class MilvusConnection:
def __init__(self, env, name="movies_L2", port="19530", param=None):
if param is None:
param = dict()
param = {
"collection_name": name,
"dimension": 128,
"index_file_size": 1024,
"metric_type": MetricType.L2,
**param,
}
self.name = name
self.client = Milvus(host="localhost", port=port)
self.statuses = {}
if not self.client.has_collection(name)[1]:
status_created_collection = self.client.create_collection(param)
vectors = env.base.embeddings.detach().cpu().numpy().astype(
"float32")
target_ids = list(range(vectors.shape[0]))
status_inserted, inserted_vector_ids = self.client.insert(
collection_name=name, records=vectors, ids=target_ids)
status_flushed = self.client.flush([name])
status_compacted = self.client.compact(collection_name=name)
self.statuses["created_collection"] = status_created_collection
self.statuses["inserted"] = status_inserted
self.statuses["flushed"] = status_flushed
self.statuses["compacted"] = status_compacted
def search(self, search_vecs, topk=10, search_param=None):
if search_param is None:
search_param = dict()
search_param = {"nprobe": 16, **search_param}
status, results = self.client.search(
collection_name=self.name,
query_records=search_vecs,
top_k=topk,
params=search_param,
)
self.statuses['last_search'] = status
return torch.tensor(results.id_array)
def get_log(self):
return self.statuses
class MilvusClient(object):
def __init__(self,
collection_name=None,
host=None,
port=None,
timeout=180):
self._collection_name = collection_name
start_time = time.time()
if not host:
host = SERVER_HOST_DEFAULT
if not port:
port = SERVER_PORT_DEFAULT
logger.debug(host)
logger.debug(port)
# retry connect remote server
i = 0
while time.time() < start_time + timeout:
try:
self._milvus = Milvus(host=host,
port=port,
try_connect=False,
pre_ping=False)
break
except Exception as e:
logger.error(str(e))
logger.error("Milvus connect failed: %d times" % i)
i = i + 1
time.sleep(i)
if time.time() > start_time + timeout:
raise Exception("Server connect timeout")
# self._metric_type = None
def __str__(self):
return 'Milvus collection %s' % self._collection_name
def check_status(self, status):
if not status.OK():
logger.error(status.message)
logger.error(self._milvus.server_status())
logger.error(self.count())
raise Exception("Status not ok")
def check_result_ids(self, result):
for index, item in enumerate(result):
if item[0].distance >= epsilon:
logger.error(index)
logger.error(item[0].distance)
raise Exception("Distance wrong")
# only support the given field name
def create_collection(self,
dimension,
data_type=DataType.FLOAT_VECTOR,
auto_id=False,
collection_name=None,
other_fields=None):
self._dimension = dimension
if not collection_name:
collection_name = self._collection_name
vec_field_name = utils.get_default_field_name(data_type)
fields = [{
"name": vec_field_name,
"type": data_type,
"params": {
"dim": dimension
}
}]
if other_fields:
other_fields = other_fields.split(",")
if "int" in other_fields:
fields.append({
"name": utils.DEFAULT_INT_FIELD_NAME,
"type": DataType.INT64
})
if "float" in other_fields:
fields.append({
"name": utils.DEFAULT_FLOAT_FIELD_NAME,
"type": DataType.FLOAT
})
create_param = {"fields": fields, "auto_id": auto_id}
try:
self._milvus.create_collection(collection_name, create_param)
logger.info("Create collection: <%s> successfully" %
collection_name)
except Exception as e:
logger.error(str(e))
raise
def create_partition(self, tag, collection_name=None):
if not collection_name:
collection_name = self._collection_name
self._milvus.create_partition(collection_name, tag)
def generate_values(self, data_type, vectors, ids):
values = None
if data_type in [DataType.INT32, DataType.INT64]:
values = ids
elif data_type in [DataType.FLOAT, DataType.DOUBLE]:
values = [(i + 0.0) for i in ids]
elif data_type in [DataType.FLOAT_VECTOR, DataType.BINARY_VECTOR]:
values = vectors
return values
def generate_entities(self, vectors, ids=None, collection_name=None):
entities = []
if collection_name is None:
collection_name = self._collection_name
info = self.get_info(collection_name)
for field in info["fields"]:
field_type = field["type"]
entities.append({
"name":
field["name"],
"type":
field_type,
"values":
self.generate_values(field_type, vectors, ids)
})
return entities
@time_wrapper
def insert(self, entities, ids=None, collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
try:
insert_ids = self._milvus.insert(tmp_collection_name,
entities,
ids=ids)
return insert_ids
except Exception as e:
logger.error(str(e))
def get_dimension(self):
info = self.get_info()
for field in info["fields"]:
if field["type"] in [
DataType.FLOAT_VECTOR, DataType.BINARY_VECTOR
]:
return field["params"]["dim"]
def get_rand_ids(self, length):
segment_ids = []
while True:
stats = self.get_stats()
segments = stats["partitions"][0]["segments"]
# random choice one segment
segment = random.choice(segments)
try:
segment_ids = self._milvus.list_id_in_segment(
self._collection_name, segment["id"])
except Exception as e:
logger.error(str(e))
if not len(segment_ids):
continue
elif len(segment_ids) > length:
return random.sample(segment_ids, length)
else:
logger.debug("Reset length: %d" % len(segment_ids))
return segment_ids
# def get_rand_ids_each_segment(self, length):
# res = []
# status, stats = self._milvus.get_collection_stats(self._collection_name)
# self.check_status(status)
# segments = stats["partitions"][0]["segments"]
# segments_num = len(segments)
# # random choice from each segment
# for segment in segments:
# status, segment_ids = self._milvus.list_id_in_segment(self._collection_name, segment["name"])
# self.check_status(status)
# res.extend(segment_ids[:length])
# return segments_num, res
# def get_rand_entities(self, length):
# ids = self.get_rand_ids(length)
# status, get_res = self._milvus.get_entity_by_id(self._collection_name, ids)
# self.check_status(status)
# return ids, get_res
def get(self):
get_ids = random.randint(1, 1000000)
self._milvus.get_entity_by_id(self._collection_name, [get_ids])
@time_wrapper
def get_entities(self, get_ids):
get_res = self._milvus.get_entity_by_id(self._collection_name, get_ids)
return get_res
@time_wrapper
def delete(self, ids, collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
self._milvus.delete_entity_by_id(tmp_collection_name, ids)
def delete_rand(self):
delete_id_length = random.randint(1, 100)
count_before = self.count()
logger.debug("%s: length to delete: %d" %
(self._collection_name, delete_id_length))
delete_ids = self.get_rand_ids(delete_id_length)
self.delete(delete_ids)
self.flush()
logger.info("%s: count after delete: %d" %
(self._collection_name, self.count()))
get_res = self._milvus.get_entity_by_id(self._collection_name,
delete_ids)
for item in get_res:
assert not item
# if count_before - len(delete_ids) < self.count():
# logger.error(delete_ids)
# raise Exception("Error occured")
@time_wrapper
def flush(self, _async=False, collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
self._milvus.flush([tmp_collection_name], _async=_async)
@time_wrapper
def compact(self, collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
status = self._milvus.compact(tmp_collection_name)
