Feat/milvus support (#671)
Co-authored-by: StyleZhang <jasonapring2015@outlook.com> Co-authored-by: JzoNg <jzongcode@gmail.com>
This commit is contained in:
Jyong
@@ -1,123 +0,0 @@
|
||||
import numpy as np
|
||||
import sklearn.decomposition
|
||||
import pickle
|
||||
import time
|
||||
|
||||
|
||||
# Apply 'Algorithm 1' to the ada-002 embeddings to make them isotropic, taken from the paper:
|
||||
# ALL-BUT-THE-TOP: SIMPLE AND EFFECTIVE POST- PROCESSING FOR WORD REPRESENTATIONS
|
||||
# Jiaqi Mu, Pramod Viswanath
|
||||
|
||||
# This uses Principal Component Analysis (PCA) to 'evenly distribute' the embedding vectors (make them isotropic)
|
||||
# For more information on PCA, see https://jamesmccaffrey.wordpress.com/2021/07/16/computing-pca-using-numpy-without-scikit/
|
||||
|
||||
|
||||
# get the file pointer of the pickle containing the embeddings
|
||||
fp = open('/path/to/your/data/Embedding-Latest.pkl', 'rb')
|
||||
|
||||
|
||||
# the embedding data here is a dict consisting of key / value pairs
|
||||
# the key is the hash of the message (SHA3-256), the value is the embedding from ada-002 (array of dimension 1536)
|
||||
# the hash can be used to lookup the orignal text in a database
|
||||
E = pickle.load(fp) # load the data into memory
|
||||
|
||||
# seperate the keys (hashes) and values (embeddings) into seperate vectors
|
||||
K = list(E.keys()) # vector of all the hash values
|
||||
X = np.array(list(E.values())) # vector of all the embeddings, converted to numpy arrays
|
||||
|
||||
|
||||
# list the total number of embeddings
|
||||
# this can be truncated if there are too many embeddings to do PCA on
|
||||
print(f"Total number of embeddings: {len(X)}")
|
||||
|
||||
# get dimension of embeddings, used later
|
||||
Dim = len(X[0])
|
||||
|
||||
# flash out the first few embeddings
|
||||
print("First two embeddings are: ")
|
||||
print(X[0])
|
||||
print(f"First embedding length: {len(X[0])}")
|
||||
print(X[1])
|
||||
print(f"Second embedding length: {len(X[1])}")
|
||||
|
||||
|
||||
# compute the mean of all the embeddings, and flash the result
|
||||
mu = np.mean(X, axis=0) # same as mu in paper
|
||||
print(f"Mean embedding vector: {mu}")
|
||||
print(f"Mean embedding vector length: {len(mu)}")
|
||||
|
||||
|
||||
# subtract the mean vector from each embedding vector ... vectorized in numpy
|
||||
X_tilde = X - mu # same as v_tilde(w) in paper
|
||||
|
||||
|
||||
|
||||
# do the heavy lifting of extracting the principal components
|
||||
# note that this is a function of the embeddings you currently have here, and this set may grow over time
|
||||
# therefore the PCA basis vectors may change over time, and your final isotropic embeddings may drift over time
|
||||
# but the drift should stabilize after you have extracted enough embedding data to characterize the nature of the embedding engine
|
||||
print(f"Performing PCA on the normalized embeddings ...")
|
||||
pca = sklearn.decomposition.PCA() # new object
|
||||
TICK = time.time() # start timer
|
||||
pca.fit(X_tilde) # do the heavy lifting!
|
||||
TOCK = time.time() # end timer
|
||||
DELTA = TOCK - TICK
|
||||
|
||||
print(f"PCA finished in {DELTA} seconds ...")
