr.Dataframe(
+                value=values_table,
+                interactive=False,
+                wrap=True,
+                label="Metric values (normalized vs baseline, lower is better)",
+            )
+            ranks_df = gr.Dataframe(
+                value=ranks_table,
+                interactive=False,
+                wrap=True,
+                label="Ranks (average per-dataset rank, lower is better)",
+            )
+
+        with gr.Tab("By Domain"):
+            domain_values_df = gr.Dataframe(
+                value=domain_values_table,
+                interactive=False,
+                wrap=True,
+                label="Metric values by domain (geometric mean across datasets in domain)",
+            )
+            domain_ranks_df = gr.Dataframe(
+                value=domain_ranks_table,
+                interactive=False,
+                wrap=True,
+                label="Ranks by domain (average per-dataset rank within domain)",
+            )
 
     with gr.Accordion("About", open=False):
         gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
@@ -84,10 +105,9 @@ with demo:
             show_copy_button=True,
         )
 
-    refresh_btn.click(refresh_leaderboard, outputs=[values_df, ranks_df, status])
-    demo.load(refresh_leaderboard, outputs=[values_df, ranks_df, status])
-    gr.Timer(value=REFRESH_SECONDS).tick(
-        refresh_leaderboard, outputs=[values_df, ranks_df, status]
-    )
+    refresh_outputs = [values_df, ranks_df, domain_values_df, domain_ranks_df, status]
+    refresh_btn.click(refresh_leaderboard, outputs=refresh_outputs)
+    demo.load(refresh_leaderboard, outputs=refresh_outputs)
+    gr.Timer(value=REFRESH_SECONDS).tick(refresh_leaderboard, outputs=refresh_outputs)
 
 demo.queue(default_concurrency_limit=20).launch()

results/chronos_2/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,Chronos-2,23.01338130609796,23.01338130609796,3.3352480199601917,0.6439920893722575,0.16218202867074588,0.1527974363779544,2.89612761434172,4.797226418056371,0.19205978695278986,0.1335286201495986,0.11504645125502568,Energy,1
-nyc/H/short,Chronos-2,64.6470916417079,64.6470916417079,5.342310084236993,0.5863098877808262,0.13170484229384805,0.1154629259043636,3.8361340444568546,8.04034151275354,0.022374767525282956,0.014866650377123296,0.011544855018525518,Econ/Fin,1
-tokyo/H/short,Chronos-2,9.481009614778609,9.481009614778609,1.9769952623690996,0.727519343475743,0.5269572770057526,0.8802218346383637,5.858950496884518,3.079124813121191,0.3399218079461366,0.21825156324346792,0.14708404992692353,Transport,1
+binance_btcusdt_1h_close/1h/short,Chronos-2,3456576.5368118286,3456576.5368118286,1623.0543619791667,2.0574986939524016,0.021986327768627852,0.021672131704816428,48.489959733293944,1859.1870634263323,0.025025205886625466,0.021846790122807665,0.017678493498572475,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,Chronos-2,16.23975015762244,16.23975015762244,3.1123096148173013,1.0075461095970417,0.11108033348814257,0.12178538779084587,22.83609291911104,4.029857337130241,0.15700998007179723,0.12126078660335818,0.10538692936596168,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,Chronos-2,0.004323079047470276,0.004323079047470276,0.06079957137505213,2.393683806447136,0.206001817304993,0.24512269731355915,15.628345381556661,0.06575012583615544,0.17055804634582453,0.1577161409277503,0.1848085402974878,Ocean,1
+nws_ksfo_observations_temperature/1h/short,Chronos-2,0.017383567121896704,0.017383567121896704,0.0405427614847819,0.2432565302441072,0.0031643645061034255,0.0031122500692226085,10.905163462136036,0.1318467562054399,0.010155077300862916,0.0031226773317569183,0.0073943384920597205,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,Chronos-2,37.03956028525954,37.03956028525954,4.665341695149739,0.256435321088093,0.10306621353030117,0.09831916418000608,4.832422174229081,6.0860134969665936,0.12207632588984392,0.09357977491315816,0.19554550178691846,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,Chronos-2,0.32586136196990384,0.32586136196990384,0.4893229007720947,0.2650959403582245,0.019500760656554057,0.019679064961596913,1.6433001426909248,0.570842677074782,0.02274273521672013,0.01949493549565023,0.022315935953188728,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,Chronos-2,15.146823170074489,15.146823170074489,3.005747159322103,0.6218787226183662,0.7787661987632756,0.3357343377482706,4.44647136227838,3.8918919782124592,0.404352413320775,0.31228541915034835,0.26149873994989664,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,Chronos-2,4195.157594045003,4195.157594045003,41.98828125,nan,0.001653054184015716,0.0016498093399364335,nan,64.77003623624896,0.002544158542831383,0.0016492941898527006,0.0016445771390252773,Hydrology,1