self.check_status(status)
@time_wrapper
def create_index(self,
field_name,
index_type,
metric_type,
_async=False,
index_param=None):
index_type = INDEX_MAP[index_type]
metric_type = utils.metric_type_trans(metric_type)
logger.info(
"Building index start, collection_name: %s, index_type: %s, metric_type: %s"
% (self._collection_name, index_type, metric_type))
if index_param:
logger.info(index_param)
index_params = {
"index_type": index_type,
"metric_type": metric_type,
"params": index_param
}
self._milvus.create_index(self._collection_name,
field_name,
index_params,
_async=_async)
# TODO: need to check
def describe_index(self, field_name):
# stats = self.get_stats()
info = self._milvus.describe_index(self._collection_name, field_name)
index_info = {"index_type": "flat", "index_param": None}
for field in info["fields"]:
for index in field['indexes']:
if not index or "index_type" not in index:
continue
else:
for k, v in INDEX_MAP.items():
if index['index_type'] == v:
index_info['index_type'] = k
index_info['index_param'] = index['params']
return index_info
return index_info
def drop_index(self, field_name):
logger.info("Drop index: %s" % self._collection_name)
return self._milvus.drop_index(self._collection_name, field_name)
@time_wrapper
def query(self, vector_query, filter_query=None, collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
must_params = [vector_query]
if filter_query:
must_params.extend(filter_query)
query = {"bool": {"must": must_params}}
result = self._milvus.search(tmp_collection_name, query)
return result
@time_wrapper
def load_and_query(self,
vector_query,
filter_query=None,
collection_name=None):
tmp_collection_name = self._collection_name if collection_name is None else collection_name
must_params = [vector_query]
if filter_query:
must_params.extend(filter_query)
query = {"bool": {"must": must_params}}
self.load_collection(tmp_collection_name)
result = self._milvus.search(tmp_collection_name, query)
return result
def get_ids(self, result):
idss = result._entities.ids
ids = []
len_idss = len(idss)
len_r = len(result)
top_k = len_idss // len_r
for offset in range(0, len_idss, top_k):
ids.append(idss[offset:min(offset + top_k, len_idss)])
return ids
def query_rand(self, nq_max=100):
# for ivf search
dimension = 128
top_k = random.randint(1, 100)
nq = random.randint(1, nq_max)
nprobe = random.randint(1, 100)
search_param = {"nprobe": nprobe}
query_vectors = [[random.random() for _ in range(dimension)]
for _ in range(nq)]
metric_type = random.choice(["l2", "ip"])
logger.info("%s, Search nq: %d, top_k: %d, nprobe: %d" %
(self._collection_name, nq, top_k, nprobe))
vec_field_name = utils.get_default_field_name()
vector_query = {
"vector": {
vec_field_name: {
"topk": top_k,
"query": query_vectors,
"metric_type": utils.metric_type_trans(metric_type),
"params": search_param
}
}
}
self.query(vector_query)
def load_query_rand(self, nq_max=100):
# for ivf search
dimension = 128
top_k = random.randint(1, 100)
nq = random.randint(1, nq_max)
nprobe = random.randint(1, 100)
search_param = {"nprobe": nprobe}
query_vectors = [[random.random() for _ in range(dimension)]
for _ in range(nq)]
metric_type = random.choice(["l2", "ip"])
logger.info("%s, Search nq: %d, top_k: %d, nprobe: %d" %
(self._collection_name, nq, top_k, nprobe))
vec_field_name = utils.get_default_field_name()
vector_query = {
"vector": {
vec_field_name: {
"topk": top_k,
"query": query_vectors,
"metric_type": utils.metric_type_trans(metric_type),
"params": search_param
}
}
}
self.load_and_query(vector_query)
# TODO: need to check
def count(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
row_count = self._milvus.get_collection_stats(
collection_name)["row_count"]
logger.debug("Row count: %d in collection: <%s>" %
(row_count, collection_name))
return row_count
def drop(self, timeout=120, collection_name=None):
timeout = int(timeout)
if collection_name is None:
collection_name = self._collection_name
logger.info("Start delete collection: %s" % collection_name)
self._milvus.drop_collection(collection_name)
i = 0
while i < timeout:
try:
row_count = self.count(collection_name=collection_name)
if row_count:
time.sleep(1)
i = i + 1
continue
else:
break
except Exception as e:
logger.debug(str(e))
break
if i >= timeout:
logger.error("Delete collection timeout")
def get_stats(self):
return self._milvus.get_collection_stats(self._collection_name)
def get_info(self, collection_name=None):
# pdb.set_trace()
if collection_name is None:
collection_name = self._collection_name
return self._milvus.get_collection_info(collection_name)
def show_collections(self):
return self._milvus.list_collections()
def exists_collection(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
res = self._milvus.has_collection(collection_name)
return res
def clean_db(self):
collection_names = self.show_collections()
for name in collection_names:
self.drop(collection_name=name)
@time_wrapper
def load_collection(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
return self._milvus.load_collection(collection_name, timeout=3000)
@time_wrapper
def release_collection(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
return self._milvus.release_collection(collection_name, timeout=3000)
@time_wrapper
def load_partitions(self, tag_names, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
return self._milvus.load_partitions(collection_name,
tag_names,
timeout=3000)
@time_wrapper
def release_partitions(self, tag_names, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
return self._milvus.release_partitions(collection_name,
tag_names,
timeout=3000)
def main():
# Specify server addr when create milvus client instance
milvus = Milvus(_HOST, _PORT)
# Create collection demo_collection if it dosen't exist.
collection_name = 'example_async_collection_'
status, ok = milvus.has_collection(collection_name)
if not ok:
param = {
'collection_name': collection_name,
'dimension': _DIM,
'index_file_size': 128, # optional
'metric_type': MetricType.L2 # optional
}
status = milvus.create_collection(param)
if not status.OK():
print("Create collection failed: {}".format(status.message),
file=sys.stderr)
print("exiting ...", file=sys.stderr)
sys.exit(1)
# Show collections in Milvus server
_, collections = milvus.list_collections()
# Describe demo_collection
_, collection = milvus.get_collection_info(collection_name)
print(collection)
# 10000 vectors with 16 dimension
# element per dimension is float32 type
# vectors should be a 2-D array
vectors = [[random.random() for _ in range(_DIM)] for _ in range(100000)]
# You can also use numpy to generate random vectors:
# `vectors = np.random.rand(10000, 16).astype(np.float32)`
def _insert_callback(status, ids):
if status.OK():
print("Insert successfully")
else:
print("Insert failed.", status.message)
# Insert vectors into demo_collection, adding callback function
insert_future = milvus.insert(collection_name=collection_name,
records=vectors,
_async=True,
_callback=_insert_callback)
# Or invoke result() to get results:
# insert_future = milvus.insert(collection_name=collection_name, records=vectors, _async=True)
# status, ids = insert_future.result()
insert_future.done()
