diff --git a/Readme.md b/Readme.md
index 254eb3d..72d8b52 100644
--- a/Readme.md
+++ b/Readme.md
@@ -35,6 +35,13 @@ $ nohup th web_backend.lua &
 $ nohup python web_server.py & 
 ```
 
+## Model conversion between cpu and gpu
+I add a script to convert a model file trained by gpu to cpu model.
+You can try it as follow:
+```bash
+$ th convert.lua gpu_model cpu_model
+```
+
 -----------------------------------------------
 ## Karpathy's raw Readme
 please follow this to setup your experiment.
diff --git a/convert.lua b/convert.lua
new file mode 100644
index 0000000..b7b0682
--- /dev/null
+++ b/convert.lua
@@ -0,0 +1,38 @@
+require 'torch'
+require 'nngraph'
+require 'optim'
+require 'lfs'
+require 'nn'
+
+require 'util.OneHot'
+require 'util.misc'
+
+cmd = torch.CmdLine()
+cmd:text()
+cmd:text('convert a gpu model to cpu one')
+cmd:text()
+cmd:text('Options')
+
+cmd:argument('-load_model','model to convert')
+cmd:argument('-save_file','the file path to save the converted model')
+cmd:option('-gpuid',0,'which gpu to use. -1 = use CPU')
+cmd:text()
+
+-- parse input params
+opt = cmd:parse(arg)
+local ok, cunn = pcall(require, 'cunn')
+local ok2, cutorch = pcall(require, 'cutorch')
+if not ok then gprint('package cunn not found!') end
+if not ok2 then gprint('package cutorch not found!') end
+if ok and ok2 then
+    print('using CUDA on GPU ' .. opt.gpuid .. '...')
+    cutorch.setDevice(opt.gpuid + 1) -- note +1 to make it 0 indexed! sigh lua
+else
+    print('No gpu found. Convert fail.')
+    os.exit()
+end
+
+checkpoint = torch.load(opt.load_model)
+checkpoint.protos.rnn:float()
+checkpoint.protos.criterion:float()
+torch.save(opt.save_file, checkpoint)
diff --git a/sample.lua b/sample.lua
index 8504017..fb58685 100644
--- a/sample.lua
+++ b/sample.lua
@@ -80,7 +80,7 @@ local num_layers = checkpoint.opt.num_layers
 current_state = {}
 for L = 1,checkpoint.opt.num_layers do
     -- c and h for all layers
-    local h_init = torch.zeros(1, checkpoint.opt.rnn_size)
+    local h_init = torch.zeros(1, checkpoint.opt.rnn_size):float()
     if opt.gpuid >= 0 then h_init = h_init:cuda() end
     table.insert(current_state, h_init:clone())
     table.insert(current_state, h_init:clone())
diff --git a/web_backend.lua b/web_backend.lua
index 7eaf133..8483e30 100644
--- a/web_backend.lua
+++ b/web_backend.lua
@@ -89,7 +89,7 @@ for msg in client:pubsub({subscribe = channels}) do
         current_state = {}
         for L = 1,checkpoint.opt.num_layers do
             -- c and h for all layers
-            local h_init = torch.zeros(1, checkpoint.opt.rnn_size)
+            local h_init = torch.zeros(1, checkpoint.opt.rnn_size):float()
             if gpuid >= 0 then h_init = h_init:cuda() end
             table.insert(current_state, h_init:clone())
             table.insert(current_state, h_init:clone())