|
||||
|
||||
# dimensional reduction stage (the only hyperparameter)
|
||||
# pick max dimension of PCA components to express embddings
|
||||
# in general this is some integer less than or equal to the dimension of your embeddings
|
||||
# it could be set as a high percentile, say 95th percentile of pca.explained_variance_ratio_
|
||||
# but just hardcoding a constant here
|
||||
D = 15 # hyperparameter on dimension (out of 1536 for ada-002), paper recommeds D = Dim/100
|
||||
|
||||
|
||||
# form the set of v_prime(w), which is the final embedding
|
||||
# this could be vectorized in numpy to speed it up, but coding it directly here in a double for-loop to avoid errors and to be transparent
|
||||
E_prime = dict() # output dict of the new embeddings
|
||||
N = len(X_tilde)
|
||||
N10 = round(N/10)
|
||||
U = pca.components_ # set of PCA basis vectors, sorted by most significant to least significant
|
||||
print(f"Shape of full set of PCA componenents {U.shape}")
|
||||
U = U[0:D,:] # take the top D dimensions (or take them all if D is the size of the embedding vector)
|
||||
print(f"Shape of downselected PCA componenents {U.shape}")
|
||||
for ii in range(N):
|
||||
v_tilde = X_tilde[ii]
|
||||
v = X[ii]
|
||||
v_projection = np.zeros(Dim) # start to build the projection
|
||||
# project the original embedding onto the PCA basis vectors, use only first D dimensions
|
||||
for jj in range(D):
|
||||
u_jj = U[jj,:] # vector
|
||||
v_jj = np.dot(u_jj,v) # scaler
|
||||
v_projection += v_jj*u_jj # vector
|
||||
v_prime = v_tilde - v_projection # final embedding vector
|
||||
v_prime = v_prime/np.linalg.norm(v_prime) # create unit vector
|
||||
E_prime[K[ii]] = v_prime
|
||||
|
||||
if (ii%N10 == 0) or (ii == N-1):
|
||||
print(f"Finished with {ii+1} embeddings out of {N} ({round(100*ii/N)}% done)")
|
||||
|
||||
|
||||
# save as new pickle
|
||||
print("Saving new pickle ...")
|
||||
embeddingName = '/path/to/your/data/Embedding-Latest-Isotropic.pkl'
|
||||
with open(embeddingName, 'wb') as f: # Python 3: open(..., 'wb')
|
||||
pickle.dump([E_prime,mu,U], f)
|
||||
print(embeddingName)
|
||||
|
||||
print("Done!")
|
||||
|
||||
# When working with live data with a new embedding from ada-002, be sure to tranform it first with this function before comparing it
|
||||
#
|
||||
def projectEmbedding(v,mu,U):
|
||||
v = np.array(v)
|
||||
v_tilde = v - mu
|
||||
v_projection = np.zeros(len(v)) # start to build the projection
|
||||
# project the original embedding onto the PCA basis vectors, use only first D dimensions
|
||||
for u in U:
|
||||
v_jj = np.dot(u,v) # scaler
|
||||
v_projection += v_jj*u # vector
|
||||
v_prime = v_tilde - v_projection # final embedding vector
|
||||
v_prime = v_prime/np.linalg.norm(v_prime) # create unit vector
|
||||
return v_prime
|
||||
@@ -7,6 +7,7 @@ import re
|
||||
import threading
|
||||
import time
|
||||
import uuid
|
||||
from concurrent.futures import ThreadPoolExecutor
|
||||
from multiprocessing import Process
|
||||
from typing import Optional, List, cast
|
||||
|
||||
@@ -14,7 +15,6 @@ import openai
|
||||
from billiard.pool import Pool
|
||||
from flask import current_app, Flask
|
||||
from flask_login import current_user
|
||||
from gevent.threadpool import ThreadPoolExecutor
|
||||
from langchain.embeddings import OpenAIEmbeddings
|
||||
from langchain.schema import Document
|
||||
from langchain.text_splitter import RecursiveCharacterTextSplitter, TextSplitter
|
||||
@@ -516,43 +516,51 @@ class IndexingRunner:
|
||||
model_name='gpt-3.5-turbo',
|
||||
max_tokens=2000
|
||||
)
|
||||
self.format_document(llm, documents, split_documents, document_form)