# Flush collection inserted data to disk.
def _flush_callback(status):
if status.OK():
print("Flush successfully")
else:
print("Flush failed.", status.message)
flush_future = milvus.flush([collection_name],
_async=True,
_callback=_flush_callback)
# Or invoke result() to get results:
# flush_future = milvus.flush([collection_name], _async=True)
# status = flush_future.result()
flush_future.done()
def _compact_callback(status):
if status.OK():
print("Compact successfully")
else:
print("Compact failed.", status.message)
compact_furure = milvus.compact(collection_name,
_async=True,
_cakkback=_compact_callback)
# Or invoke result() to get results:
# compact_future = milvus.compact(collection_name, _async=True)
# status = compact_future.result()
compact_furure.done()
# Get demo_collection row count
status, result = milvus.count_entities(collection_name)
# present collection info
_, info = milvus.get_collection_stats(collection_name)
print(info)
# create index of vectors, search more rapidly
index_param = {'nlist': 2048}
def _index_callback(status):
if status.OK():
print("Create index successfully")
else:
print("Create index failed.", status.message)
# Create ivflat index in demo_collection
# You can search vectors without creating index. however, Creating index help to
# search faster
print("Creating index: {}".format(index_param))
index_future = milvus.create_index(collection_name,
IndexType.IVF_FLAT,
index_param,
_async=True,
_callback=_index_callback)
# Or invoke result() to get results:
# index_future = milvus.create_index(collection_name, IndexType.IVF_FLAT, index_param, _async=True)
# status = index_future.result()
index_future.done()
# describe index, get information of index
status, index = milvus.get_index_info(collection_name)
print(index)
# Use the top 10 vectors for similarity search
query_vectors = vectors[0:10]
# execute vector similarity search
search_param = {"nprobe": 16}
print("Searching ... ")
def _search_callback(status, results):
# if status.OK():
# print("Search successfully")
# else:
# print("Search failed.", status.message)
if status.OK():
# indicate search result
# also use by:
# `results.distance_array[0][0] == 0.0 or results.id_array[0][0] == ids[0]`
if results[0][0].distance == 0.0: # or results[0][0].id == ids[0]:
print('Query result is correct')
else:
print('Query result isn\'t correct')
# print results
print(results)
else:
print("Search failed. ", status)
param = {
'collection_name': collection_name,
'query_records': query_vectors,
'top_k': 1,
'params': search_param,
"_async": True,
"_callback": _search_callback
}
search_future = milvus.search(**param)
# Or invoke result() to get results:
#
# param = {
# 'collection_name': collection_name,
# 'query_records': query_vectors,
# 'top_k': 1,
# 'params': search_param,
# "_async": True,
# }
# search_future = milvus.search(param)
# status, results = index_future.result()
search_future.done()
# Delete demo_collection
status = milvus.drop_collection(collection_name)
class MilvusDocumentStore(SQLDocumentStore):
"""
Milvus (https://milvus.io/) is a highly reliable, scalable Document Store specialized on storing and processing vectors.
Therefore, it is particularly suited for Haystack users that work with dense retrieval methods (like DPR).
In contrast to FAISS, Milvus ...
- runs as a separate service (e.g. a Docker container) and can scale easily in a distributed environment
- allows dynamic data management (i.e. you can insert/delete vectors without recreating the whole index)
- encapsulates multiple ANN libraries (FAISS, ANNOY ...)
This class uses Milvus for all vector related storage, processing and querying.
The meta-data (e.g. for filtering) and the document text are however stored in a separate SQL Database as Milvus
does not allow these data types (yet).
Usage:
1. Start a Milvus server (see https://milvus.io/docs/v1.0.0/install_milvus.md)
2. Init a MilvusDocumentStore in Haystack
"""
def __init__(
self,
sql_url: str = "sqlite:///",
milvus_url: str = "tcp://localhost:19530",
connection_pool: str = "SingletonThread",
index: str = "document",
vector_dim: int = 768,
index_file_size: int = 1024,
similarity: str = "dot_product",
index_type: IndexType = IndexType.FLAT,
index_param: Optional[Dict[str, Any]] = None,
search_param: Optional[Dict[str, Any]] = None,
update_existing_documents: bool = False,
return_embedding: bool = False,
embedding_field: str = "embedding",
progress_bar: bool = True,
**kwargs,
):
"""
:param sql_url: SQL connection URL for storing document texts and metadata. It defaults to a local, file based SQLite DB. For large scale
deployment, Postgres is recommended. If using MySQL then same server can also be used for
Milvus metadata. For more details see https://milvus.io/docs/v1.0.0/data_manage.md.
:param milvus_url: Milvus server connection URL for storing and processing vectors.
Protocol, host and port will automatically be inferred from the URL.
See https://milvus.io/docs/v1.0.0/install_milvus.md for instructions to start a Milvus instance.
:param connection_pool: Connection pool type to connect with Milvus server. Default: "SingletonThread".
:param index: Index name for text, embedding and metadata (in Milvus terms, this is the "collection name").
:param vector_dim: The embedding vector size. Default: 768.
:param index_file_size: Specifies the size of each segment file that is stored by Milvus and its default value is 1024 MB.
When the size of newly inserted vectors reaches the specified volume, Milvus packs these vectors into a new segment.
Milvus creates one index file for each segment. When conducting a vector search, Milvus searches all index files one by one.
As a rule of thumb, we would see a 30% ~ 50% increase in the search performance after changing the value of index_file_size from 1024 to 2048.
Note that an overly large index_file_size value may cause failure to load a segment into the memory or graphics memory.
(From https://milvus.io/docs/v1.0.0/performance_faq.md#How-can-I-get-the-best-performance-from-Milvus-through-setting-index_file_size)
:param similarity: The similarity function used to compare document vectors. 'dot_product' is the default and recommended for DPR embeddings.
'cosine' is recommended for Sentence Transformers, but is not directly supported by Milvus.
However, you can normalize your embeddings and use `dot_product` to get the same results.
See https://milvus.io/docs/v1.0.0/metric.md?Inner-product-(IP)#floating.
:param index_type: Type of approximate nearest neighbour (ANN) index used. The choice here determines your tradeoff between speed and accuracy.
Some popular options:
- FLAT (default): Exact method, slow
- IVF_FLAT, inverted file based heuristic, fast
- HSNW: Graph based, fast
- ANNOY: Tree based, fast
See: https://milvus.io/docs/v1.0.0/index.md
:param index_param: Configuration parameters for the chose index_type needed at indexing time.
For example: {"nlist": 16384} as the number of cluster units to create for index_type IVF_FLAT.
See https://milvus.io/docs/v1.0.0/index.md
:param search_param: Configuration parameters for the chose index_type needed at query time
For example: {"nprobe": 10} as the number of cluster units to query for index_type IVF_FLAT.
See https://milvus.io/docs/v1.0.0/index.md
:param update_existing_documents: Whether to update any existing documents with the same ID when adding
documents. When set as True, any document with an existing ID gets updated.
If set to False, an error is raised if the document ID of the document being
added already exists.
:param return_embedding: To return document embedding.
:param embedding_field: Name of field containing an embedding vector.
:param progress_bar: Whether to show a tqdm progress bar or not.
Can be helpful to disable in production deployments to keep the logs clean.
"""
self.milvus_server = Milvus(uri=milvus_url, pool=connection_pool)
self.vector_dim = vector_dim
self.index_file_size = index_file_size
if similarity == "dot_product":
self.metric_type = MetricType.IP
self.similarity = similarity
else:
raise ValueError(
"The Milvus document store can currently only support dot_product similarity. "
"Please set similarity=\"dot_product\"")
self.index_type = index_type
self.index_param = index_param or {"nlist": 16384}
self.search_param = search_param or {"nprobe": 10}
self.index = index
self._create_collection_and_index_if_not_exist(self.index)
self.return_embedding = return_embedding
self.embedding_field = embedding_field
self.progress_bar = progress_bar
super().__init__(url=sql_url,
update_existing_documents=update_existing_documents,
index=index)
def __del__(self):
return self.milvus_server.close()
def _create_collection_and_index_if_not_exist(
self,
index: Optional[str] = None,
index_param: Optional[Dict[str, Any]] = None):
index = index or self.index
index_param = index_param or self.index_param
status, ok = self.milvus_server.has_collection(collection_name=index)
if not ok:
collection_param = {
'collection_name': index,
'dimension': self.vector_dim,
'index_file_size': self.index_file_size,
'metric_type': self.metric_type
}
status = self.milvus_server.create_collection(collection_param)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Collection creation on Milvus server failed: {status}')
status = self.milvus_server.create_index(index, self.index_type,
index_param)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Index creation on Milvus server failed: {status}')
def _create_document_field_map(self) -> Dict:
return {
self.index: self.embedding_field,
}
def write_documents(self,
documents: Union[List[dict], List[Document]],
index: Optional[str] = None,
batch_size: int = 10_000):
"""
Add new documents to the DocumentStore.