results/chronos_2/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "Chronos-2",
     "api_model_id": "amazon/chronos-2",
-    "evaluated_at": "2026-05-28T10:12:27.369342+00:00",
+    "evaluated_at": "2026-05-28T11:28:37.128740+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:12:07.138621+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:12:16.626231+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:12:22.950374+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

results/chronos_bolt_base/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,Chronos-Bolt-Base,18.228861802930954,18.228861802930954,2.9531471265686884,0.5820599244735587,0.14223163408323697,0.13481896730470858,3.4027020979674396,4.269527117015532,0.17093303442938024,0.11823098568668831,0.10213142932311715,Energy,1
-nyc/H/short,Chronos-Bolt-Base,127.86634941859094,127.86634941859094,7.446101744969686,0.7713159302956321,0.2014185006333762,0.16535848331344657,7.195686584286532,11.307800379321831,0.03146749479587533,0.020721109327139026,0.014262716797478302,Econ/Fin,1
-tokyo/H/short,Chronos-Bolt-Base,8.307242700559431,8.307242700559431,1.8618432140081294,0.6809096260138134,0.5041955416902912,0.8758267896424199,5.4925886624835085,2.88222877311282,0.318185352960008,0.20553928464379514,0.13544487699281893,Transport,1
+binance_btcusdt_1h_close/1h/short,Chronos-Bolt-Base,4607018.989560445,4607018.989560445,1858.7041015625,2.35622499837002,0.0251826589192142,0.024764974818790447,45.85497342100289,2146.3967456088926,0.028891132866563078,0.025018705077580722,0.018658445215217687,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,Chronos-Bolt-Base,11.714788128538506,11.714788128538506,2.731881300608317,0.8843902814824524,0.09843378317034886,0.10633484760999325,8.014284235604487,3.4226872671248403,0.13335361891147196,0.10643866337771665,0.08533284538268805,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,Chronos-Bolt-Base,0.009506709623320342,0.009506709623320342,0.07536611706018448,2.9671698316575204,0.2695283864134343,0.21253978219318506,26.998647571987323,0.0975023570141786,0.252924406074776,0.19550225224644832,0.26257767644673047,Ocean,1
+nws_ksfo_observations_temperature/1h/short,Chronos-Bolt-Base,0.0034732855160655163,0.0034732855160655163,0.020564476648966473,0.12338684028198876,0.0016022806741837464,0.0015917658941863241,5.9058213383640705,0.05893458675570328,0.0045392492119111045,0.0015839134464305355,0.00307946348055061,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,Chronos-Bolt-Base,38.52759882238721,38.52759882238721,4.519460995992024,0.24841683790433056,0.09322053873410907,0.0921955378323242,4.593873175357852,6.207060401058396,0.12450434569611492,0.09065362632996957,0.19122388302773416,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,Chronos-Bolt-Base,0.44280348156295685,0.44280348156295685,0.4794563452402751,0.25975062786073605,0.019098214658702924,0.01917220895287895,2.213553795098679,0.6654348063957557,0.026511345793905183,0.01910184564975589,0.024546680423780484,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,Chronos-Bolt-Base,17.587611950468727,17.587611950468727,3.2559762795766196,0.6736502647399903,0.8449049942530357,0.3541087048686474,5.527254602004742,4.193758690061784,0.43571518857784763,0.3382832498261423,0.2748536965433022,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,Chronos-Bolt-Base,4222.804605325063,4222.804605325063,42.295654296875,nan,0.0016651289991267179,0.0016618627932662316,nan,64.98311015429366,0.0025525280584337935,0.001661367762888707,0.0017887527135615566,Hydrology,1