|
||||
threads = []
|
||||
for doc in documents:
|
||||
document_format_thread = threading.Thread(target=self.format_document, kwargs={
|
||||
'llm': llm, 'document_node': doc, 'split_documents': split_documents, 'document_form': document_form})
|
||||
threads.append(document_format_thread)
|
||||
document_format_thread.start()
|
||||
for thread in threads:
|
||||
thread.join()
|
||||
all_documents.extend(split_documents)
|
||||
|
||||
return all_documents
|
||||
|
||||
def format_document(self, llm: StreamableOpenAI, documents: List[Document], split_documents: List, document_form: str):
|
||||
for document_node in documents:
|
||||
format_documents = []
|
||||
if document_node.page_content is None or not document_node.page_content.strip():
|
||||
return format_documents
|
||||
if document_form == 'text_model':
|
||||
# text model document
|
||||
doc_id = str(uuid.uuid4())
|
||||
hash = helper.generate_text_hash(document_node.page_content)
|
||||
def format_document(self, llm: StreamableOpenAI, document_node, split_documents: List, document_form: str):
|
||||
print(document_node.page_content)
|
||||
format_documents = []
|
||||
if document_node.page_content is None or not document_node.page_content.strip():
|
||||
return format_documents
|
||||
if document_form == 'text_model':
|
||||
# text model document
|
||||
doc_id = str(uuid.uuid4())
|
||||
hash = helper.generate_text_hash(document_node.page_content)
|
||||
|
||||
document_node.metadata['doc_id'] = doc_id
|
||||
document_node.metadata['doc_hash'] = hash
|
||||
document_node.metadata['doc_id'] = doc_id
|
||||
document_node.metadata['doc_hash'] = hash
|
||||
|
||||
format_documents.append(document_node)
|
||||
elif document_form == 'qa_model':
|
||||
try:
|
||||
# qa model document
|
||||
response = LLMGenerator.generate_qa_document_sync(llm, document_node.page_content)
|
||||
document_qa_list = self.format_split_text(response)
|
||||
qa_documents = []
|
||||
for result in document_qa_list:
|
||||
qa_document = Document(page_content=result['question'], metadata=document_node.metadata.copy())
|
||||
doc_id = str(uuid.uuid4())
|
||||
hash = helper.generate_text_hash(result['question'])
|
||||
qa_document.metadata['answer'] = result['answer']
|
||||
qa_document.metadata['doc_id'] = doc_id
|
||||
qa_document.metadata['doc_hash'] = hash
|
||||
qa_documents.append(qa_document)
|
||||
format_documents.extend(qa_documents)
|
||||
except Exception:
|
||||
logging.error("sss")
|
||||
split_documents.extend(format_documents)
|
||||
|
||||
format_documents.append(document_node)
|
||||
elif document_form == 'qa_model':
|
||||
try:
|
||||
# qa model document
|
||||
response = LLMGenerator.generate_qa_document_sync(llm, document_node.page_content)
|
||||
document_qa_list = self.format_split_text(response)
|
||||
qa_documents = []
|
||||
for result in document_qa_list:
|
||||
qa_document = Document(page_content=result['question'], metadata=document_node.metadata.copy())
|
||||
doc_id = str(uuid.uuid4())
|
||||
hash = helper.generate_text_hash(result['question'])
|
||||
qa_document.metadata['answer'] = result['answer']
|
||||
qa_document.metadata['doc_id'] = doc_id
|
||||
qa_document.metadata['doc_hash'] = hash
|
||||
qa_documents.append(qa_document)
|
||||
format_documents.extend(qa_documents)
|
||||
except Exception:
|
||||
continue
|
||||
split_documents.extend(format_documents)
|
||||
|
||||
def _split_to_documents_for_estimate(self, text_docs: List[Document], splitter: TextSplitter,
|
||||
processing_rule: DatasetProcessRule) -> List[Document]:
|
||||
|
||||
Reference in New Issue
Block a user