:param documents: List of `Dicts` or List of `Documents`. If they already contain the embeddings, we'll index
them right away in Milvus. If not, you can later call update_embeddings() to create & index them.
:param index: (SQL) index name for storing the docs and metadata
:param batch_size: When working with large number of documents, batching can help reduce memory footprint.
:return:
"""
index = index or self.index
self._create_collection_and_index_if_not_exist(index)
field_map = self._create_document_field_map()
if len(documents) == 0:
logger.warning(
"Calling DocumentStore.write_documents() with empty list")
return
document_objects = [
Document.from_dict(d, field_map=field_map)
if isinstance(d, dict) else d for d in documents
]
add_vectors = False if document_objects[0].embedding is None else True
batched_documents = get_batches_from_generator(document_objects,
batch_size)
with tqdm(total=len(document_objects),
disable=not self.progress_bar) as progress_bar:
for document_batch in batched_documents:
vector_ids = []
if add_vectors:
doc_ids = []
embeddings = []
for doc in document_batch:
doc_ids.append(doc.id)
if isinstance(doc.embedding, np.ndarray):
embeddings.append(doc.embedding.tolist())
elif isinstance(doc.embedding, list):
embeddings.append(doc.embedding)
else:
raise AttributeError(
f'Format of supplied document embedding {type(doc.embedding)} is not '
f'supported. Please use list or numpy.ndarray')
if self.update_existing_documents:
existing_docs = super().get_documents_by_id(
ids=doc_ids, index=index)
self._delete_vector_ids_from_milvus(
documents=existing_docs, index=index)
status, vector_ids = self.milvus_server.insert(
collection_name=index, records=embeddings)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Vector embedding insertion failed: {status}')
docs_to_write_in_sql = []
for idx, doc in enumerate(document_batch):
meta = doc.meta
if add_vectors:
meta["vector_id"] = vector_ids[idx]
docs_to_write_in_sql.append(doc)
super().write_documents(docs_to_write_in_sql, index=index)
progress_bar.update(batch_size)
progress_bar.close()
self.milvus_server.flush([index])
if self.update_existing_documents:
self.milvus_server.compact(collection_name=index)
def update_embeddings(
self,
retriever: BaseRetriever,
index: Optional[str] = None,
batch_size: int = 10_000,
update_existing_embeddings: bool = True,
filters: Optional[Dict[str, List[str]]] = None,
):
def main():
milvus = Milvus(_HOST, _PORT)
# num = random.randint(1, 100000)
num = 100000
# Create collection demo_collection if it dosen't exist.
collection_name = 'example_hybrid_collections_{}'.format(num)
if milvus.has_collection(collection_name):
milvus.drop_collection(collection_name)
collection_param = {
"fields": [{
"field": "A",
"type": DataType.INT32
}, {
"field": "B",
"type": DataType.INT32
}, {
"field": "C",
"type": DataType.INT64
}, {
"field": "Vec",
"type": DataType.FLOAT_VECTOR,
"params": {
"dim": 128,
"metric_type": "L2"
}
}],
"segment_size":
100
}
milvus.create_collection(collection_name, collection_param)
milvus.compact(collection_name)
# milvus.create_partition(collection_name, "p_01", timeout=1800)
# pars = milvus.list_partitions(collection_name)
# ok = milvus.has_partition(collection_name, "p_01", timeout=1800)
# assert ok
# ok = milvus.has_partition(collection_name, "p_02")
# assert not ok
# for p in pars:
# if p == "_default":
# continue
# milvus.drop_partition(collection_name, p)
# milvus.drop_collection(collection_name)
# sys.exit(0)
A_list = [random.randint(0, 255) for _ in range(num)]
vec = [[random.random() for _ in range(128)] for _ in range(num)]
hybrid_entities = [{
"field": "A",
"values": A_list,
"type": DataType.INT32
}, {
"field": "B",
"values": A_list,
"type": DataType.INT32
}, {
"field": "C",
"values": A_list,
"type": DataType.INT64
}, {
"field": "Vec",
"values": vec,
"type": DataType.FLOAT_VECTOR,
"params": {
"dim": 128
}
}]
for slice_e in utils.entities_slice(hybrid_entities):
ids = milvus.insert(collection_name, slice_e)
milvus.flush([collection_name])
print("Flush ... ")
# time.sleep(3)
count = milvus.count_entities(collection_name)
milvus.delete_entity_by_id(collection_name, ids[:1])
milvus.flush([collection_name])
print("Get entity be id start ...... ")
entities = milvus.get_entity_by_id(collection_name, ids[:1])
et = entities.dict()
milvus.delete_entity_by_id(collection_name, ids[1:2])
milvus.flush([collection_name])
print("Create index ......")
milvus.create_index(collection_name, "Vec", {
"index_type": "IVF_FLAT",
"metric_type": "L2",
"params": {
"nlist": 100
}
})
print("Create index done.")
info = milvus.get_collection_info(collection_name)
print(info)
stats = milvus.get_collection_stats(collection_name)
print("\nstats\n")
print(stats)
query_hybrid = \
{
"bool": {
"must": [
{
"term": {
"A": [1, 2, 5]
}
},
{
"range": {
"B": {"GT": 1, "LT": 100}
}
},
{
"vector": {
"Vec": {
"topk": 10, "query": vec[: 10000], "params": {"nprobe": 10}
}
}
}
],
},
}
# print("Start searach ..", flush=True)
# results = milvus.search(collection_name, query_hybrid)
# print(results)
#
# for r in list(results):
# print("ids", r.ids)
# print("distances", r.distances)
t0 = time.time()
count = 0
results = milvus.search(collection_name, query_hybrid, fields=["B"])
for r in list(results):
# print("ids", r.ids)
# print("distances", r.distances)
for rr in r:
count += 1
# print(rr.entity.get("B"))
print("Search cost {} s".format(time.time() - t0))
