[Example] fix print info (dmlc#2305)

examples/pytorch/graphsage/experimental/train_dist.py

@@ -237,7 +237,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Part {} | Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB | time {:.3f} s'.format(
+                print('Part {} | Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MB | time {:.3f} s'.format(
                     g.rank(), epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), gpu_mem_alloc, np.sum(step_time[-args.log_every:])))
             start = time.time()
 

examples/pytorch/graphsage/train_sampling.py

@@ -174,7 +174,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), gpu_mem_alloc))
             tic_step = time.time()
 

examples/pytorch/graphsage/train_sampling_multi_gpu.py

@@ -195,7 +195,7 @@ def run(proc_id, n_gpus, args, devices, data):
                 iter_tput.append(len(seeds) * n_gpus / (time.time() - tic_step))
             if step % args.log_every == 0 and proc_id == 0:
                 acc = compute_acc(batch_pred, batch_labels)
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), th.cuda.max_memory_allocated() / 1000000))
 
         if n_gpus > 1:

examples/pytorch/graphsage/train_sampling_unsupervised.py

@@ -271,7 +271,7 @@ def run(proc_id, n_gpus, args, devices, data):
             iter_t.append(t - d_step)
             if step % args.log_every == 0:
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('[{}]Epoch {:05d} | Step {:05d} | Loss {:.4f} | Speed (samples/sec) {:.4f}|{:.4f} | Load {:.4f}| train {:.4f} | GPU {:.1f} MiB'.format(
+                print('[{}]Epoch {:05d} | Step {:05d} | Loss {:.4f} | Speed (samples/sec) {:.4f}|{:.4f} | Load {:.4f}| train {:.4f} | GPU {:.1f} MB'.format(
                     proc_id, epoch, step, loss.item(), np.mean(iter_pos[3:]), np.mean(iter_neg[3:]), np.mean(iter_d[3:]), np.mean(iter_t[3:]), gpu_mem_alloc))
             tic_step = time.time()
 

examples/pytorch/ogb/cluster-gat/main.py

@@ -189,7 +189,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), gpu_mem_alloc))
                 tic_start = time.time()
 

examples/pytorch/ogb/cluster-sage/main.py

@@ -158,7 +158,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), gpu_mem_alloc))
 
         toc = time.time()

examples/pytorch/ogb/ogbn-products/gat/main.py

@@ -184,7 +184,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), gpu_mem_alloc))
 
         toc = time.time()

examples/pytorch/ogb/ogbn-products/graphsage/main.py

@@ -180,7 +180,7 @@ def run(args, device, data):
             if step % args.log_every == 0:
                 acc = compute_acc(batch_pred, batch_labels)
                 gpu_mem_alloc = th.cuda.max_memory_allocated() / 1000000 if th.cuda.is_available() else 0
-                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB'.format(
+                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MB'.format(
                     epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), gpu_mem_alloc))
 
         toc = time.time()