results/chronos_bolt_base/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "Chronos-Bolt-Base",
     "api_model_id": "amazon/chronos-bolt-base",
-    "evaluated_at": "2026-05-28T10:12:05.182036+00:00",
+    "evaluated_at": "2026-05-28T11:28:13.203571+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:10:35.736508+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:11:53.331496+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:12:00.889157+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

results/chronos_bolt_mini/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,Chronos-Bolt-Mini,23.02803795462124,23.02803795462124,3.1608052717314825,0.6035438633560637,0.15491954678594283,0.1448325903393575,3.115631327236734,4.798753791831921,0.19212093626207102,0.1265447019142281,0.1081969714859019,Energy,1
-nyc/H/short,Chronos-Bolt-Mini,67.67541915984299,67.67541915984299,5.3830602169036865,0.6020463455851162,0.13388074766003255,0.11707076108417061,4.7290140050230995,8.226507105682398,0.022892831572236993,0.014980050379299257,0.01192178705488713,Econ/Fin,1
-tokyo/H/short,Chronos-Bolt-Mini,5.129979933399418,5.129979933399418,1.3517586653534737,0.5165812565468629,0.37651829322351843,0.8238071342871242,3.8948858261463672,2.264945900766598,0.250040044563296,0.14922819869975965,0.106032322595721,Transport,1
+binance_btcusdt_1h_close/1h/short,Chronos-Bolt-Mini,3239695.1700490317,3239695.1700490317,1535.5286458333333,1.946544895437788,0.02081274235848315,0.020517944101278912,21.28989139791247,1799.9153230219003,0.02422739078987778,0.020668668184455112,0.013425687663908592,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,Chronos-Bolt-Mini,9.135275099050887,9.135275099050887,2.4557649294535318,0.7950032956155635,0.08925746825784069,0.09554263032551523,7.220700491609023,3.022461761387708,0.1177601640015367,0.09568070787069057,0.07960250786567027,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,Chronos-Bolt-Mini,0.015534220397449413,0.015534220397449413,0.09950688978036244,3.9175939124286194,0.3815415510549928,0.28434516110611735,26.805949948798798,0.12463635263216512,0.3233109068351303,0.2581242317494595,0.2691398090715768,Ocean,1
+nws_ksfo_observations_temperature/1h/short,Chronos-Bolt-Mini,0.003377165089053354,0.003377165089053354,0.018387635548909504,0.11032579575764406,0.0014345831973890761,0.0014243600664891444,5.698881197153537,0.05811338132524517,0.004475998473961995,0.001416249180133919,0.0031020092876858536,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,Chronos-Bolt-Mini,38.374950237810175,38.374950237810175,4.645793914794922,0.25536085716682044,0.09648939445349469,0.09567172423600463,4.506314561404428,6.194751830203545,0.12425745417875896,0.09318767568330809,0.1884475140151591,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,Chronos-Bolt-Mini,0.8058134480226423,0.8058134480226423,0.6428094704945883,0.3482489390605959,0.025220451344546868,0.025546879110079343,2.13034581480904,0.8976711246456813,0.03576378837708663,0.025609938025610044,0.027764242765793087,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,Chronos-Bolt-Mini,18.296158545855253,18.296158545855253,3.3327626387278237,0.6895370976678258,0.8633187411836727,0.35962223692216266,5.221781747094517,4.277400910115307,0.4444052893626293,0.34626105337431934,0.2820276711754296,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,Chronos-Bolt-Mini,4301.3119861284895,4301.3119861284895,43.243245442708336,nan,0.0017023483385521204,0.001699021475602972,nan,65.58438828050842,0.0025761461845044227,0.0016985890190261866,0.0017997282886661212,Hydrology,1

results/chronos_bolt_mini/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "Chronos-Bolt-Mini",
     "api_model_id": "amazon/chronos-bolt-mini",
-    "evaluated_at": "2026-05-28T10:10:33.484071+00:00",
+    "evaluated_at": "2026-05-28T11:27:37.713441+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:10:08.921860+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:10:20.642074+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:10:28.780465+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

