Initial commit

.gitattributes

@@ -1,35 +0,0 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,6 @@
+**/*
+
+!app.py
+!utils.py
+!.gitignore
+!requirements.txt

README.md

@@ -1,12 +0,0 @@
----
-title: QNLPDemoApp
-emoji: 📊
-colorFrom: red
-colorTo: pink
-sdk: streamlit
-sdk_version: 1.39.0
-app_file: app.py
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,76 @@
+import streamlit as st
+from utils import QNLP
+import numpy as np
+import matplotlib.pyplot as plt
+import pandas as pd
+
+st.write("### QNLP demo")
+lang = st.selectbox("Select Language", ("EN", "ZH"))
+input = st.text_input("Text Input")
+
+def plot_data(array:np.ndarray):
+    fig, ax = plt.subplots()
+
+    ax.set_xlabel("Value")
+    ax.set_ylabel("% Total")
+    value, count = np.unique(array,return_counts=True)
+    count = count * 100 / len(array)
+    ax.set_xlim([-5,260])
+    ax.bar(value, count, edgecolor="blue", align="edge")
+
+    return fig
+
+if input.strip():
+    results = QNLP(lang.lower()).process_sentence(input)
+    subtabs = []
+    if len(results) > 1:
+        subtabs = [f"Sentence {n+1}" for n in range(len(results))]
+
+    tabs = st.tabs(["Overall"] + subtabs)
+
+    overall_tab = tabs[0]
+    detail_tabs = tabs[1:]
+
+
+    with overall_tab:
+        st.header("Overall")
+        whole_array = np.concatenate([np.sum(res.array, axis=-1) for res in results if res.job.done()])
+
+        total = len(whole_array)
+        value, count = np.unique(whole_array,return_counts=True)
+        count = count/total*100
+
+        df = pd.DataFrame({    
+            "Bit Value" : value, 
+            "Percentage" : count,
+        })
+
+        col1, col2 = st.columns([1,2])
+        with col1:
+            st.dataframe(df, hide_index=True)
+        with col2:
+            st.pyplot(plot_data(whole_array))
+
+    for idx, (tab, result) in enumerate(zip(detail_tabs, results)):
+        with tab:
+            st.header(f"Sentence {idx}")
+            st.write(' '.join(result.tokens))
+            
+            if result.valid:
+                value, count = np.unique(result.array,return_counts=True)
+                count = count/total*100
+
+                df = pd.DataFrame({    
+                    "Bit Value" : value, 
+                    "Percentage" : count,
+                })
+
+                col1, col2 = st.columns([1,2])
+                with col1:
+                    st.dataframe(df, hide_index=True)
+                with col2:
+                    st.pyplot(plot_data(result.array))
+            else:
+                st.write(f"Sentence Discarded due to lack of qubits to process")
+else:
+    st.write("Choose a langauge and input some sentences to start !")

requirements.txt

@@ -0,0 +1,13 @@
+ipython
+joblib
+lambeq
+matplotlib
+numpy
+pandas
+pytket
+pytket_qiskit
+qiskit
+qiskit_aer
+spacy
+streamlit
+tqdm

utils.py

@@ -0,0 +1,97 @@
+import numpy as np
+from qiskit.primitives.primitive_job import PrimitiveJob
+from qiskit import QuantumCircuit
+import re
+import spacy
+import spacy.lang.en
+import spacy.cli
+import logging
+from lambeq import Rewriter, BobcatParser
+from tqdm import tqdm
+from lambeq import AtomicType, IQPAnsatz
+from pytket.extensions.qiskit import tk_to_qiskit
+from qiskit_aer.primitives import SamplerV2
+from dataclasses import dataclass
+from joblib import Parallel, delayed
+
+LANGUAGE_2_SPACY = dict(
+    en = 'en_core_web_sm',
+    zh = 'zh_core_web_sm',
+)
+
+MAPPING = {
+    AtomicType.PREPOSITIONAL_PHRASE: 0, 
+    AtomicType.NOUN: 0, 
+    AtomicType.SENTENCE: 1,
+    AtomicType.CONJUNCTION: 0,
+    AtomicType.PUNCTUATION: 0,
+}
+
+ANSATZ = IQPAnsatz(MAPPING, n_layers=1, discard=True)
+LOGGER = logging.getLogger(__name__) 
+
+@dataclass
+class QNLP_OUTPUT():
+    tokens: str
+    circuit: QuantumCircuit
+    job: PrimitiveJob
+
+    @property
+    def array(self):
+        if not self.job.done():
+            return np.array([])
+        return self.job.result()[0].data.meas.array
+
+    @property
+    def valid(self):
+        return self.job.done()
+
+class QNLP():
+    def __init__(self, langauge = "en") -> None:
+        model = LANGUAGE_2_SPACY.get(langauge)
+        
+        try:
+            self.nlp = spacy.load(model)
+        except OSError:
+            LOGGER.warning('Downloading SpaCy tokeniser. '
+                           'This action only has to happen once.')
+            spacy.cli.download(model)
+            self.nlp = spacy.load(model)
+        
+        self.rewriter = Rewriter([
+            'auxiliary',
+            'connector',
+            'determiner',
+            'postadverb',
+            'preadverb',
+            'prepositional_phrase',
+        ])
+        self.parser = BobcatParser()
+    
+    def process_sentence(self, 
+                         input_sentence: str, 
+                         shots = 1024
+                         ) -> list[QNLP_OUTPUT]:
+        input_sentence = re.sub(r'\n+', '', input_sentence)
+        docs = self.nlp(input_sentence)
+        sentences = list([str(s).strip() for s in chunks] for chunks in docs.sents)
+
+        def sentence2diagrams(sent: list[str], pb: tqdm = None):
+            diagram = self.parser.sentence2diagram(sent, tokenised=True)
+            diagram = self.rewriter(diagram).normal_form()
+
+            if pb: pb.update(1)
+            return diagram
+        
+        pb = tqdm(total = len(sentences), desc="Splitting sentences")
+        diagrams = Parallel(4, require='sharedmem')(delayed(sentence2diagrams)(s, pb) for s in sentences)
+
+        qiskit_circuits = list(tk_to_qiskit(ANSATZ(diagram).to_tk()) for diagram in diagrams)
+        for qc in qiskit_circuits : qc.measure_all()
+
+        sampler = SamplerV2()
+        jobs = [sampler.run([(qc, [1] * qc.num_parameters)], shots=shots) 
+                for qc in qiskit_circuits]
+        
+        return [QNLP_OUTPUT(*params) for params in zip(sentences, qiskit_circuits, jobs)]
+