adding active mem stat

utils/memory_utils.py

@@ -50,6 +50,7 @@ class MemoryTrace:
         self.end = byte2gb(torch.cuda.memory_allocated())
         self.peak = byte2gb(torch.cuda.max_memory_allocated())
         cuda_info = torch.cuda.memory_stats()
+        self.peak_active_gb = byte2gb(cuda_info["active_bytes.all.peak"])
         self.cuda_malloc_retires = cuda_info.get("num_alloc_retries", 0)
         self.m_cuda_ooms = cuda_info.get("num_ooms", 0)
         self.used = byte2gb(self.end - self.begin)

utils/train_utils.py

@@ -78,13 +78,11 @@ def train(model, train_dataloader,eval_dataloader, tokenizer, optimizer, lr_sche
             model.train()
             total_loss = 0.0
             data_set_len = 0
-            
             for step, batch in enumerate(tqdm(train_dataloader,colour="blue", desc=f"Training Epoch{epoch}")):
                 for key in batch.keys():
                     if train_config.enable_fsdp:
                         batch[key] = batch[key].to(local_rank)
                     else:
-
                         batch[key] = batch[key].to('cuda:0')              
                 loss = model(**batch).loss
                 loss = loss / gradient_accumulation_steps
@@ -117,6 +115,7 @@ def train(model, train_dataloader,eval_dataloader, tokenizer, optimizer, lr_sche
         
         print(f"Max CUDA memory allocated was {memtrace.peak} GB")
         print(f"Max CUDA memory reserved was {memtrace.max_reserved} GB")
+        print(f"Peak active CUDA memory was {memtrace.peak_active_gb} GB")
         print(f"Cuda Malloc retires : {memtrace.cuda_malloc_retires}")
         print(f"CPU Total Peak Memory consumed during the train (max): {memtrace.cpu_peaked + memtrace.cpu_begin} GB")