fixing a bug introduced in previous commit. We have to use doubles no…

…t floats when we convert GPU models to CPU, because CPU trained models use doubles. This introduced an error where a CPU checkpoint could not be loaded and used on CPU because the initial state was of float type. Anyone who has converted a model with the old script must rerun the conversion script so that double() CPU models are saved instead of float().

convert_gpu_cpu_checkpoint.lua

@@ -64,7 +64,7 @@ protos = checkpoint.protos
 -- convert the networks to be CPU models
 for k,v in pairs(protos) do
 	print('converting ' .. k .. ' to CPU')
-	protos[k]:float()
+	protos[k]:double()
 end
 
 local savefile = opt.model .. '_cpu.t7' -- append "cpu.t7" to filename

sample.lua

@@ -51,6 +51,7 @@ if opt.gpuid >= 0 and opt.opencl == 0 then
     if not ok2 then gprint('package cutorch not found!') end
     if ok and ok2 then
         gprint('using CUDA on GPU ' .. opt.gpuid .. '...')
+        gprint('Make sure that your saved checkpoint was also trained with GPU. If it was trained with CPU use -gpuid -1 for sampling as well')
         cutorch.setDevice(opt.gpuid + 1) -- note +1 to make it 0 indexed! sigh lua
         cutorch.manualSeed(opt.seed)
     else
@@ -66,7 +67,8 @@ if opt.gpuid >= 0 and opt.opencl == 1 then
     if not ok then print('package clnn not found!') end
     if not ok2 then print('package cltorch not found!') end
     if ok and ok2 then
-        print('using OpenCL on GPU ' .. opt.gpuid .. '...')
+        gprint('using OpenCL on GPU ' .. opt.gpuid .. '...')
+        gprint('Make sure that your saved checkpoint was also trained with GPU. If it was trained with CPU use -gpuid -1 for sampling as well')
         cltorch.setDevice(opt.gpuid + 1) -- note +1 to make it 0 indexed! sigh lua
         torch.manualSeed(opt.seed)
     else
@@ -96,7 +98,7 @@ local current_state
 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):float()
+    local h_init = torch.zeros(1, checkpoint.opt.rnn_size):double()
     if opt.gpuid >= 0 and opt.opencl == 0 then h_init = h_init:cuda() end
     if opt.gpuid >= 0 and opt.opencl == 1 then h_init = h_init:cl() end
     table.insert(current_state, h_init:clone())