Error in embeddings pipe

[gopig99]

Hi @ALL,

I am trying to run this code
%%time
from towhee import pipe, ops
import numpy as np
from towhee.datacollection import DataCollection

insert_pipe = (
pipe.input('id', 'question', 'answer')
.map('question', 'vec', ops.text_embedding.dpr(model_name='facebook/dpr-ctx_encoder-single-nq-base'))
.map('vec', 'vec', lambda x: x / np.linalg.norm(x, axis=0))
.map(('id', 'vec'), 'insert_status', ops.ann_insert.milvus_client(host='127.0.0.1', port='19530', collection_name='question_answer'))
.output()
)

import csv
with open('question_answer.csv', encoding='utf-8') as f:
reader = csv.reader(f)
next(reader)
for row in reader:
insert_pipe(*row)

And I am getting the following error
RuntimeError Traceback (most recent call last)
File :10

File ~\anaconda3\lib\site-packages\towhee\runtime\pipeline.py:116, in Pipeline.output(self, *output_schema, **config_kws)
113 dag_dict[self._clo_node]['next_nodes'].append(uid)
115 run_pipe = RuntimePipeline(dag_dict, config=config_kws)
--> 116 run_pipe.preload()
117 return run_pipe

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:140, in RuntimePipeline.preload(self)
136 def preload(self):
137 """
138 Preload the operators.
139 """
--> 140 return _Graph(self._dag_repr.nodes, self._dag_repr.edges, self._operator_pool, self._thread_pool)

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:64, in _Graph.init(self, nodes, edges, operator_pool, thread_pool, enable_trance)
62 self.features = None
63 self.time_profiler.record(Event.pipe_name, Event.pipe_in)
---> 64 self.initialize()
65 self._input_queue = self._data_queues[0]

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:75, in _Graph.initialize(self)
73 node = create_node(self._nodes[name], self._operator_pool, in_queues, out_queues, self._time_profiler)
74 if not node.initialize():
---> 75 raise RuntimeError(node.err_msg)
76 self._node_runners.append(node)

RuntimeError: Node-text-embedding/dpr-0 runs failed, error msg: Create text-embedding/dpr-0 operator text-embedding/dpr:main with args None and kws {'model_name': 'facebook/dpr-ctx_encoder-single-nq-base'} failed, err: Load operator failed, Traceback (most recent call last):
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\nodes\node.py", line 88, in initialize
self._op = self._op_pool.acquire_op(
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\operator_manager\operator_pool.py", line 99, in acquire_op
op = self._op_loader.load_operator(hub_op_id, op_args, op_kws, tag)
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\operator_manager\operator_loader.py", line 151, in load_operator
raise RuntimeError('Load operator failed')
RuntimeError: Load operator failed

I am using towhee 1.0.0

Please help in solving the error. I am have read other issues in this repo and tried those. But still getting the same error.

Thanks in adavance

[junjiejiangjjj]

Try this code to check if it runs success

>>> from transformers import DPRContextEncoder
>>> DPRContextEncoder.from_pretrained("facebook/dpr-ctx_encoder-single-nq-base")

[gopig99]

No, Still getting same error.

I am thinking it is because of ops.text_embedding.dpr

[junjiejiangjjj]

If you run failed with this code, it means you can't access the huggingface hub to download models. You need to make sure your net can access huggingface hub.

[gopig99]

Change code to this
from transformers import DPRContextEncoder
model = DPRContextEncoder.from_pretrained("facebook/dpr-ctx_encoder-single-nq-base")

insert_pipe = (
pipe.input('id', 'question', 'answer')
.map('question', 'vec', ops.text_embedding.dpr(model_name=model))
.map('vec', 'vec', lambda x: x / np.linalg.norm(x, axis=0))
.map(('id', 'vec'), 'insert_status', ops.ann_insert.milvus_client(host='127.0.0.1', port='19530', collection_name='question_answer'))
.output()
)
And following is the error

RuntimeError Traceback (most recent call last)
File :12

File ~\anaconda3\lib\site-packages\towhee\runtime\pipeline.py:116, in Pipeline.output(self, *output_schema, **config_kws)
113 dag_dict[self._clo_node]['next_nodes'].append(uid)
115 run_pipe = RuntimePipeline(dag_dict, config=config_kws)
--> 116 run_pipe.preload()
117 return run_pipe

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:140, in RuntimePipeline.preload(self)
136 def preload(self):
137 """
138 Preload the operators.
139 """
--> 140 return _Graph(self._dag_repr.nodes, self._dag_repr.edges, self._operator_pool, self._thread_pool)

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:64, in _Graph.init(self, nodes, edges, operator_pool, thread_pool, enable_trance)
62 self.features = None
63 self.time_profiler.record(Event.pipe_name, Event.pipe_in)
---> 64 self.initialize()
65 self._input_queue = self._data_queues[0]

File ~\anaconda3\lib\site-packages\towhee\runtime\runtime_pipeline.py:75, in _Graph.initialize(self)
73 node = create_node(self._nodes[name], self._operator_pool, in_queues, out_queues, self._time_profiler)
74 if not node.initialize():
---> 75 raise RuntimeError(node.err_msg)
76 self._node_runners.append(node)

RuntimeError: Node-text-embedding/dpr-0 runs failed, error msg: Create text-embedding/dpr-0 operator text-embedding/dpr:main with args None and kws {'model_name': DPRContextEncoder(
(ctx_encoder): DPREncoder(
(bert_model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(30522, 768, padding_idx=0)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(1): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(2): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(3): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(4): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(5): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(6): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(7): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(8): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(9): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(10): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(11): BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
)
)
)} failed, err: Load operator failed, Traceback (most recent call last):
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\nodes\node.py", line 88, in initialize
self._op = self._op_pool.acquire_op(
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\operator_manager\operator_pool.py", line 99, in acquire_op
op = self._op_loader.load_operator(hub_op_id, op_args, op_kws, tag)
File "C:\Users\Manoj\anaconda3\lib\site-packages\towhee\runtime\operator_manager\operator_loader.py", line 151, in load_operator
raise RuntimeError('Load operator failed')
RuntimeError: Load operator failed

[junjiejiangjjj]

Use towhee=1.1.3 and try this code, it works OK in my env.

from towhee import pipe, ops, register, operator
from transformers import DPRContextEncoder, DPRContextEncoderTokenizer

import numpy as np


@register
class DPR(operator.NNOperator):
    def __init__(self):
        self.tokenizer = DPRContextEncoderTokenizer.from_pretrained('facebook/dpr-ctx_encoder-single-nq-base')
        self.model = DPRContextEncoder.from_pretrained('facebook/dpr-ctx_encoder-single-nq-base')
        self.model.eval()

    def __call__(self, text: str):
        input_ids = self.tokenizer(text, return_tensors="pt")["input_ids"]
        embeddings = self.model(input_ids).pooler_output
        return embeddings.squeeze(0).detach().numpy()

insert_pipe = (
pipe.input('id', 'question', 'answer')
    .map('question', 'vec', ops.DPR())
    .map('vec', 'vec', lambda x: x / np.linalg.norm(x, axis=0))
    .output('vec')
)

[gopig99]

Sorry for delayed reply. Yes, this code is working. Thank you