# for result in results:
# for r in result:
# print(f"{r}")
# itertor entity id
# for result in results:
# for r in result:
# # get distance
# dis = r.distance
# id_ = r.id
# # obtain all field name
# fields = r.entity.fields
# for f in fields:
# # get field value by field name
# # fv = r.entity.
# fv = r.entity.value_of_field(f)
# print(fv)
milvus.drop_collection(collection_name)
class MilvusClient(object):
def __init__(self, collection_name=None, ip=None, port=None, timeout=60):
self._collection_name = collection_name
try:
i = 1
start_time = time.time()
if not ip:
self._milvus = Milvus(host=SERVER_HOST_DEFAULT,
port=SERVER_PORT_DEFAULT)
else:
# retry connect for remote server
while time.time() < start_time + timeout:
try:
self._milvus = Milvus(host=ip, port=port)
if self._milvus.server_status():
logger.debug(
"Try connect times: %d, %s" %
(i, round(time.time() - start_time, 2)))
break
except Exception as e:
logger.debug("Milvus connect failed")
i = i + 1
except Exception as e:
raise e
def __str__(self):
return 'Milvus collection %s' % self._collection_name
def check_status(self, status):
if not status.OK():
logger.error(status.message)
raise Exception("Status not ok")
def check_result_ids(self, result):
for index, item in enumerate(result):
if item[0].distance >= epsilon:
logger.error(index)
logger.error(item[0].distance)
raise Exception("Distance wrong")
def create_collection(self, collection_name, dimension, index_file_size,
metric_type):
if not self._collection_name:
self._collection_name = collection_name
if metric_type == "l2":
metric_type = MetricType.L2
elif metric_type == "ip":
metric_type = MetricType.IP
elif metric_type == "jaccard":
metric_type = MetricType.JACCARD
elif metric_type == "hamming":
metric_type = MetricType.HAMMING
elif metric_type == "sub":
metric_type = MetricType.SUBSTRUCTURE
elif metric_type == "super":
metric_type = MetricType.SUPERSTRUCTURE
else:
logger.error("Not supported metric_type: %s" % metric_type)
create_param = {
'collection_name': collection_name,
'dimension': dimension,
'index_file_size': index_file_size,
"metric_type": metric_type
}
status = self._milvus.create_collection(create_param)
self.check_status(status)
@time_wrapper
def insert(self, X, ids=None):
status, result = self._milvus.add_vectors(self._collection_name, X,
ids)
self.check_status(status)
return status, result
@time_wrapper
def delete_vectors(self, ids):
status = self._milvus.delete_by_id(self._collection_name, ids)
self.check_status(status)
@time_wrapper
def flush(self):
status = self._milvus.flush([self._collection_name])
self.check_status(status)
@time_wrapper
def compact(self):
status = self._milvus.compact(self._collection_name)
self.check_status(status)
@time_wrapper
def create_index(self, index_type, index_param=None):
index_type = INDEX_MAP[index_type]
logger.info(
"Building index start, collection_name: %s, index_type: %s" %
(self._collection_name, index_type))
if index_param:
logger.info(index_param)
status = self._milvus.create_index(self._collection_name, index_type,
index_param)
self.check_status(status)
def describe_index(self):
status, result = self._milvus.describe_index(self._collection_name)
self.check_status(status)
index_type = None
for k, v in INDEX_MAP.items():
if result._index_type == v:
index_type = k
break
return {"index_type": index_type, "index_param": result._params}
def drop_index(self):
logger.info("Drop index: %s" % self._collection_name)
return self._milvus.drop_index(self._collection_name)
@time_wrapper
def query(self, X, top_k, search_param=None):
status, result = self._milvus.search_vectors(self._collection_name,
top_k,
query_records=X,
params=search_param)
self.check_status(status)
return result
@time_wrapper
def query_ids(self, top_k, ids, search_param=None):
status, result = self._milvus.search_by_ids(self._collection_name,
ids,
top_k,
params=search_param)
self.check_result_ids(result)
return result
def count(self):
return self._milvus.count_collection(self._collection_name)[1]
def delete(self, timeout=120):
timeout = int(timeout)
logger.info("Start delete collection: %s" % self._collection_name)
self._milvus.drop_collection(self._collection_name)
i = 0
while i < timeout:
if self.count():
time.sleep(1)
i = i + 1
continue
else:
break
if i >= timeout:
logger.error("Delete collection timeout")
def describe(self):
return self._milvus.describe_collection(self._collection_name)
def show_collections(self):
return self._milvus.show_collections()
def exists_collection(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status, res = self._milvus.has_collection(collection_name)
# self.check_status(status)
return res
@time_wrapper
def preload_collection(self):
status = self._milvus.preload_collection(self._collection_name,
timeout=3000)
self.check_status(status)
return status
def get_server_version(self):
status, res = self._milvus.server_version()
return res
def get_server_mode(self):
return self.cmd("mode")
def get_server_commit(self):
return self.cmd("build_commit_id")
def get_server_config(self):
return json.loads(self.cmd("get_config *"))
def get_mem_info(self):
result = json.loads(self.cmd("get_system_info"))
result_human = {
# unit: Gb
"memory_used":
round(int(result["memory_used"]) / (1024 * 1024 * 1024), 2)
}
return result_human
def cmd(self, command):
status, res = self._milvus._cmd(command)
logger.info("Server command: %s, result: %s" % (command, res))
self.check_status(status)
return res
class Indexer:
'''
索引器。
'''
def __init__(self, name, host='127.0.0.1', port='19531'):
'''
初始化。
'''
self.client = Milvus(host=host, port=port)
self.collection = name
def init(self, lenient=False):
'''
创建集合。
'''
if lenient:
status, result = self.client.has_collection(
collection_name=self.collection)
if status.code != 0:
raise ExertMilvusException(status)
if result:
return
status = self.client.create_collection({
'collection_name': self.collection,
'dimension': 512,
'index_file_size': 1024,
'metric_type': MetricType.L2
})
if status.code != 0 and not (lenient and status.code == 9):
raise ExertMilvusException(status)
# 创建索引。
status = self.client.create_index(collection_name=self.collection,
index_type=IndexType.IVF_FLAT,
params={'nlist': 16384})
if status.code != 0:
raise ExertMilvusException(status)
return status
def drop(self):
'''
删除集合。
'''
status = self.client.drop_collection(collection_name=self.collection)
if status.code != 0:
raise ExertMilvusException(status)
def flush(self):
'''
写入到硬盘。
'''
status = self.client.flush([self.collection])
if status.code != 0:
raise ExertMilvusException(status)
def compact(self):
'''
压缩集合。
'''
status = self.client.compact(collection_name=self.collection)
if status.code != 0:
raise ExertMilvusException(status)
def close(self):
'''
关闭链接。
'''
self.client.close()
def new_tag(self, tag):
'''
建分块标签。
'''
status = self.client.create_partition(collection_name=self.collection,
partition_tag=tag)
if status.code != 0:
raise ExertMilvusException(status)
def list_tag(self):
'''
列举分块标签。
'''
status, result = self.client.list_partitions(
collection_name=self.collection)
if status.code != 0:
raise ExertMilvusException(status)
return result
def drop_tag(self, tag):
'''
删除分块标签。
'''
status = self.client.drop_partition(collection_name=self.collection,
partition_tag=tag)
if status.code != 0:
raise ExertMilvusException(status)
def index(self, vectors, tag=None, ids=None):
'''
添加索引
'''
params = {}
if tag != None:
params['tag'] = tag
if ids != None:
params['ids'] = ids
status, result = self.client.insert(collection_name=self.collection,
records=vectors,
**params)
if status.code != 0:
raise ExertMilvusException(status)
return result
def listing(self, ids):
'''
列举信息。
'''
status, result = self.client.get_entity_by_id(
collection_name=self.collection, ids=ids)
if status.code != 0:
raise ExertMilvusException(status)
return result
def counting(self):
'''
计算索引数。
'''
status, result = self.client.count_entities(
collection_name=self.collection)
if status.code != 0:
raise ExertMilvusException(status)
return result
def unindex(self, ids):
'''
去掉索引。
'''
status = self.client.delete_entity_by_id(
collection_name=self.collection, id_array=ids)
if status.code != 0:
raise ExertMilvusException(status)
def search(self, vectors, top_count=100, tags=None):
'''
搜索。
'''
params = {'params': {'nprobe': 16}}
if tags != None:
params['partition_tags'] = tags
status, results = self.client.search(collection_name=self.collection,
query_records=vectors,
top_k=top_count,
**params)
if status.code != 0:
raise ExertMilvusException(status)
return results
class MilvusClient(object):
def __init__(self, collection_name=None, host=None, port=None, timeout=60):
"""
Milvus client wrapper for python-sdk.
