Hyperliquid End-to-End

Build a model on CrowdCent's training data and submit¶

In [ ]:

!pip install crowdcent-challenge
import crowdcent_challenge as cc
import polars as pl
from xgboost import XGBRegressor

!pip install crowdcent-challenge import crowdcent_challenge as cc import polars as pl from xgboost import XGBRegressor

For this tutorial, you will need:

1. CrowdCent account: register for free
2. CrowdCent API Key: generate an API key from your user profile

Load API key¶

In [ ]:

CROWDCENT_API_KEY = "API_KEY_HERE"

CROWDCENT_API_KEY = "API_KEY_HERE"

Initialize the client¶

In [ ]:

client = cc.ChallengeClient(
    challenge_slug="hyperliquid-ranking",
    api_key=CROWDCENT_API_KEY,
)

client = cc.ChallengeClient( challenge_slug="hyperliquid-ranking", api_key=CROWDCENT_API_KEY, )

2025-07-25 16:47:55,298 - INFO - ChallengeClient initialized for 'hyperliquid-ranking' at URL: https://crowdcent.com/api

Get CrowdCent's training data¶

In [4]:

client.download_training_dataset(version="latest", dest_path="training_data.parquet")

training_data = pl.read_parquet("training_data.parquet")
training_data.head()

client.download_training_dataset(version="latest", dest_path="training_data.parquet") training_data = pl.read_parquet("training_data.parquet") training_data.head()

2025-07-25 16:47:57,090 - INFO - Downloading training data for challenge 'hyperliquid-ranking' v1.0 to training_data.parquet
Downloading training_data.parquet: 100%|██████████| 85.1M/85.1M [00:04<00:00, 18.3MB/s]
2025-07-25 16:48:02,297 - INFO - Successfully downloaded training data to training_data.parquet

Out[4]:

shape: (5, 85)

