Upload llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4

.gitattributes

@@ -131,3 +131,4 @@ llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-16/profiler/ip-26-0-171-88_1107214.1720050881
 llama-1B/8_GPUS/dp-8_tp-1_pp-1_mbz-1/profiler/ip-26-0-171-62_4129089.1720072629336630088.pt.trace.json filter=lfs diff=lfs merge=lfs -text
 llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-8/profiler/ip-26-0-171-249_403602.1720133712161162722.pt.trace.json filter=lfs diff=lfs merge=lfs -text
 llama-1B/8_GPUS/dp-2_tp-4_pp-1_mbz-4/profiler/ip-26-0-163-226_4070653.1720134059871777463.pt.trace.json filter=lfs diff=lfs merge=lfs -text
+llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/profiler/ip-26-0-165-164_3524991.1720134704668975445.pt.trace.json filter=lfs diff=lfs merge=lfs -text

llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/bench.slurm

@@ -5,7 +5,7 @@
 #SBATCH --partition=hopper-prod
 #SBATCH --nodes=1
 #SBATCH --gres=gpu:8
-#SBATCH --qos=normal
+#SBATCH --qos=low
 #SBATCH --ntasks-per-node=1
 #SBATCH --cpus-per-task=96
 #SBATCH --exclusive

llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/log_metrics.csv

@@ -0,0 +1,21 @@
+iteration,consumed_tokens,elapsed_time_per_iteration_ms,tokens_per_sec,tokens_per_sec_per_gpu,global_batch_size,lm_loss,lr,model_tflops_per_gpu,hardware_tflops_per_gpu,grad_norm,memory_usage_MiB,peak_allocated_MiB,peak_reserved_MiB
+1,4190000.0000000005,44500.0,94300.0,11800.0,1020.0,11.5,0.0001,107.0,107.0,15.7,2488.12,9689.3,10130.0
+2,8390000.0,35800.0,117000.0,14600.0,1020.0,11.5,9.53e-05,133.0,133.0,16.0,2488.12,9689.3,10130.0
+3,12600000.0,42300.0,99100.0,12400.0,1020.0,12.8,9.05e-05,112.0,112.0,137.0,2488.12,9689.3,10130.0
+4,16800000.0,55600.0,75400.0,9420.0,1020.0,12.2,8.58e-05,85.5,85.5,22.4,2488.0,2488.2,10130.0
+5,21000000.0,55900.0,75000.0,9380.0,1020.0,12.4,8.11e-05,85.1,85.1,42.8,2488.0,9689.3,10130.0
+6,25200000.0,56700.0,73900.0,9240.0,1020.0,11.1,7.63e-05,83.9,83.9,24.8,2488.0,9689.3,10130.0
+7,29400000.0,43300.0,96800.0,12100.0,1020.0,10.2,7.16e-05,110.0,110.0,12.1,2488.0,9689.3,10130.0
+8,33600000.0,41400.0,101000.0,12700.0,1020.0,9.8,6.68e-05,115.0,115.0,7.31,2488.0,9689.3,10130.0
+9,37700000.0,41200.0,102000.0,12700.0,1020.0,9.32,6.21e-05,116.0,116.0,6.66,2488.0,9689.3,10130.0
+10,41900000.0,41200.0,102000.0,12700.0,1020.0,9.22,5.74e-05,115.0,115.0,16.3,2488.0,9689.3,10130.0
+11,46100000.0,41300.0,102000.0,12700.0,1020.0,8.63,5.26e-05,115.0,115.0,7.95,2488.0,9689.3,10130.0
+12,50300000.0,41400.0,101000.0,12700.0,1020.0,8.27,4.79e-05,115.0,115.0,5.43,2488.0,9689.3,10130.0
+13,54500000.0,41300.0,102000.0,12700.0,1020.0,8.1,4.32e-05,115.0,115.0,5.53,2488.0,9689.3,10130.0
+14,58700000.0,41500.0,101000.0,12600.0,1020.0,7.93,3.84e-05,115.0,115.0,5.77,2488.0,9689.3,10130.0
+15,62900000.0,41300.0,102000.0,12700.0,1020.0,7.72,3.37e-05,115.0,115.0,5.17,2488.0,9689.3,10130.0
+16,67099999.99999999,41500.0,101000.0,12600.0,1020.0,7.56,2.89e-05,115.0,115.0,4.94,2488.0,9689.3,10130.0
+17,71300000.0,41200.0,102000.0,12700.0,1020.0,7.45,2.42e-05,116.0,116.0,4.93,2488.0,9689.3,10130.0
+18,75500000.0,41500.0,101000.0,12600.0,1020.0,7.35,1.95e-05,115.0,115.0,4.02,2488.0,9689.3,10130.0
+19,79700000.0,41200.0,102000.0,12700.0,1020.0,7.29,1.47e-05,115.0,115.0,4.11,2488.0,9689.3,10130.0
+20,83900000.0,41400.0,101000.0,12700.0,1020.0,7.23,1e-05,115.0,115.0,3.96,,,

llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/profiler.csv

@@ -0,0 +1,2 @@
+forward,backward
+1ms 145μs,0ms 317μs

llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/profiler/ip-26-0-165-164_3524991.1720134704668975445.pt.trace.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a52ec46da5883e26b13f3661861e5376f495a0ac0c550bc419f408ef00f41fb1
+size 36780932278

llama-1B/8_GPUS/dp-1_tp-8_pp-1_mbz-4/status.txt

@@ -1 +1 @@
-timeout
+completed