Token trend action

awesome-giza-actions/README.md

@@ -7,6 +7,10 @@ A curated list of models and algorithms built with [Giza Actions SDK](https://ac
 ### Basics
 - [MNIST Classification with Giza Actions](https://github.com/gizatechxyz/actions-sdk/blob/main/examples/verifiable_mnist/verifiable_mnist.ipynb)
 
+## DeFi
+- [Binary Liquidation Prediction in Aave v2 & v3](Aave_Liquidation_Model)
+- [Predicting Cryptocurrency Token Trends with Giza: An End-to-End Machine Learning Approach](trend_token_prediction)
+
 ## Learn the basics
 
 Familiarize yourself with Giza Actions SDK. How to build your own ML workflow and perform verifiableinference.

awesome-giza-actions/trend_token_prediction/download_proof.py

@@ -0,0 +1,19 @@
+import requests
+from giza import API_HOST
+
+MODEL_ID = None  # Update with your model ID
+VERSION_ID = None  # Update with your version ID
+DEPLOYMENT_ID = None  # Update with your deployment id
+REQUEST_ID = None  # Update with your request id
+API_KEY = None  # Update with your API key
+
+url = f"{API_HOST}/api/v1/models/{MODEL_ID}/versions/{VERSION_ID}/deployments/{DEPLOYMENT_ID}/proofs/{REQUEST_ID}:download"
+
+headers = {"X-API-KEY": API_KEY}
+
+d_url = requests.get(url, headers=headers).json()["download_url"]
+
+proof = requests.get(d_url)
+
+with open("zk.proof", "wb") as f:
+    f.write(proof.content)

awesome-giza-actions/trend_token_prediction/predict_cairo_action.py

@@ -0,0 +1,29 @@
+from giza_actions.action import Action, action
+from giza_actions.model import GizaModel
+from giza_actions.task import task
+import polars as pl
+import numpy as np
+import torch
+
+
+@task(name="Prediction with Cairo")
+def prediction(X_test, model):
+    (result, request_id) = model.predict(
+        input_feed={"input_feed": X_test},
+        verifiable=True,
+        output_dtype="arr_fixed_point",
+    )
+    return result, request_id
+
+
+@action(name="Execution: Prediction with Cairo", log_prints=True)
+def execution():
+    model = GizaModel(id=293, version=1)
+    df = pl.read_csv("./example_token_trend.csv")
+    model_input = df.to_numpy().astype(np.float32)
+    (result, request_id) = prediction(model_input, model)
+    return result, request_id
+
+
+(result, request_id) = execution()
+print(f"Result: {result}, Request ID: {request_id}")