id	eodhd_id	date	feature_16_lag15	feature_13_lag15	feature_14_lag15	feature_15_lag15	feature_8_lag15	feature_5_lag15	feature_6_lag15	feature_7_lag15	feature_12_lag15	feature_9_lag15	feature_10_lag15	feature_11_lag15	feature_4_lag15	feature_1_lag15	feature_2_lag15	feature_3_lag15	feature_20_lag15	feature_17_lag15	feature_18_lag15	feature_19_lag15	feature_16_lag10	feature_13_lag10	feature_14_lag10	feature_15_lag10	feature_8_lag10	feature_5_lag10	feature_6_lag10	feature_7_lag10	feature_12_lag10	feature_9_lag10	feature_10_lag10	feature_11_lag10	feature_4_lag10	feature_1_lag10	…	feature_5_lag5	feature_6_lag5	feature_7_lag5	feature_12_lag5	feature_9_lag5	feature_10_lag5	feature_11_lag5	feature_4_lag5	feature_1_lag5	feature_2_lag5	feature_3_lag5	feature_20_lag5	feature_17_lag5	feature_18_lag5	feature_19_lag5	feature_16_lag0	feature_13_lag0	feature_14_lag0	feature_15_lag0	feature_8_lag0	feature_5_lag0	feature_6_lag0	feature_7_lag0	feature_12_lag0	feature_9_lag0	feature_10_lag0	feature_11_lag0	feature_4_lag0	feature_1_lag0	feature_2_lag0	feature_3_lag0	feature_20_lag0	feature_17_lag0	feature_18_lag0	feature_19_lag0	target_10d	target_30d
str	str	datetime[μs]	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	…	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64
"BRETT"	"BRETT29743-USD.CC"	2024-05-06 00:00:00	0.55721	0.540086	0.512754	0.508386	0.692426	0.55288	0.549131	0.533897	0.736748	0.617263	0.603953	0.623957	0.450295	0.479222	0.519303	0.528181	0.527312	0.525878	0.497428	0.518564	0.239416	0.398313	0.451205	0.487323	0.118248	0.405337	0.448927	0.492024	0.086131	0.41144	0.485798	0.550351	0.236496	0.343396	…	0.177372	0.365126	0.493638	0.351825	0.218978	0.418121	0.546089	0.268613	0.252555	0.365888	0.495271	0.420438	0.551095	0.538486	0.542948	0.233333	0.249513	0.323913	0.447919	0.289855	0.263176	0.334256	0.435102	0.428986	0.390405	0.400922	0.508626	0.272464	0.270538	0.306967	0.431848	0.401449	0.410944	0.507738	0.507471	0.507246	0.985507
"BRETT"	"BRETT29743-USD.CC"	2024-05-07 00:00:00	0.456339	0.471133	0.514562	0.508886	0.54709	0.519471	0.515427	0.532992	0.613331	0.547406	0.606655	0.627392	0.563407	0.461703	0.512452	0.548603	0.517968	0.518984	0.491352	0.514658	0.233577	0.344958	0.446484	0.48631	0.116788	0.331939	0.402621	0.492332	0.056934	0.335133	0.450072	0.55014	0.110949	0.337178	…	0.175161	0.347316	0.470913	0.351814	0.204374	0.37589	0.527665	0.230646	0.170798	0.31625	0.470839	0.374209	0.602433	0.560709	0.53946	0.272464	0.269793	0.307375	0.428926	0.407246	0.32039	0.326165	0.426455	0.484058	0.417936	0.376534	0.491264	0.365217	0.297932	0.317555	0.420723	0.281159	0.327684	0.500999	0.487154	0.869565	0.985507
"BRETT"	"BRETT29743-USD.CC"	2024-05-08 00:00:00	0.566488	0.453614	0.507711	0.529308	0.383759	0.434102	0.471631	0.518237	0.448497	0.491656	0.563988	0.615796	0.37807	0.391257	0.465007	0.524881	0.46842	0.55421	0.480155	0.521811	0.106569	0.336529	0.407477	0.490267	0.287591	0.335675	0.417691	0.514816	0.230657	0.339577	0.45308	0.574943	0.291971	0.33502	…	0.262028	0.348065	0.471507	0.332974	0.281815	0.386736	0.525698	0.239405	0.265688	0.328473	0.472002	0.498498	0.616402	0.585306	0.538537	0.365217	0.297916	0.317222	0.417801	0.255072	0.245769	0.290722	0.407634	0.314493	0.323733	0.331655	0.468477	0.171014	0.20521	0.270115	0.396829	0.205797	0.352147	0.476755	0.47601	0.913043	0.985507
"BRETT"	"BRETT29743-USD.CC"	2024-05-09 00:00:00	0.382578	0.383157	0.46101	0.505768	0.401234	0.42504	0.441107	0.496157	0.321855	0.413705	0.524751	0.593716	0.40717	0.410203	0.472939	0.505173	0.590994	0.579914	0.497107	0.513647	0.286131	0.334355	0.427468	0.513196	0.268613	0.334924	0.391343	0.490949	0.337226	0.329541	0.440073	0.566184	0.254015	0.330593	…	0.261943	0.343492	0.450375	0.403047	0.370136	0.391921	0.52686	0.242272	0.248143	0.329173	0.449071	0.554988	0.581873	0.580894	0.524828	0.171014	0.205194	0.269774	0.393907	0.268116	0.261695	0.298309	0.387545	0.276812	0.339929	0.334735	0.464148	0.321739	0.282006	0.306299	0.392148	0.265217	0.410103	0.50499	0.468336	0.934783	0.992754
"BRETT"	"BRETT29743-USD.CC"	2024-05-10 00:00:00	0.407299	0.398432	0.467491	0.485709	0.268613	0.414399	0.446095	0.489301	0.232117	0.413673	0.488628	0.5732	0.240876	0.396751	0.456653	0.506786	0.60292	0.598595	0.535617	0.530595	0.252555	0.329927	0.390503	0.489134	0.235036	0.251825	0.366699	0.490923	0.348905	0.290511	0.440729	0.546316	0.267153	0.254015	…	0.268333	0.341366	0.43541	0.420322	0.384613	0.399143	0.530058	0.233365	0.250259	0.323505	0.451219	0.38994	0.498619	0.548607	0.510192	0.321739	0.28199	0.305958	0.389226	0.249275	0.275452	0.263639	0.385433	0.344928	0.382625	0.336568	0.452784	0.343478	0.288422	0.271218	0.394539	0.273913	0.331926	0.468518	0.468964	0.949275	0.985507

