paths

1 files changed, 8 insertions, 7 deletions

diff --git a/inversion/image_inversion.py b/inversion/image_inversion.py
index 5556778..2927eac 100644
--- a/inversion/image_inversion.py
+++ b/inversion/image_inversion.py
@@ -68,11 +68,12 @@ params = params.Params(sys.argv[1])
 # --------------------------
 # Global directories.
 # --------------------------
+LATENT_TAG = 'latent_' if params.inv_layer == 'latent' else 'dense_'
 BATCH_SIZE = params.batch_size
 SAMPLE_SIZE = params.sample_size
-LOGS_DIR = 'logs'
-SAMPLES_DIR = 'samples'
-INVERSES_DIR = 'inverses'
+LOGS_DIR = os.path.join('inverses', params.tag, LATENT_TAG, 'logs')
+SAMPLES_DIR = os.path.join('inverses', params.tag, LATENT_TAG, 'samples')
+INVERSES_DIR = os.path.join('inverses', params.tag)
 if not os.path.exists(LOGS_DIR):
   os.makedirs(LOGS_DIR)
 if not os.path.exists(SAMPLES_DIR):
@@ -394,7 +395,7 @@ for image_batch, label_batch in image_gen:
       sess.run(clip_latent)
 
     # Every 100 iterations save logs with training information.
-    if it % 100 == 99:
+    if it < 100 or it % 100 == 0:
       # Log losses.
       etime = time.time() - start_time
       print('It [{:8d}] time [{:5.1f}] total [{:.4f}] mse [{:.4f}] '
@@ -436,7 +437,7 @@ for image_batch, label_batch in image_gen:
         inv_batch = vs.interleave(vs.data2img(image_batch[BATCH_SIZE - SAMPLE_SIZE:]),
                         vs.data2img(gen_images[BATCH_SIZE - SAMPLE_SIZE:]))
         inv_batch = vs.grid_transform(inv_batch)
-        vs.save_image('{}/progress_{}.png'.format(SAMPLES_DIR, it), inv_batch)
+        vs.save_image('{}/progress_{}_{}.png'.format(SAMPLES_DIR, params.tag, it), inv_batch)
 
         # Save linear interpolation between the actual and generated encodings.
         if params.dist_loss and it % 1000 == 999:
@@ -448,7 +449,7 @@ for image_batch, label_batch in image_gen:
             inv_batch = vs.interleave(vs.data2img(image_batch[BATCH_SIZE - SAMPLE_SIZE:]),
                 vs.data2img(gen_images[BATCH_SIZE - SAMPLE_SIZE:]))
             inv_batch = vs.grid_transform(inv_batch)
-            vs.save_image('{}/progress_{}_lat_{}.png'.format(SAMPLES_DIR,it,j),
+            vs.save_image('{}/progress_{}_{}_lat_{}.png'.format(SAMPLES_DIR,params.tag,it,j),
                           inv_batch)
           sess.run(encoding.assign(enc_batch))
 
@@ -462,7 +463,7 @@ for image_batch, label_batch in image_gen:
     inv_batch = vs.interleave(vs.data2img(image_batch[BATCH_SIZE - SAMPLE_SIZE:]),
                     vs.data2img(gen_images[BATCH_SIZE - SAMPLE_SIZE:]))
     inv_batch = vs.grid_transform(inv_batch)
-    vs.save_image('{}/{}.png'.format(SAMPLES_DIR, out_pos), inv_batch)
+    vs.save_image('{}/{}_{}.png'.format(SAMPLES_DIR, params.tag, out_pos), inv_batch)
     print('Saved samples for out_pos: {}.'.format(out_pos))
 
   # Save images that are ready.