Default timeout set 60s
"""
self._collection_name = collection_name
try:
start_time = time.time()
if not host:
host = SERVER_HOST_DEFAULT
if not port:
port = SERVER_PORT_DEFAULT
logger.debug(host)
logger.debug(port)
# retry connect for remote server
i = 0
while time.time() < start_time + timeout:
try:
self._milvus = Milvus(host=host, port=port, try_connect=False, pre_ping=False)
if self._milvus.server_status():
logger.debug("Try connect times: %d, %s" % (i, round(time.time() - start_time, 2)))
break
except Exception as e:
logger.debug("Milvus connect failed: %d times" % i)
i = i + 1
if time.time() > start_time + timeout:
raise Exception("Server connect timeout")
except Exception as e:
raise e
self._metric_type = None
if self._collection_name and self.exists_collection():
self._metric_type = metric_type_to_str(self.describe()[1].metric_type)
self._dimension = self.describe()[1].dimension
def __str__(self):
return 'Milvus collection %s' % self._collection_name
def set_collection(self, name):
self._collection_name = name
def check_status(self, status):
if not status.OK():
logger.error(self._collection_name)
logger.error(status.message)
logger.error(self._milvus.server_status())
logger.error(self.count())
raise Exception("Status not ok")
def check_result_ids(self, result):
for index, item in enumerate(result):
if item[0].distance >= epsilon:
logger.error(index)
logger.error(item[0].distance)
raise Exception("Distance wrong")
def create_collection(self, collection_name, dimension, index_file_size, metric_type):
if not self._collection_name:
self._collection_name = collection_name
if metric_type not in METRIC_MAP.keys():
raise Exception("Not supported metric_type: %s" % metric_type)
metric_type = METRIC_MAP[metric_type]
create_param = {'collection_name': collection_name,
'dimension': dimension,
'index_file_size': index_file_size,
"metric_type": metric_type}
status = self._milvus.create_collection(create_param)
self.check_status(status)
def create_partition(self, tag_name):
status = self._milvus.create_partition(self._collection_name, tag_name)
self.check_status(status)
def drop_partition(self, tag_name):
status = self._milvus.drop_partition(self._collection_name, tag_name)
self.check_status(status)
def list_partitions(self):
status, tags = self._milvus.list_partitions(self._collection_name)
self.check_status(status)
return tags
@time_wrapper
def insert(self, X, ids=None, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status, result = self._milvus.insert(collection_name, X, ids)
self.check_status(status)
return status, result
def insert_rand(self):
insert_xb = random.randint(1, 100)
X = [[random.random() for _ in range(self._dimension)] for _ in range(insert_xb)]
X = utils.normalize(self._metric_type, X)
count_before = self.count()
status, _ = self.insert(X)
self.check_status(status)
self.flush()
if count_before + insert_xb != self.count():
raise Exception("Assert failed after inserting")
def get_rand_ids(self, length):
while True:
status, stats = self._milvus.get_collection_stats(self._collection_name)
self.check_status(status)
segments = stats["partitions"][0]["segments"]
# random choice one segment
segment = random.choice(segments)
status, segment_ids = self._milvus.list_id_in_segment(self._collection_name, segment["name"])
if not status.OK():
logger.error(status.message)
continue
if len(segment_ids):
break
if length >= len(segment_ids):
logger.debug("Reset length: %d" % len(segment_ids))
return segment_ids
return random.sample(segment_ids, length)
def get_rand_ids_each_segment(self, length):
res = []
status, stats = self._milvus.get_collection_stats(self._collection_name)
self.check_status(status)
segments = stats["partitions"][0]["segments"]
segments_num = len(segments)
# random choice from each segment
for segment in segments:
status, segment_ids = self._milvus.list_id_in_segment(self._collection_name, segment["name"])
self.check_status(status)
res.extend(segment_ids[:length])
return segments_num, res
def get_rand_entities(self, length):
ids = self.get_rand_ids(length)
status, get_res = self._milvus.get_entity_by_id(self._collection_name, ids)
self.check_status(status)
return ids, get_res
@time_wrapper
def get_entities(self, get_ids):
status, get_res = self._milvus.get_entity_by_id(self._collection_name, get_ids)
self.check_status(status)
return get_res
@time_wrapper
def delete(self, ids, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status = self._milvus.delete_entity_by_id(collection_name, ids)
self.check_status(status)
def delete_rand(self):
delete_id_length = random.randint(1, 100)
count_before = self.count()
logger.info("%s: length to delete: %d" % (self._collection_name, delete_id_length))
delete_ids = self.get_rand_ids(delete_id_length)
self.delete(delete_ids)
self.flush()
logger.info("%s: count after delete: %d" % (self._collection_name, self.count()))
status, get_res = self._milvus.get_entity_by_id(self._collection_name, delete_ids)
self.check_status(status)
for item in get_res:
if item:
raise Exception("Assert failed after delete")
if count_before - len(delete_ids) != self.count():
raise Exception("Assert failed after delete")
@time_wrapper
def flush(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status = self._milvus.flush([collection_name])
self.check_status(status)
@time_wrapper
def compact(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status = self._milvus.compact(collection_name)
self.check_status(status)
@time_wrapper
def create_index(self, index_type, index_param=None):
index_type = INDEX_MAP[index_type]
logger.info("Building index start, collection_name: %s, index_type: %s" % (self._collection_name, index_type))
if index_param:
logger.info(index_param)
status = self._milvus.create_index(self._collection_name, index_type, index_param)
self.check_status(status)
def describe_index(self):
status, result = self._milvus.get_index_info(self._collection_name)
self.check_status(status)
index_type = None
for k, v in INDEX_MAP.items():
if result._index_type == v:
index_type = k
break
return {"index_type": index_type, "index_param": result._params}
def drop_index(self):
logger.info("Drop index: %s" % self._collection_name)
return self._milvus.drop_index(self._collection_name)
def query(self, X, top_k, search_param=None, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
status, result = self._milvus.search(collection_name, top_k, query_records=X, params=search_param)
self.check_status(status)
return result
def query_rand(self):
top_k = random.randint(1, 100)
nq = random.randint(1, 100)
nprobe = random.randint(1, 100)
search_param = {"nprobe": nprobe}
_, X = self.get_rand_entities(nq)
logger.info("%s, Search nq: %d, top_k: %d, nprobe: %d" % (self._collection_name, nq, top_k, nprobe))
status, _ = self._milvus.search(self._collection_name, top_k, query_records=X, params=search_param)
self.check_status(status)
# for i, item in enumerate(search_res):
# if item[0].id != ids[i]:
# logger.warning("The index of search result: %d" % i)
# raise Exception("Query failed")
# @time_wrapper
# def query_ids(self, top_k, ids, search_param=None):
# status, result = self._milvus.search_by_id(self._collection_name, ids, top_k, params=search_param)
# self.check_result_ids(result)
# return result
def count(self, name=None):
if name is None:
name = self._collection_name
logger.debug(self._milvus.count_entities(name))
row_count = self._milvus.count_entities(name)[1]
if not row_count:
row_count = 0
logger.debug("Row count: %d in collection: <%s>" % (row_count, name))
return row_count
def drop(self, timeout=120, name=None):
timeout = int(timeout)
if name is None:
name = self._collection_name
logger.info("Start delete collection: %s" % name)
status = self._milvus.drop_collection(name)
self.check_status(status)
i = 0
while i < timeout:
if self.count(name=name):
time.sleep(1)
i = i + 1
continue
else:
break
if i >= timeout:
logger.error("Delete collection timeout")
def describe(self):
# logger.info(self._milvus.get_collection_info(self._collection_name))
return self._milvus.get_collection_info(self._collection_name)
def show_collections(self):
return self._milvus.list_collections()
def exists_collection(self, collection_name=None):
if collection_name is None:
collection_name = self._collection_name
_, res = self._milvus.has_collection(collection_name)
# self.check_status(status)
return res
def clean_db(self):
collection_names = self.show_collections()[1]
for name in collection_names:
logger.debug(name)
self.drop(name=name)
@time_wrapper
def preload_collection(self):
status = self._milvus.load_collection(self._collection_name, timeout=3000)
self.check_status(status)
return status
def get_server_version(self):
_, res = self._milvus.server_version()
return res
def get_server_mode(self):
return self.cmd("mode")
def get_server_commit(self):
return self.cmd("build_commit_id")
def get_server_config(self):
return json.loads(self.cmd("get_config *"))
def get_mem_info(self):
result = json.loads(self.cmd("get_system_info"))
result_human = {
# unit: Gb
"memory_used": round(int(result["memory_used"]) / (1024*1024*1024), 2)
}
return result_human
def cmd(self, command):
status, res = self._milvus._cmd(command)
logger.info("Server command: %s, result: %s" % (command, res))
self.check_status(status)
return res
# ------
print("\n----------search----------")
search_future = client.search(collection_name, dsl, _async=True)
search_results = search_future.result()