Train a model on the training data¶

In [5]:

xgb_regressor = XGBRegressor(n_estimators=2000)
feature_names = [col for col in training_data.columns if col.startswith("feature")]

xgb_regressor.fit(
    training_data[feature_names],
    training_data[["target_10d", "target_30d"]],
)

xgb_regressor = XGBRegressor(n_estimators=2000) feature_names = [col for col in training_data.columns if col.startswith("feature")] xgb_regressor.fit( training_data[feature_names], training_data[["target_10d", "target_30d"]], )

Out[5]:

XGBRegressor(base_score=None, booster=None, callbacks=None,
             colsample_bylevel=None, colsample_bynode=None,
             colsample_bytree=None, device=None, early_stopping_rounds=None,
             enable_categorical=False, eval_metric=None, feature_types=None,
             feature_weights=None, gamma=None, grow_policy=None,
             importance_type=None, interaction_constraints=None,
             learning_rate=None, max_bin=None, max_cat_threshold=None,
             max_cat_to_onehot=None, max_delta_step=None, max_depth=None,
             max_leaves=None, min_child_weight=None, missing=nan,
             monotone_constraints=None, multi_strategy=None, n_estimators=2000,
             n_jobs=None, num_parallel_tree=None, ...)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Get CrowdCent's latest inference data¶

In [6]:

client.download_inference_data("current", "inference_data.parquet")

inference_data = pl.read_parquet("inference_data.parquet")
inference_data.head()

client.download_inference_data("current", "inference_data.parquet") inference_data = pl.read_parquet("inference_data.parquet") inference_data.head()

2025-07-25 16:48:34,787 - INFO - Downloading inference data for challenge 'hyperliquid-ranking' current to inference_data.parquet
Downloading inference_data.parquet: 100%|██████████| 125k/125k [00:00<00:00, 3.39MB/s]
2025-07-25 16:48:35,220 - INFO - Successfully downloaded inference data to inference_data.parquet
2025-07-25 16:48:35,221 - INFO - Successfully downloaded inference data after 1 attempt(s) to inference_data.parquet

Out[6]:

shape: (5, 83)

