Refactor Dockerfile to update Python version to 3.11.9, fix `relik se…

…rve`, black everything, update version to 1.0.5

dockerfiles/fastapi/Dockerfile.cpu

@@ -1,4 +1,4 @@
-FROM python:3.10.13-slim-bullseye
+FROM python:3.11.9-slim-bullseye
 
 ARG DEBIAN_FRONTEND=noninteractive
 
@@ -10,6 +10,11 @@ ENV PATH=${PATH}:/home/relik-user/.local/bin
 COPY --chown=relik-user:relik-user . /home/relik-user/relik
 WORKDIR /home/relik-user/relik
 
+# mount huggingface cache dir
+RUN mkdir -p /home/relik-user/.cache/huggingface
+# ENV HF_HOME=/home/relik-user/.cache/huggingface
+# mount huggingface
+
 RUN pip install --upgrade --no-cache-dir .[serve] -c constraints.cpu.txt \
     && chmod +x scripts/docker/start-gunic.sh
 

dockerfiles/fastapi/Dockerfile.cuda

@@ -1,11 +1,11 @@
-FROM nvidia/cuda:12.1.0-base-ubuntu22.04
+FROM nvidia/cuda:12.0.0-base-ubuntu22.04
 
 ARG DEBIAN_FRONTEND=noninteractive
 
 RUN adduser --disabled-password --gecos '' relik-user \
     && apt-get update \
     && apt-get install -y --no-install-recommends \
-    curl wget python3.10 python3-distutils python3-pip \
+    curl wget python3.11 python3-distutils python3-pip \
     && rm -rf /var/lib/apt/lists/*
 
 USER relik-user
@@ -15,6 +15,9 @@ ENV PATH=${PATH}:/home/relik-user/.local/bin
 COPY --chown=relik-user:relik-user . /home/relik-user/relik
 WORKDIR /home/relik-user/relik
 
+RUN mkdir -p /home/relik-user/.cache/huggingface
+
+
 RUN pip install --upgrade --no-cache-dir .[serve] \
     && chmod +x scripts/docker/start-gunic.sh
 

examples/cie.py

@@ -0,0 +1,22 @@
+from relik import Relik
+
+relik = Relik.from_pretrained("relik-ie/relik-cie-small", device="cuda")
+
+text = """When Noah Lyles put his spike into Stade de France’s purple track for his first stride Sunday night of the Paris Olympics 100-meter final, he was already behind. In an event in which margin for error is slimmest, his reaction time to the starting gun was the slowest. 
+
+Halfway through, Lyles, 27, of the U.S., was still in seventh place in an eight-man field, trying to chase down Jamaica’s Kishane Thompson, who owned not only this season’s fastest time but also the fastest time in the semifinal round contested earlier Sunday. 
+
+By the final steps Lyles had caught up so much to Thompson, American Fred Kerley and South Africa’s Akani Simbine that he did something he rarely practices — dipping his shoulder at the finish.
+
+Even then, Lyles was unconvinced he had won the gold medal he had so boldly predicted, and so badly wanted, for three years. The scoreboard offered no indication of who had won gold, silver or bronze as it processed a photo finish, a sold-out, raucous stadium sharing in the uncertainty.
+
+“I think you got that one, big dog,” Lyles told Thompson. 
+
+“I’m not even sure,” Thompson replied. “It was that close.”"""
+
+output = relik(text)
+
+# Entities
+print(output.spans)
+# Relations
+print(output.triplets)

