Add label semantic examples with new Fluid api #10368
Conversation
sidgoyal78
requested review from
helinwang,
cs2be,
wangkuiyi,
cxysteven,
reyoung and
abhinavarora
and removed request for
cs2be and
cxysteven
|
|
||
|
|
||
| def inference_network(): | ||
| word = fluid.layers.data( |
There was a problem hiding this comment.
Can we put data layer definitions into lstm_net?
| mark_dict_len = 2 | ||
|
|
||
| def lstm_net(word, predicate, ctx_n2, ctx_n1, ctx_0, ctx_p1, ctx_p2, mark, | ||
| **ignored): |
There was a problem hiding this comment.
Why do we need **ignored?
| return crf_decode | ||
|
|
||
| def train_network(): | ||
| mix_hidden_lr = 1e-3 |
There was a problem hiding this comment.
Some of the constants in this file is upper case snake case (BATCH_SIZE), some are snake case (mix_hidden_lr). Could you change all constants to upper case snake case (BATCH_SIZE)?
| event.batch_id + 1, avg_cost)) | ||
|
|
||
| if avg_cost > 0.01: # Low threshold for speeding up CI | ||
| trainer.params.save(save_path) |
There was a problem hiding this comment.
Sorry that the API changed a little, please use trainer.save_params (https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/fluid/tests/book/word2vec/no_test_word2vec_new_api.py#L103)
| def infer(use_cuda, save_path): | ||
| params = fluid.Params(save_path) | ||
| place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace() | ||
| inferencer = fluid.Inferencer(inference_network, params, place=place) |
There was a problem hiding this comment.
Sorry that the API changed a little (param.py is removed), please use
https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/fluid/tests/book/word2vec/no_test_word2vec_new_api.py#L108 and https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/fluid/tests/book/word2vec/no_test_word2vec_new_api.py#L118
| if (event.batch_id % 10) == 0: | ||
| avg_cost = trainer.test(reader=test_reader) | ||
|
|
||
| print('BatchID {1:04}, Loss {2:2.2}'.format(event.batch_id + 1, |
There was a problem hiding this comment.
I think it should be BatchID {0:04}, Loss {1:2.2}
… simplify_fluid_labelrole
Issues:
train_network()andinference_network().