results/chronos_bolt_tiny/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,Chronos-Bolt-Tiny,26.466026654661675,26.466026654661675,3.9281906419330173,0.801173140704702,0.2664836929777801,0.2181852629665049,4.3784708273154,5.144514229221421,0.20596365915123496,0.15726742747849448,0.11943187145790482,Energy,1
-nyc/H/short,Chronos-Bolt-Tiny,60.21970314708268,60.21970314708268,4.967222028308445,0.5249371955261508,0.1250899007550242,0.10975090945337941,3.9523726692591126,7.760135510871101,0.021595006598292,0.013822850429123644,0.011242553046380919,Econ/Fin,1
-tokyo/H/short,Chronos-Bolt-Tiny,4.9888988149732105,4.9888988149732105,1.3425839688732393,0.5246587213755459,0.3603745541647132,0.8183498665142726,4.24507431889489,2.233584297709225,0.24657786181385902,0.14821535264635233,0.1081772060090852,Transport,1
+binance_btcusdt_1h_close/1h/short,Chronos-Bolt-Tiny,4129613.9859161377,4129613.9859161377,1748.8697916666667,2.2169912459735355,0.02369875444577939,0.02332389509979645,34.5717005688838,2032.1451685143306,0.02735327296214127,0.023540302891684937,0.01647334941480239,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,Chronos-Bolt-Tiny,7.575305215077758,7.575305215077758,2.2041639486948648,0.7135526622571996,0.07970524238906422,0.08480597509902231,2.145182820424518,2.752327236190813,0.10723527121511543,0.08587791296502906,0.07227905580469843,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,Chronos-Bolt-Tiny,0.00212926727120751,0.00212926727120751,0.037918721636136375,1.4928629904591733,0.10135372044194513,0.09963429626483145,22.4661209949781,0.046143984128025936,0.11969905279118213,0.09836244417701523,0.2008313014153419,Ocean,1
+nws_ksfo_observations_temperature/1h/short,Chronos-Bolt-Tiny,0.003261453246371578,0.003261453246371578,0.019474267959594727,0.11684558918546213,0.0015177768583167321,0.0015079460138350757,5.598889176227384,0.057109134526549934,0.004398649556450488,0.0014999435875331973,0.0030681055610885257,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,Chronos-Bolt-Tiny,45.61863695330218,45.61863695330218,4.8366044362386065,0.26584895440185635,0.09554627026749872,0.09781298110450203,4.594367458116666,6.754157012781253,0.13547828525428338,0.09701504928518727,0.19375042860214714,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,Chronos-Bolt-Tiny,0.46900437186999017,0.46900437186999017,0.5150312582651774,0.27902371932368997,0.020414068294647703,0.020514924162286192,2.0817677095534717,0.6848389386344721,0.02728441876013251,0.020519172804461968,0.02454554141837925,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,Chronos-Bolt-Tiny,17.11004541089532,17.11004541089532,3.2204874753952026,0.666307753530042,0.8362585107064778,0.35154779990197677,5.960524353487738,4.136429065135207,0.4297588639101514,0.3345961013397613,0.27332769139791235,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,Chronos-Bolt-Tiny,4292.419397036235,4292.419397036235,43.07421875,nan,0.0016957134831663966,0.0016923934893262177,nan,65.51655818979073,0.0025734818274222217,0.0016919496726677578,0.0017651717215630117,Hydrology,1

results/chronos_bolt_tiny/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "Chronos-Bolt-Tiny",
     "api_model_id": "amazon/chronos-bolt-tiny",
-    "evaluated_at": "2026-05-28T10:09:59.581604+00:00",
+    "evaluated_at": "2026-05-28T11:27:13.348223+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:09:00.615970+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:09:46.973725+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:09:54.263751+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

results/dataset_properties.csv

@@ -1,4 +1,9 @@
 dataset,domain,frequency,num_variates
-berlin,Energy,H,3
-nyc,Econ/Fin,H,3
-tokyo,Transport,H,3
+binance_btcusdt_1h_close,Finance,1h,1
+nasa_power_shanghai_hourly_t2m,Weather,1h,1
+noaa_coops_sf_water_level_water_level,Ocean,6min,1
+nws_ksfo_observations_temperature,Weather,1h,1
+open_meteo_shanghai_air_quality_hourly_pm2_5,Air Quality,1h,1
+open_meteo_shanghai_hourly_temperature_2m,Weather,1h,1
+usgs_earthquake_all_week_hourly_earthquake_count,Hazards,1h,1
+usgs_potomac_iv_usgs_00060,Hydrology,15min,1