id	eodhd_id	date	feature_16_lag15	feature_13_lag15	feature_14_lag15	feature_15_lag15	feature_8_lag15	feature_5_lag15	feature_6_lag15	feature_7_lag15	feature_12_lag15	feature_9_lag15	feature_10_lag15	feature_11_lag15	feature_4_lag15	feature_1_lag15	feature_2_lag15	feature_3_lag15	feature_20_lag15	feature_17_lag15	feature_18_lag15	feature_19_lag15	feature_16_lag10	feature_13_lag10	feature_14_lag10	feature_15_lag10	feature_8_lag10	feature_5_lag10	feature_6_lag10	feature_7_lag10	feature_12_lag10	feature_9_lag10	feature_10_lag10	feature_11_lag10	feature_4_lag10	feature_1_lag10	…	feature_15_lag5	feature_8_lag5	feature_5_lag5	feature_6_lag5	feature_7_lag5	feature_12_lag5	feature_9_lag5	feature_10_lag5	feature_11_lag5	feature_4_lag5	feature_1_lag5	feature_2_lag5	feature_3_lag5	feature_20_lag5	feature_17_lag5	feature_18_lag5	feature_19_lag5	feature_16_lag0	feature_13_lag0	feature_14_lag0	feature_15_lag0	feature_8_lag0	feature_5_lag0	feature_6_lag0	feature_7_lag0	feature_12_lag0	feature_9_lag0	feature_10_lag0	feature_11_lag0	feature_4_lag0	feature_1_lag0	feature_2_lag0	feature_3_lag0	feature_20_lag0	feature_17_lag0	feature_18_lag0	feature_19_lag0
str	str	datetime[μs]	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	…	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64	f64
"ENS"	"ENS-USD.CC"	2025-07-24 00:00:00	0.563743	0.553216	0.510526	0.529579	0.588304	0.551462	0.50731	0.542609	0.51345	0.494152	0.461111	0.492697	0.590643	0.577193	0.516082	0.527347	0.595322	0.545029	0.521053	0.499538	0.569591	0.566667	0.544737	0.539454	0.552047	0.570175	0.542398	0.532056	0.6	0.556725	0.511404	0.501226	0.603509	0.597076	…	0.575047	0.818713	0.68538	0.618421	0.572337	0.721637	0.660819	0.577485	0.545107	0.752047	0.677778	0.627485	0.579438	0.549708	0.551462	0.548246	0.514757	0.54152	0.667251	0.616959	0.572807	0.585965	0.702339	0.636257	0.575731	0.464327	0.592982	0.574854	0.524708	0.603509	0.677778	0.637427	0.579094	0.518129	0.533918	0.554094	0.517544
"DOOD"	"DOOD-USD.CC"	2025-07-24 00:00:00	0.567251	0.6	0.47076	0.482089	0.291228	0.356725	0.384795	0.423368	0.693567	0.676023	0.523392	0.56311	0.418713	0.517544	0.456433	0.479155	0.293567	0.433333	0.487135	0.503057	0.419883	0.493567	0.540351	0.456486	0.562573	0.426901	0.458772	0.429866	0.550877	0.622222	0.630702	0.529764	0.409357	0.414035	…	0.43514	0.410526	0.48655	0.421637	0.444858	0.568421	0.559649	0.617836	0.541846	0.396491	0.402924	0.460234	0.451797	0.424561	0.484795	0.459064	0.496462	0.819883	0.525146	0.509357	0.501754	0.422222	0.416374	0.421637	0.450585	0.749708	0.659064	0.640643	0.59152	0.797661	0.597076	0.505556	0.50614	0.518129	0.471345	0.445322	0.50614
"ETH"	"ETH-USD.CC"	2025-07-24 00:00:00	0.698246	0.647953	0.586257	0.58237	0.678363	0.649123	0.580702	0.601509	0.651462	0.628655	0.550292	0.557504	0.653801	0.684795	0.578947	0.57374	0.561404	0.571345	0.538596	0.541827	0.505263	0.601754	0.583041	0.584499	0.437427	0.557895	0.566667	0.581041	0.480702	0.566082	0.560234	0.551857	0.54386	0.59883	…	0.615683	0.8	0.618713	0.633918	0.610746	0.791813	0.636257	0.632456	0.596275	0.706433	0.625146	0.654971	0.61135	0.57193	0.522222	0.546784	0.523853	0.536842	0.663158	0.632456	0.596199	0.587135	0.693567	0.625731	0.597368	0.561404	0.676608	0.621345	0.579386	0.654971	0.680702	0.639766	0.606433	0.519298	0.545614	0.531287	0.527485
"TAO"	"TAO22974-USD.CC"	2025-07-24 00:00:00	0.392982	0.448538	0.438304	0.485539	0.547368	0.487135	0.462573	0.4896	0.384795	0.439181	0.444737	0.499779	0.417544	0.475439	0.472222	0.489614	0.561404	0.511111	0.523392	0.499633	0.624561	0.508772	0.479532	0.517433	0.680702	0.614035	0.512281	0.525129	0.663158	0.523977	0.490936	0.525174	0.635088	0.526316	…	0.500015	0.404678	0.54269	0.514912	0.492479	0.554386	0.608772	0.523977	0.514325	0.359064	0.497076	0.486257	0.485795	0.467836	0.496491	0.503801	0.495814	0.545029	0.516959	0.512865	0.497515	0.625731	0.515205	0.56462	0.516374	0.500585	0.527485	0.525731	0.51462	0.574269	0.466667	0.496491	0.5	0.48655	0.477193	0.510234	0.508626
"PNUT"	"PNUT-USD.CC"	2025-07-24 00:00:00	0.34386	0.479532	0.487135	0.49155	0.37193	0.483041	0.459357	0.468705	0.533333	0.578363	0.553509	0.5457	0.346199	0.415205	0.467836	0.496082	0.563743	0.552047	0.515205	0.491196	0.535673	0.439766	0.49152	0.492038	0.569591	0.47076	0.488596	0.479315	0.589474	0.561404	0.573977	0.561296	0.509942	0.42807	…	0.484984	0.31462	0.442105	0.462573	0.452025	0.415205	0.502339	0.540351	0.531432	0.484211	0.497076	0.45614	0.475645	0.409357	0.491228	0.521637	0.493562	0.617544	0.505848	0.472807	0.515936	0.529825	0.422222	0.446491	0.479094	0.535673	0.475439	0.518421	0.548392	0.51462	0.499415	0.463743	0.504678	0.397661	0.403509	0.485965	0.488304