examples/langchain.py

@@ -0,0 +1,21 @@
+from langchain_experimental.graph_transformers import RelikGraphTransformer
+from langchain_core.documents import Document
+
+relik = RelikGraphTransformer("relik-ie/relik-relation-extraction-small-wikipedia")
+
+text = """When Noah Lyles put his spike into Stade de France’s purple track for his first stride Sunday night of the Paris Olympics 100-meter final, he was already behind. In an event in which margin for error is slimmest, his reaction time to the starting gun was the slowest. 
+
+Halfway through, Lyles, 27, of the U.S., was still in seventh place in an eight-man field, trying to chase down Jamaica’s Kishane Thompson, who owned not only this season’s fastest time but also the fastest time in the semifinal round contested earlier Sunday. 
+
+By the final steps Lyles had caught up so much to Thompson, American Fred Kerley and South Africa’s Akani Simbine that he did something he rarely practices — dipping his shoulder at the finish.
+
+Even then, Lyles was unconvinced he had won the gold medal he had so boldly predicted, and so badly wanted, for three years. The scoreboard offered no indication of who had won gold, silver or bronze as it processed a photo finish, a sold-out, raucous stadium sharing in the uncertainty.
+
+“I think you got that one, big dog,” Lyles told Thompson. 
+
+“I’m not even sure,” Thompson replied. “It was that close.”"""
+
+documents = [Document(page_content=text)]
+output = relik.convert_to_graph_documents(documents)
+# triplets
+print(output)

relik/cli/cli.py

@@ -144,61 +144,19 @@ def inference(
 
 @app.command(context_settings=dict(ignore_unknown_options=True, allow_extra_args=True))
 def serve(
-    relik_pretrained: str,
-    device: Annotated[
+    relik_pretrained: Annotated[
         str,
-        typer.Argument(
-            help="The device to use for relik (e.g., 'cuda', 'cpu').",
-        ),
-    ] = None,
-    retriever_device: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The device to use for the retriever (e.g., 'cuda', 'cpu').",
-        ),
-    ] = None,
-    document_index_device: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The device to use for the index (e.g., 'cuda', 'cpu').",
-        ),
-    ] = None,
-    reader_device: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The device to use for the reader (e.g., 'cuda', 'cpu').",
-        ),
-    ] = None,
-    precision: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The precision to use for relik (e.g., '32', '16').",
-        ),
-    ] = "32",
-    retriever_precision: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The precision to use for the retriever (e.g., '32', '16').",
-        ),
-    ] = None,
-    document_index_precision: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The precision to use for the index (e.g., '32', '16').",
-        ),
-    ] = None,
-    reader_precision: Annotated[
-        Union[str, None],
-        typer.Argument(
-            help="The precision to use for the reader (e.g., '32', '16').",
-        ),
-    ] = None,
-    annotation_type: Annotated[
-        str,
-        typer.Argument(
-            help="The type of annotation to use (e.g., 'CHAR', 'WORD').",
-        ),
-    ] = "char",
+        typer.Argument(help="The device to use for relik (e.g., 'cuda', 'cpu')."),
+    ],
+    device: str = "cpu",
+    retriever_device: str = None,
+    document_index_device: str = None,
+    reader_device: str = None,
+    precision: int = 32,
+    retriever_precision: int = None,
+    document_index_precision: int = None,
+    reader_precision: int = None,
+    annotation_type: str = "char",
     host: str = "0.0.0.0",
     port: int = 8000,
     frontend: bool = False,
@@ -216,10 +174,9 @@ def serve(
         annotation_type=annotation_type,
         host=host,
         port=port,
-        frontend=frontend
+        frontend=frontend,
     )
 
 
-
 if __name__ == "__main__":
     app()

relik/inference/data/window/manager.py

@@ -509,7 +509,7 @@ def _merge_window_pair(
 
         # merge text, take into account overlapping chars
         if isinstance(self.splitter, SpacySentenceSplitter):
-            m_text = window1.text[: window2.offset] + " "  + window2.text
+            m_text = window1.text[: window2.offset] + " " + window2.text
         else:
             m_text = window1.text[: window2.offset] + window2.text
 