results/moirai_1_1_small/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,Moirai-1.1-Small,25.942115142478016,25.914808326787288,4.134999871253967,0.8981191360243419,0.2807477289705378,0.2263154350433037,4.74372356414242,5.0933402735805915,0.20391487967713806,0.16554715685998675,0.13114703596375427,Energy,1
-nyc/H/short,Moirai-1.1-Small,115.85427334839227,120.1817511764708,7.214917302131653,0.797558713399391,0.19310537226504765,0.15686588488083114,4.257123747918823,10.763562298253877,0.029952981945521127,0.02007776623583399,0.014418479379084874,Econ/Fin,1
-tokyo/H/short,Moirai-1.1-Small,5.042828195691895,4.650267012598303,1.3240045196531962,0.49590462426539694,0.38585838156843166,0.8294245926005037,4.263774493409405,2.2456242329677276,0.2479070175996799,0.1461642633424671,0.11333818582137908,Transport,1
+binance_btcusdt_1h_close/1h/short,Moirai-1.1-Small,3390792.0338694253,3250596.685723623,1596.3206380208333,2.0236091315216753,0.021615965498509516,0.02132047128913711,55.05904119187469,1841.4103382650553,0.024785925926097418,0.021486946318311433,0.018204560989331663,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,Moirai-1.1-Small,17.161671841229992,19.652554899722645,3.3657895723978677,1.0896050229857455,0.11970495933232987,0.13274751685468023,12.267177338715925,4.142664823664834,0.16140514836958095,0.13113678958778802,0.1139907354264086,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,Moirai-1.1-Small,0.032676079013027236,0.03367299428084677,0.15551247944434485,6.122538289786621,0.5620811233437056,0.3930596061461844,26.333062841452506,0.18076525941957775,0.468911186126516,0.4034046222591011,0.31140152062426113,Ocean,1
+nws_ksfo_observations_temperature/1h/short,Moirai-1.1-Small,0.05592957999988357,0.0033324138969893133,0.016686856746673584,0.10012112456621734,0.0013036336316992689,0.0012935427405916388,7.563268532233472,0.23649435511209052,0.018215225967629194,0.0012852520990873227,0.00433598717861285,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,Moirai-1.1-Small,282.3376806496176,228.05316836794057,13.081255078315735,0.7190246857439403,0.2855619235227987,0.3445380131709041,7.350156328629539,16.80290691069904,0.33704117497432257,0.26239040691983084,0.2608272141742082,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,Moirai-1.1-Small,0.7041840767868356,0.6881678424846314,0.7153110901514689,0.3875274707011718,0.029217001935354953,0.028838413590723245,3.0817251996444437,0.8391567653226872,0.03343253909614939,0.028498448651846527,0.02684884140550711,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,Moirai-1.1-Small,18.806281264192233,14.745847608713348,2.9538458585739136,0.6111405224635683,0.7604717681290337,0.3316750803113991,5.700842945740141,4.336620950024597,0.45055802078177626,0.3068930762154715,0.26118736757917893,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,Moirai-1.1-Small,5757.071295579274,4171.65254219373,41.727945963541664,nan,0.0016428269019221863,0.0016396002202848034,nan,75.87536685630768,0.0029803744755341807,0.0016390682538870705,0.0017124873911733953,Hydrology,1

results/moirai_1_1_small/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "Moirai-1.1-Small",
     "api_model_id": "Salesforce/moirai-1.1-R-small",
-    "evaluated_at": "2026-05-28T10:18:27.617821+00:00",
+    "evaluated_at": "2026-05-28T11:30:01.959858+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:18:02.944773+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:18:14.313951+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:18:22.281838+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