Make predictions on the inference data¶

In [7]:

preds = xgb_regressor.predict(inference_data[feature_names])
pred_df = pl.from_numpy(preds, ["pred_10d", "pred_30d"])

preds = xgb_regressor.predict(inference_data[feature_names]) pred_df = pl.from_numpy(preds, ["pred_10d", "pred_30d"])

Submit to the hyperliquid-ranking challenge on CrowdCent¶

In [8]:

pred_df = pred_df.with_columns(inference_data["id"]).select(
    ["id", "pred_10d", "pred_30d"]
)

# ensure predictions are between 0 and 1
pred_df = pred_df.with_columns(pl.col(["pred_10d", "pred_30d"]).clip(0, 1))


with pl.Config(tbl_rows=20):
    display(pred_df.sort("pred_30d", descending=True))

pred_df = pred_df.with_columns(inference_data["id"]).select( ["id", "pred_10d", "pred_30d"] ) # ensure predictions are between 0 and 1 pred_df = pred_df.with_columns(pl.col(["pred_10d", "pred_30d"]).clip(0, 1)) with pl.Config(tbl_rows=20): display(pred_df.sort("pred_30d", descending=True))

shape: (171, 3)

id	pred_10d	pred_30d
str	f32	f32
"ENS"	0.64009	0.766982
"PEOPLE"	0.655082	0.759009
"APE"	0.553019	0.754386
"TRUMP"	0.540669	0.751174
"FXS"	0.746214	0.745177
"LDO"	0.449098	0.717584
"LAYER"	0.513393	0.710931
"TURBO"	0.586166	0.700337
"BERA"	0.708864	0.698633
"RENDER"	0.440773	0.692121
…	…	…
"HYPER"	0.279913	0.25558
"USUAL"	0.607879	0.239557
"TRB"	0.481452	0.223786
"FTT"	0.25645	0.202582
"LAUNCHCOIN"	0.640741	0.19282
"IMX"	0.322157	0.177864
"HMSTR"	0.525798	0.175566
"kNEIRO"	0.178402	0.085106
"INIT"	0.397368	0.06886
"JTO"	0.592458	0.0

In [ ]:

# directly submit a dataframe to slot 1
client.submit_predictions(df=pred_df, slot=1)

# directly submit a dataframe to slot 1 client.submit_predictions(df=pred_df, slot=1)

2025-07-25 16:48:45,241 - INFO - Wrote DataFrame to temporary file: submission.parquet
2025-07-25 16:48:45,242 - INFO - Submitting predictions from submission.parquet to challenge 'hyperliquid-ranking' (Slot: 1)
2025-07-25 16:48:45,839 - INFO - Successfully submitted predictions to challenge 'hyperliquid-ranking'

Out[ ]:

{'match_info': {'matched_ids': 171,
  'unmatched_ids': 0,
  'message': '171 IDs matched inference data'},
 'id': 814,
 'username': 'jrai',
 'challenge_slug': 'hyperliquid-ranking',
 'inference_data_release_date': '2025-07-25T14:00:29.177700Z',
 'submitted_at': '2025-07-25T16:48:45.671926Z',
 'status': 'pending',
 'slot': 1,
 'score_details': None,
 'percentile_details': None,
 'prediction_file': 'https://cc-challenge-storage.s3.amazonaws.com/challenges/hyperliquid-ranking/submissions/4/57/submission_9QQtZah.parquet?AWSAccessKeyId=AKIATBWNDLY2W4VNIT4P&Signature=Tlvu7YNhkONh2c52JQLjlQ1%2FQjQ%3D&Expires=1753465725'}