relik/inference/serve/frontend/gradio_fe.py

@@ -318,7 +318,9 @@ def text_analysis(Text, Relation_Threshold, Window_Size, Window_Stride):
     #     window_stride=Window_Stride,
     # )
     print(f"Using ReLiK at {relik}")
-    print(f"Querying ReLiK with ?text={Text}&relation_threshold={Relation_Threshold}&window_size={Window_Size}&window_stride={Window_Stride}&annotation_type=word&remove_nmes=False")
+    print(
+        f"Querying ReLiK with ?text={Text}&relation_threshold={Relation_Threshold}&window_size={Window_Size}&window_stride={Window_Stride}&annotation_type=word&remove_nmes=False"
+    )
     response = requests.get(
         f"http://{relik}/?text={Text}&relation_threshold={Relation_Threshold}&window_size={Window_Size}&window_stride={Window_Stride}&annotation_type=word&remove_nmes=False",
     )

relik/reader/data/relik_reader_re_data.py

@@ -893,7 +893,8 @@ def flip_cands(flip_candidates, candidates):
                 "end_labels": end_labels,
                 "disambiguation_labels": disambiguation_labels,
                 "relation_labels": relation_labels,
-                "predictable_candidates": candidates_symbols + candidates_entities_symbols,
+                "predictable_candidates": candidates_symbols
+                + candidates_entities_symbols,
             }
         if self.materialize_samples:
             self.samples = data_acc

relik/retriever/pytorch_modules/model.py

@@ -648,15 +648,18 @@ def to_config(cls, *args, **kwargs):
         }
         return config
 
+
 class GoldenSillyRetriever(GoldenRetriever):
     def __init__(self, documents: List[str], *args, **kwargs):
         self.documents = DocumentStore([Document(doc) for doc in documents])
         self.document_index = BaseDocumentIndex(self.documents)
-    def retrieve(self,
-                text: Optional[Union[str, List[str]]] = None,
-                k: int = 100,
-                *args,
-                **kwargs,
+
+    def retrieve(
+        self,
+        text: Optional[Union[str, List[str]]] = None,
+        k: int = 100,
+        *args,
+        **kwargs,
     ) -> List[List[RetrievedSample]]:
         if isinstance(text, str):
             text = [text]
@@ -666,11 +669,12 @@ def retrieve(self,
             [RetrievedSample(score=1.0, document=doc) for doc in self.documents[:k]]
             for _ in text
         ]
+
     def index(self):
-        pass
-    def eval(self):
-        pass
+        raise NotImplementedError
+
     def save_pretrained(self):
-        pass
+        raise NotImplementedError
+
     def to(self, device):
-        pass
+        raise NotImplementedError

relik/version.py

@@ -4,7 +4,7 @@
 _MINOR = "0"
 # On main and in a nightly release the patch should be one ahead of the last
 # released build.
-_PATCH = "4"
+_PATCH = "5"
 # This is mainly for nightly builds which have the suffix ".dev$DATE". See
 # https://semver.org/#is-v123-a-semantic-version for the semantics.
 _SUFFIX = os.environ.get("RELIK_VERSION_SUFFIX", "")

scripts/build_all.sh

@@ -12,4 +12,4 @@ docker build -f dockerfiles/fastapi/Dockerfile.cpu -t relik:$LATEST_VERSION-cpu-
 
 echo "==== Building GPU images ===="
 # docker build -f dockerfiles/ray/Dockerfile.cuda -t relik:$VERSION-cuda-ray .
-docker build -f dockerfiles/fastapi/Dockerfile.cuda -t relik:$LATEST_VERSION-cuda-fastapi .
+# docker build -f dockerfiles/fastapi/Dockerfile.cuda -t relik:$LATEST_VERSION-cuda-fastapi .

scripts/docker/start-gunic.sh

@@ -196,9 +196,7 @@ if [ -z "$GUNICORN_CONF" ]; then
 fi
 
 # Start Ray Serve with the app
-# exec gunicorn -k uvicorn.workers.UvicornWorker -c "$GUNICORN_CONF" "$APP_MODULE" -b 0.0.0.0:8000
-exec relik serve $RELIK_PRETRAINED \
-  --device $DEVICE \
+exec relik serve $RELIK_PRETRAINED --device $DEVICE \
   --retriever-device $RETRIEVER_DEVICE \
   --index-device $INDEX_DEVICE \
   --reader-device $READER_DEVICE \