results/online_status.json

@@ -1,32 +1,283 @@
 {
     "status": "ok",
-    "started_at": "2026-05-28T10:17:24.399004+00:00",
-    "finished_at": "2026-05-28T10:18:27.618686+00:00",
-    "data_source": "open_meteo",
-    "data_config": "configs/datasets/open_meteo.yaml",
+    "started_at": "2026-05-28T11:26:01.650755+00:00",
+    "finished_at": "2026-05-28T11:30:01.962344+00:00",
+    "data_source": "ts_bench",
+    "data_config": "configs/datasets/ts_bench.yaml",
+    "ts_bench_root": "/Users/zhouziyu/Desktop/TS-Bench",
     "baseline_model": "Chronos-Bolt-Base",
     "models": {
+        "Chronos-Bolt-Tiny": {
+            "model": "Chronos-Bolt-Tiny",
+            "api_model_id": "amazon/chronos-bolt-tiny",
+            "evaluated_at": "2026-05-28T11:27:13.348223+00:00",
+            "datasets": [
+                {
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                }
+            ]
+        },
+        "Chronos-Bolt-Mini": {
+            "model": "Chronos-Bolt-Mini",
+            "api_model_id": "amazon/chronos-bolt-mini",
+            "evaluated_at": "2026-05-28T11:27:37.713441+00:00",
+            "datasets": [
+                {
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                }
+            ]
+        },
+        "Chronos-Bolt-Base": {
+            "model": "Chronos-Bolt-Base",
+            "api_model_id": "amazon/chronos-bolt-base",
+            "evaluated_at": "2026-05-28T11:28:13.203571+00:00",
+            "datasets": [
+                {
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                }
+            ]
+        },
+        "Chronos-2": {
+            "model": "Chronos-2",
+            "api_model_id": "amazon/chronos-2",
+            "evaluated_at": "2026-05-28T11:28:37.128740+00:00",
+            "datasets": [
+                {
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                }
+            ]
+        },
         "TimesFM-2.5": {
             "model": "TimesFM-2.5",
             "api_model_id": "google/timesfm-2.5-200m-pytorch",
-            "evaluated_at": "2026-05-28T10:18:01.120947+00:00",
+            "evaluated_at": "2026-05-28T11:29:36.583998+00:00",
             "datasets": [
                 {
-                    "dataset": "berlin/H",
-                    "data_fetched_at": "2026-05-28T10:17:26.673802+00:00",
-                    "context_length": 504,
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 },
                 {
-                    "dataset": "nyc/H",
-                    "data_fetched_at": "2026-05-28T10:17:42.004619+00:00",
-                    "context_length": 504,
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 },
                 {
-                    "dataset": "tokyo/H",
-                    "data_fetched_at": "2026-05-28T10:17:52.661277+00:00",
-                    "context_length": 504,
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 }
             ]
@@ -34,24 +285,54 @@
         "Moirai-1.1-Small": {
             "model": "Moirai-1.1-Small",
             "api_model_id": "Salesforce/moirai-1.1-R-small",
-            "evaluated_at": "2026-05-28T10:18:27.617821+00:00",
+            "evaluated_at": "2026-05-28T11:30:01.959858+00:00",
             "datasets": [
                 {
-                    "dataset": "berlin/H",
-                    "data_fetched_at": "2026-05-28T10:18:02.944773+00:00",
-                    "context_length": 504,
+                    "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 84,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
+                    "prediction_length": 24
+                },
+                {
+                    "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 },
                 {
-                    "dataset": "nyc/H",
-                    "data_fetched_at": "2026-05-28T10:18:14.313951+00:00",
-                    "context_length": 504,
+                    "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 },
                 {
-                    "dataset": "tokyo/H",
-                    "data_fetched_at": "2026-05-28T10:18:22.281838+00:00",
-                    "context_length": 504,
+                    "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+                    "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+                    "context_length": 96,
                     "prediction_length": 24
                 }
             ]