# ------
# Basic delete:
# Now let's see how to delete things in Milvus.
# You can simply delete entities by their ids.
#
# After deleted, we invoke compact collection in a asynchronous way.
# ------
print("\n----------delete id = 1, id = 2----------")
client.delete_entity_by_id(collection_name, ids=[1, 4])
client.flush() # flush is important
compact_future = client.compact(collection_name, _async=True)
compact_future.result()
# ------
# Basic delete:
# You can drop partitions we create, and drop the collection we create.
# ------
client.drop_partition(collection_name, partition_tag='American')
if collection_name in client.list_collections():
client.drop_collection(collection_name)
# ------
# Summary:
# Now we've went through all basic communications pymilvus can do with Milvus server, hope it's helpful!
# ------
class MilvusDocumentStore(SQLDocumentStore):
"""
Milvus (https://milvus.io/) is a highly reliable, scalable Document Store specialized on storing and processing vectors.
Therefore, it is particularly suited for Haystack users that work with dense retrieval methods (like DPR).
In contrast to FAISS, Milvus ...
- runs as a separate service (e.g. a Docker container) and can scale easily in a distributed environment
- allows dynamic data management (i.e. you can insert/delete vectors without recreating the whole index)
- encapsulates multiple ANN libraries (FAISS, ANNOY ...)
This class uses Milvus for all vector related storage, processing and querying.
The meta-data (e.g. for filtering) and the document text are however stored in a separate SQL Database as Milvus
does not allow these data types (yet).
Usage:
1. Start a Milvus server (see https://milvus.io/docs/v0.10.5/install_milvus.md)
2. Init a MilvusDocumentStore in Haystack
"""
def __init__(
self,
sql_url: str = "sqlite:///",
milvus_url: str = "tcp://localhost:19530",
connection_pool: str = "SingletonThread",
index: str = "document",
vector_dim: int = 768,
index_file_size: int = 1024,
similarity: str = "dot_product",
index_type: IndexType = IndexType.FLAT,
index_param: Optional[Dict[str, Any]] = None,
search_param: Optional[Dict[str, Any]] = None,
update_existing_documents: bool = False,
return_embedding: bool = False,
embedding_field: str = "embedding",
**kwargs,
):
"""
:param sql_url: SQL connection URL for storing document texts and metadata. It defaults to a local, file based SQLite DB. For large scale
deployment, Postgres is recommended. If using MySQL then same server can also be used for
Milvus metadata. For more details see https://milvus.io/docs/v0.10.5/data_manage.md.
:param milvus_url: Milvus server connection URL for storing and processing vectors.
Protocol, host and port will automatically be inferred from the URL.
See https://milvus.io/docs/v0.10.5/install_milvus.md for instructions to start a Milvus instance.
:param connection_pool: Connection pool type to connect with Milvus server. Default: "SingletonThread".
:param index: Index name for text, embedding and metadata (in Milvus terms, this is the "collection name").
:param vector_dim: The embedding vector size. Default: 768.
:param index_file_size: Specifies the size of each segment file that is stored by Milvus and its default value is 1024 MB.
When the size of newly inserted vectors reaches the specified volume, Milvus packs these vectors into a new segment.
Milvus creates one index file for each segment. When conducting a vector search, Milvus searches all index files one by one.
As a rule of thumb, we would see a 30% ~ 50% increase in the search performance after changing the value of index_file_size from 1024 to 2048.
Note that an overly large index_file_size value may cause failure to load a segment into the memory or graphics memory.
(From https://milvus.io/docs/v0.10.5/performance_faq.md#How-can-I-get-the-best-performance-from-Milvus-through-setting-index_file_size)
:param similarity: The similarity function used to compare document vectors. 'dot_product' is the default and recommended for DPR embeddings.
'cosine' is recommended for Sentence Transformers, but is not directly supported by Milvus.
However, you can normalize your embeddings and use `dot_product` to get the same results.
See https://milvus.io/docs/v0.10.5/metric.md?Inner-product-(IP)#floating.
:param index_type: Type of approximate nearest neighbour (ANN) index used. The choice here determines your tradeoff between speed and accuracy.
Some popular options:
- FLAT (default): Exact method, slow
- IVF_FLAT, inverted file based heuristic, fast
- HSNW: Graph based, fast
- ANNOY: Tree based, fast
See: https://milvus.io/docs/v0.10.5/index.md
:param index_param: Configuration parameters for the chose index_type needed at indexing time.
For example: {"nlist": 16384} as the number of cluster units to create for index_type IVF_FLAT.
See https://milvus.io/docs/v0.10.5/index.md
:param search_param: Configuration parameters for the chose index_type needed at query time
For example: {"nprobe": 10} as the number of cluster units to query for index_type IVF_FLAT.
See https://milvus.io/docs/v0.10.5/index.md
:param update_existing_documents: Whether to update any existing documents with the same ID when adding
documents. When set as True, any document with an existing ID gets updated.
If set to False, an error is raised if the document ID of the document being
added already exists.
:param return_embedding: To return document embedding.
:param embedding_field: Name of field containing an embedding vector.
"""
self.milvus_server = Milvus(uri=milvus_url, pool=connection_pool)
self.vector_dim = vector_dim
self.index_file_size = index_file_size
if similarity == "dot_product":
self.metric_type = MetricType.L2
else:
raise ValueError(
"The Milvus document store can currently only support dot_product similarity. "
"Please set similarity=\"dot_product\"")
self.index_type = index_type
self.index_param = index_param or {"nlist": 16384}
self.search_param = search_param or {"nprobe": 10}
self.index = index
self._create_collection_and_index_if_not_exist(self.index)
self.return_embedding = return_embedding
self.embedding_field = embedding_field
super().__init__(url=sql_url,
update_existing_documents=update_existing_documents,
index=index)
def __del__(self):
return self.milvus_server.close()
def _create_collection_and_index_if_not_exist(
self,
index: Optional[str] = None,
index_param: Optional[Dict[str, Any]] = None):
index = index or self.index
index_param = index_param or self.index_param
status, ok = self.milvus_server.has_collection(collection_name=index)
if not ok:
collection_param = {
'collection_name': index,
'dimension': self.vector_dim,
'index_file_size': self.index_file_size,
'metric_type': self.metric_type
}
status = self.milvus_server.create_collection(collection_param)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Collection creation on Milvus server failed: {status}')
status = self.milvus_server.create_index(index, self.index_type,
index_param)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Index creation on Milvus server failed: {status}')
def _create_document_field_map(self) -> Dict:
return {
self.index: self.embedding_field,
}
def write_documents(self,
documents: Union[List[dict], List[Document]],
index: Optional[str] = None,
batch_size: int = 10_000):
"""
Add new documents to the DocumentStore.
:param documents: List of `Dicts` or List of `Documents`. If they already contain the embeddings, we'll index
them right away in Milvus. If not, you can later call update_embeddings() to create & index them.
:param index: (SQL) index name for storing the docs and metadata
:param batch_size: When working with large number of documents, batching can help reduce memory footprint.
