[ci] Fix plotly bug

fastvideo/tests/modal/pr_test.py

@@ -15,8 +15,6 @@ image = (modal.Image.from_registry(
     image_tag, add_python="3.12"
 ).run_commands("rm -rf /FastVideo").apt_install(
     "cmake", "pkg-config", "build-essential", "curl", "libssl-dev", "ffmpeg"
-).pip_install("huggingface_hub", "pandas", "plotly"
-).run_commands("pip show plotly || echo 'plotly NOT installed'"
 ).run_commands(
     "curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y --default-toolchain stable"
 ).run_commands("echo 'source ~/.cargo/env' >> ~/.bashrc").env({

fastvideo/tests/performance/dashboard.py

@@ -1,9 +1,10 @@
 import os
 import json
+import glob
 import pandas as pd
 from datetime import datetime, timedelta, timezone
 import plotly.express as px
-from huggingface_hub import HfApi, hf_hub_download
+from huggingface_hub import snapshot_download
 
 
 HF_REPO_ID = os.environ.get("HF_REPO_ID", "FastVideo/performance-tracking")
@@ -13,39 +14,67 @@ HF_TOKEN = os.environ.get("HF_API_KEY")
 # -----------------------------
 # 1. Load HF JSON artifacts
 # -----------------------------
-def load_hf_dataset(days: int = 30) -> pd.DataFrame:
-    """
-    Loads only recent benchmark JSONs from HF repo.
-    Avoids full history scan when possible.
-    """
-    api = HfApi(token=HF_TOKEN)
+# def load_hf_dataset(days: int = 30) -> pd.DataFrame:
+#     """
+#     Loads only recent benchmark JSONs from HF repo.
+#     Avoids full history scan when possible.
+#     """
+#     api = HfApi(token=HF_TOKEN)
 
-    files = api.list_repo_files(
+#     files = api.list_repo_files(
+#         repo_id=HF_REPO_ID,
+#         repo_type="dataset",
+#     )
+
+#     cutoff = datetime.now(timezone.utc) - timedelta(days=days)
+#     records = []
+
+#     for f in files:
+#         if not f.endswith(".json"):
+#             continue
+
+#         path = hf_hub_download(
+#             repo_id=HF_REPO_ID,
+#             repo_type="dataset",
+#             filename=f,
+#             token=HF_TOKEN,
+#         )
+
+#         try:
+#             with open(path, "r") as fp:
+#                 data = json.load(fp)
+
+#             # fast skip: timestamp filter BEFORE dataframe creation
+#             ts = pd.to_datetime(data.get("timestamp"), utc=True)
+#             if ts < cutoff:
+#                 continue
+
+#             records.append(data)
+
+#         except Exception:
+#             continue
+
+#     return pd.DataFrame(records)
+
+def load_hf_dataset(days: int = 30) -> pd.DataFrame:
+    cutoff = datetime.now(timezone.utc) - timedelta(days=days)
+
+    local_dir = snapshot_download(
         repo_id=HF_REPO_ID,
         repo_type="dataset",
+        token=HF_TOKEN,
+        allow_patterns="*.json"
     )
 
-    cutoff = datetime.now(timezone.utc) - timedelta(days=days)
     records = []
 
-    for f in files:
-        if not f.endswith(".json"):
-            continue
-
-        path = hf_hub_download(
-            repo_id=HF_REPO_ID,
-            repo_type="dataset",
-            filename=f,
-            token=HF_TOKEN,
-        )
-
+    for path in glob.glob(f"{local_dir}/**/*.json", recursive=True):
         try:
             with open(path, "r") as fp:
                 data = json.load(fp)
 
-            # fast skip: timestamp filter BEFORE dataframe creation
-            ts = pd.to_datetime(data.get("timestamp"), utc=True)
-            if ts < cutoff:
+            ts = pd.to_datetime(data.get("timestamp"), utc=True, errors="coerce")
+            if pd.isna(ts) or ts < cutoff:
                 continue
 
             records.append(data)
@@ -55,7 +84,6 @@ def load_hf_dataset(days: int = 30) -> pd.DataFrame:
 
     return pd.DataFrame(records)
 
-
 # -----------------------------
 # 2. Normalize schema (your CI format)
 # -----------------------------
@@ -79,7 +107,7 @@ def normalize(df: pd.DataFrame) -> pd.DataFrame:
 # 3. Grouping (your requested design)
 # -----------------------------
 def group_data(df: pd.DataFrame):
-    keys = ["benchmark_id", "model_id", "gpu_type", "config_id"]
+    keys = ["model_id", "gpu_type", "config_id"]
     return df.groupby(keys, dropna=False)
 
 
@@ -94,9 +122,9 @@ def build_plots(df: pd.DataFrame):
     for name, g in grouped:
         g = g.sort_values("timestamp")
 
-        benchmark_id, model_id, gpu_type, config_id = name
+        model_id, gpu_type, config_id = name
 
-        title = f"{benchmark_id} | {model_id} | {gpu_type} | {config_id}"
+        title = f"{model_id} | {gpu_type} | {config_id}"
 
         fig = px.line(
             g,
@@ -146,17 +174,30 @@ def annotate_buildkite(html: str):
 # 6. Buildkite upload
 # -----------------------------
 
-def upload_artifact(path: str):
-    if not os.environ.get("BUILDKITE"):
-        return
+# def upload_artifact(path: str):
+#     if not os.environ.get("BUILDKITE"):
+#         return
 
+#     import subprocess
+
+#     subprocess.run(
+#         ["buildkite-agent", "artifact", "upload", path],
+#         check=False,
+#     )
+
+def upload_artifact(path: str):
+    import shutil
     import subprocess
 
+    if shutil.which("buildkite-agent") is None:
+        print("buildkite-agent not found, skipping artifact upload")
+        return
+
     subprocess.run(
         ["buildkite-agent", "artifact", "upload", path],
-        check=False,
+        check=True,
     )
-
+    
 # -----------------------------
 # 7. Main
 # -----------------------------

pyproject.toml

@@ -124,6 +124,8 @@ test = [
     "av",
     "pytorch-msssim==1.0.0",
     "pytest",
+    "pandas",
+    "plotly",
 ]
 
 dev = [ "fastvideo[lint]", "fastvideo[test]", ]