results/timesfm_2_5/all_results.csv

@@ -1,4 +1,9 @@
 dataset,model,eval_metrics/MSE[mean],eval_metrics/MSE[0.5],eval_metrics/MAE[0.5],eval_metrics/MASE[0.5],eval_metrics/MAPE[0.5],eval_metrics/sMAPE[0.5],eval_metrics/MSIS,eval_metrics/RMSE[mean],eval_metrics/NRMSE[mean],eval_metrics/ND[0.5],eval_metrics/mean_weighted_sum_quantile_loss,domain,num_variates
-berlin/H/short,TimesFM-2.5,26.54830902205753,31.322212017679387,3.443761388460795,0.6303993503146956,0.1386495615012084,0.13621565370038174,18.01169611505954,5.152505121012257,0.20628357921361393,0.13787301681124278,0.1312424613907817,Energy,1
-nyc/H/short,TimesFM-2.5,66.51865488618301,44.39263497040171,4.487805022133721,0.4947887756747412,0.09551055813760116,0.08531003986906531,15.677110160428015,8.155896939404213,0.02269633667797287,0.012488722513809137,0.014150956333775139,Econ/Fin,1
-tokyo/H/short,TimesFM-2.5,4.1984207631854,2.872721300592553,1.081747318436909,0.411514142106336,0.31230229584942987,0.6226950610386188,18.191892843171864,2.0490048226359545,0.2262011013105863,0.11942013609095183,0.13494832121177208,Transport,1
+binance_btcusdt_1h_close/1h/short,TimesFM-2.5,2945013.3387781777,2681467.9179967246,1428.6839192708333,1.8111009506706026,0.019353734596483634,0.019109486135747686,79.32768631088803,1716.1041165320296,0.02309926724638121,0.019230506671433854,0.02050973154774111,Finance,1
+nasa_power_shanghai_hourly_T2M/1h/short,TimesFM-2.5,7.564045018159201,11.792191819793516,2.7090757687886557,0.8770074604569356,0.09732200338765,0.10518331037226293,5.763399002020428,2.7502808980464524,0.10715554245938361,0.10555012172549984,0.07059366255091025,Weather,1
+noaa_coops_sf_water_level_water_level/6min/short,TimesFM-2.5,0.0012317975151250697,0.0006751451120904731,0.021281978736321133,0.8378731415068383,0.07262858551861906,0.06855572589795482,59.23829910967593,0.035096973019408235,0.09104273299429477,0.055206171387193116,0.08958580834201618,Ocean,1
+nws_ksfo_observations_temperature/1h/short,TimesFM-2.5,0.003379698177809587,0.0033795199761546733,0.017791271209716797,0.10674761029122508,0.0013887252558803381,0.0013784933480735714,6.0172820063778545,0.058135171607294554,0.004477676801172975,0.001370316112546397,0.0017051250960053892,Weather,1
+open_meteo_shanghai_air_quality_hourly_pm2_5/1h/short,TimesFM-2.5,115.98230007174395,208.12528911732048,11.97671381632487,0.6583124353490453,0.22580687277819833,0.2659550464465625,5.12798307368112,10.769507884380973,0.21602021665285698,0.2402349616312552,0.17359867126943784,Air Quality,1
+open_meteo_shanghai_hourly_temperature_2m/1h/short,TimesFM-2.5,0.3277569304599031,0.3587956641404162,0.49057690302530926,0.265775309555952,0.019524724242541983,0.01952001086124053,9.464577663910285,0.5725005942878165,0.02280878769265011,0.019544895742757497,0.0218208472167031,Weather,1
+usgs_earthquake_all_week_hourly_earthquake_count/1h/short,TimesFM-2.5,18.241773770916666,14.520057190405927,3.0709699789683023,0.6353730990968901,0.7601302723747606,0.3445192624252451,24.76184288386641,4.2710389568483995,0.4437443072050285,0.31906181599670674,0.3453991472979128,Hazards,1
+usgs_potomac_iv_usgs_00060/15min/short,TimesFM-2.5,4170.709945042928,4152.292089939117,41.726806640625,nan,0.0016427671452028975,0.0016395554622701321,nan,64.58103394219488,0.0025367345574675566,0.0016390235014320785,0.0017658538768412437,Hydrology,1

results/timesfm_2_5/online_meta.json

@@ -1,24 +1,54 @@
 {
     "model": "TimesFM-2.5",
     "api_model_id": "google/timesfm-2.5-200m-pytorch",
-    "evaluated_at": "2026-05-28T10:18:01.120947+00:00",
+    "evaluated_at": "2026-05-28T11:29:36.583998+00:00",
     "datasets": [
         {
-            "dataset": "berlin/H",
-            "data_fetched_at": "2026-05-28T10:17:26.673802+00:00",
-            "context_length": 504,
+            "dataset": "task_binance_btcusdt_1h_crypto_btcusdt_close_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         },
         {
-            "dataset": "nyc/H",
-            "data_fetched_at": "2026-05-28T10:17:42.004619+00:00",
-            "context_length": 504,
+            "dataset": "task_nasa_power_shanghai_hourly_point_shanghai_31.2304_121.4737_t2m_20260524t230000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 84,
             "prediction_length": 24
         },
         {
-            "dataset": "tokyo/H",
-            "data_fetched_at": "2026-05-28T10:17:52.661277+00:00",
-            "context_length": 504,
+            "dataset": "task_noaa_coops_sf_water_level_coops_station_9414290_water_level_20260528t085400z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_nws_ksfo_observations_nws_station_ksfo_temperature_20260528t092500z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_air_quality_hourly_grid_shanghai_31.2304_121.4737_pm2_5_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_open_meteo_shanghai_hourly_grid_shanghai_31.2304_121.4737_temperature_2m_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_earthquake_all_week_hourly_earth_global_earthquake_count_20260527t110000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
+            "prediction_length": 24
+        },
+        {
+            "dataset": "task_usgs_potomac_iv_usgs_site_01646500_usgs_00060_20260528t085000z",
+            "data_fetched_at": "2026-05-28T11:26:01.551610Z",
+            "context_length": 96,
             "prediction_length": 24
         }
     ]

