From d8375cb28fb762fd6f0aed7566f7d1359cb6202d Mon Sep 17 00:00:00 2001
From: "dhsig552@163.com" <dhsig552@163.com>
Date: Wed, 6 Sep 2017 10:27:48 +0100
Subject: [PATCH] inferencing with top k 2

---
 README.md              | 2 +-
 main_simple_seq2seq.py | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 4dc6f5e2..fcfd81df 100644
--- a/README.md
+++ b/README.md
@@ -34,7 +34,7 @@ Query > donald trump won last nights presidential debate according to snap onlin
  > i think he was a racist
  > he is not a racist
  > he is a liar
- > i dont think he is a disgrace
+ > trump needs to be president
 
 ```
 <!---
diff --git a/main_simple_seq2seq.py b/main_simple_seq2seq.py
index 913aad27..9c78d53e 100644
--- a/main_simple_seq2seq.py
+++ b/main_simple_seq2seq.py
@@ -213,7 +213,7 @@ def model(encode_seqs, decode_seqs, is_train=True, reuse=False):
                         o, state = sess.run([y, net_rnn.final_state_decode],
                                         {net_rnn.initial_state_decode: state,
                                         decode_seqs2: [[w_id]]})
-                        w_id = tl.nlp.sample_top(o[0], top_k=3)
+                        w_id = tl.nlp.sample_top(o[0], top_k=2)
                         w = idx2w[w_id]
                         if w_id == end_id:
                             break