:return:
"""
index = index or self.index
self._create_collection_and_index_if_not_exist(index)
field_map = self._create_document_field_map()
if len(documents) == 0:
logger.warning(
"Calling DocumentStore.write_documents() with empty list")
return
document_objects = [
Document.from_dict(d, field_map=field_map)
if isinstance(d, dict) else d for d in documents
]
add_vectors = False if document_objects[0].embedding is None else True
batched_documents = get_batches_from_generator(document_objects,
batch_size)
with tqdm(total=len(document_objects)) as progress_bar:
for document_batch in batched_documents:
vector_ids = []
if add_vectors:
doc_ids = []
embeddings = []
for doc in document_batch:
doc_ids.append(doc.id)
if isinstance(doc.embedding, np.ndarray):
embeddings.append(doc.embedding.tolist())
elif isinstance(doc.embedding, list):
embeddings.append(doc.embedding)
else:
raise AttributeError(
f'Format of supplied document embedding {type(doc.embedding)} is not '
f'supported. Please use list or numpy.ndarray')
if self.update_existing_documents:
existing_docs = super().get_documents_by_id(
ids=doc_ids, index=index)
self._delete_vector_ids_from_milvus(
documents=existing_docs, index=index)
status, vector_ids = self.milvus_server.insert(
collection_name=index, records=embeddings)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Vector embedding insertion failed: {status}')
docs_to_write_in_sql = []
for idx, doc in enumerate(document_batch):
meta = doc.meta
if add_vectors:
meta["vector_id"] = vector_ids[idx]
docs_to_write_in_sql.append(doc)
super().write_documents(docs_to_write_in_sql, index=index)
progress_bar.update(batch_size)
progress_bar.close()
self.milvus_server.flush([index])
if self.update_existing_documents:
self.milvus_server.compact(collection_name=index)
def update_embeddings(self,
retriever: BaseRetriever,
index: Optional[str] = None,
batch_size: int = 10_000):
"""
Updates the embeddings in the the document store using the encoding model specified in the retriever.
This can be useful if want to add or change the embeddings for your documents (e.g. after changing the retriever config).
:param retriever: Retriever to use to get embeddings for text
:param index: (SQL) index name for storing the docs and metadata
:param batch_size: When working with large number of documents, batching can help reduce memory footprint.
:return: None
"""
index = index or self.index
self._create_collection_and_index_if_not_exist(index)
document_count = self.get_document_count(index=index)
if document_count == 0:
logger.warning(
"Calling DocumentStore.update_embeddings() on an empty index")
return
logger.info(f"Updating embeddings for {document_count} docs...")
result = self.get_all_documents_generator(index=index,
batch_size=batch_size,
return_embedding=False)
batched_documents = get_batches_from_generator(result, batch_size)
with tqdm(total=document_count) as progress_bar:
for document_batch in batched_documents:
self._delete_vector_ids_from_milvus(documents=document_batch,
index=index)
embeddings = retriever.embed_passages(
document_batch) # type: ignore
embeddings_list = [
embedding.tolist() for embedding in embeddings
]
assert len(document_batch) == len(embeddings_list)
status, vector_ids = self.milvus_server.insert(
collection_name=index, records=embeddings_list)
if status.code != Status.SUCCESS:
raise RuntimeError(
f'Vector embedding insertion failed: {status}')
vector_id_map = {}
for vector_id, doc in zip(vector_ids, document_batch):
vector_id_map[doc.id] = vector_id
self.update_vector_ids(vector_id_map, index=index)
progress_bar.update(batch_size)
progress_bar.close()
self.milvus_server.flush([index])
self.milvus_server.compact(collection_name=index)
def query_by_embedding(
self,
query_emb: np.array,
filters: Optional[dict] = None,
top_k: int = 10,
index: Optional[str] = None,
return_embedding: Optional[bool] = None) -> List[Document]:
"""
Find the document that is most similar to the provided `query_emb` by using a vector similarity metric.
:param query_emb: Embedding of the query (e.g. gathered from DPR)
:param filters: Optional filters to narrow down the search space.
Example: {"name": ["some", "more"], "category": ["only_one"]}
:param top_k: How many documents to return
:param index: (SQL) index name for storing the docs and metadata
:param return_embedding: To return document embedding
:return:
"""
if filters:
raise Exception(
"Query filters are not implemented for the MilvusDocumentStore."
)
index = index or self.index
status, ok = self.milvus_server.has_collection(collection_name=index)
if status.code != Status.SUCCESS:
raise RuntimeError(f'Milvus has collection check failed: {status}')
if not ok:
raise Exception(
"No index exists. Use 'update_embeddings()` to create an index."
)
if return_embedding is None:
return_embedding = self.return_embedding
index = index or self.index
query_emb = query_emb.reshape(1, -1).astype(np.float32)
status, search_result = self.milvus_server.search(
collection_name=index,
query_records=query_emb,
top_k=top_k,
params=self.search_param)
if status.code != Status.SUCCESS:
raise RuntimeError(f'Vector embedding search failed: {status}')
vector_ids_for_query = []
scores_for_vector_ids: Dict[str, float] = {}
for vector_id_list, distance_list in zip(search_result.id_array,
search_result.distance_array):
for vector_id, distance in zip(vector_id_list, distance_list):
vector_ids_for_query.append(str(vector_id))
scores_for_vector_ids[str(vector_id)] = distance
documents = self.get_documents_by_vector_ids(vector_ids_for_query,
index=index)
if return_embedding:
self._populate_embeddings_to_docs(index=index, docs=documents)
for doc in documents:
doc.score = scores_for_vector_ids[doc.meta["vector_id"]]
doc.probability = float(expit(np.asarray(doc.score / 100)))
return documents
def delete_all_documents(self,
index: Optional[str] = None,
filters: Optional[Dict[str, List[str]]] = None):
"""
Delete all documents (from SQL AND Milvus).
:param index: (SQL) index name for storing the docs and metadata
:param filters: Optional filters to narrow down the search space.
Example: {"name": ["some", "more"], "category": ["only_one"]}
:return: None
"""
index = index or self.index
super().delete_all_documents(index=index, filters=filters)
status, ok = self.milvus_server.has_collection(collection_name=index)
if status.code != Status.SUCCESS:
raise RuntimeError(f'Milvus has collection check failed: {status}')
if ok:
status = self.milvus_server.drop_collection(collection_name=index)
if status.code != Status.SUCCESS:
raise RuntimeError(f'Milvus drop collection failed: {status}')
self.milvus_server.flush([index])
self.milvus_server.compact(collection_name=index)
def get_all_documents_generator(
self,
index: Optional[str] = None,
filters: Optional[Dict[str, List[str]]] = None,
return_embedding: Optional[bool] = None,
batch_size: int = 10_000,
) -> Generator[Document, None, None]:
"""
Get all documents from the document store. Under-the-hood, documents are fetched in batches from the
document store and yielded as individual documents. This method can be used to iteratively process
a large number of documents without having to load all documents in memory.
:param index: Name of the index to get the documents from. If None, the
DocumentStore's default index (self.index) will be used.
:param filters: Optional filters to narrow down the documents to return.
Example: {"name": ["some", "more"], "category": ["only_one"]}
:param return_embedding: Whether to return the document embeddings.
:param batch_size: When working with large number of documents, batching can help reduce memory footprint.
"""
index = index or self.index
documents = super().get_all_documents_generator(index=index,
filters=filters,
batch_size=batch_size)
if return_embedding is None:
return_embedding = self.return_embedding
for doc in documents:
if return_embedding:
self._populate_embeddings_to_docs(index=index, docs=[doc])
yield doc