src/utils.py

@@ -65,9 +65,9 @@ def format_number(value):
 def format_df(df: pd.DataFrame) -> pd.DataFrame:
     formatted = df.copy()
     for col in formatted.columns:
-        formatted[col] = formatted[col].map(format_number)
-    if len(formatted.columns) > 1:
-        formatted.iloc[:, 1:] = formatted.iloc[:, 1:].astype(float)
+        if pd.api.types.is_numeric_dtype(formatted[col]):
+            formatted[col] = formatted[col].map(format_number)
+            formatted[col] = formatted[col].astype(float)
     return formatted
 
 
@@ -173,9 +173,26 @@ def get_grouped_dfs(
         columns={f"Rank_{metric}": f"{METRIC_LABELS[metric]}_Rank" for metric in DISPLAY_METRICS}
     )
 
+    domain_values: dict[str, pd.DataFrame] = {}
+    domain_ranks: dict[str, pd.DataFrame] = {}
+    domains = sorted(df["domain"].dropna().unique()) if "domain" in df.columns else []
+    for domain in domains:
+        subset = df[df["domain"] == domain]
+        if subset.empty:
+            continue
+        values = subset.groupby(["model"])[DISPLAY_METRICS].agg(stats.gmean)
+        ranks = subset.groupby(["model"])[rank_cols].mean()
+        domain_values[domain] = values.rename(columns=METRIC_LABELS)
+        domain_ranks[domain] = ranks.rename(
+            columns={f"Rank_{metric}": f"{METRIC_LABELS[metric]}_Rank" for metric in DISPLAY_METRICS}
+        )
+
     return {
         "overall_values": overall_values,
         "overall_ranks": overall_ranks,
+        "domain_values": domain_values,
+        "domain_ranks": domain_ranks,
+        "domains": domains,
     }
 
 
@@ -191,3 +208,29 @@ def prepare_ranks_table(ranks_df: pd.DataFrame) -> pd.DataFrame:
     table = format_df(table)
     cols = ["Model"] + [col for col in RANK_COLUMNS if col in table.columns]
     return table[cols].sort_values(by=["MASE_Rank"])
+
+
+def prepare_domain_summary(values_by_domain: dict[str, pd.DataFrame]) -> pd.DataFrame:
+    rows = []
+    for domain, values_df in sorted(values_by_domain.items()):
+        for model, metrics in values_df.iterrows():
+            row = {"Domain": domain, "Model": model}
+            row.update(metrics.to_dict())
+            rows.append(row)
+    if not rows:
+        return pd.DataFrame(columns=["Domain", "Model"] + VALUE_COLUMNS)
+    table = pd.DataFrame(rows)
+    return format_df(table.sort_values(by=["Domain", "MASE"]))
+
+
+def prepare_domain_rank_summary(ranks_by_domain: dict[str, pd.DataFrame]) -> pd.DataFrame:
+    rows = []
+    for domain, ranks_df in sorted(ranks_by_domain.items()):
+        for model, metrics in ranks_df.iterrows():
+            row = {"Domain": domain, "Model": model}
+            row.update(metrics.to_dict())
+            rows.append(row)
+    if not rows:
+        return pd.DataFrame(columns=["Domain", "Model"] + RANK_COLUMNS)
+    table = pd.DataFrame(rows)
+    return format_df(table.sort_values(by=["Domain", "MASE_Rank"]))