Upload llama-1B/8_GPUS/dp-2_tp-2_pp-2_mbz-1

.gitattributes

@@ -132,3 +132,4 @@ llama-1B/8_GPUS/dp-8_tp-1_pp-1_mbz-1/profiler/ip-26-0-171-62_4129089.17200726293
 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-2_tp-2_pp-2_mbz-1/profiler/ip-26-0-163-134_3359572.1720135363501644918.pt.trace.json filter=lfs diff=lfs merge=lfs -text

llama-1B/8_GPUS/dp-2_tp-2_pp-2_mbz-1/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-2_tp-2_pp-2_mbz-1/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
+1,4190000.0000000005,56600.0,74100.0,9260.0,1020.0,11.2,0.0001,84.0,84.0,21.2,4034.71,8570.26,9170.0
+2,8390000.0,41600.0,101000.0,12600.0,1020.0,11.2,9.53e-05,114.0,114.0,21.3,4034.71,5682.98,9170.0
+3,12600000.0,48700.0,86100.0,10800.0,1020.0,9.94,9.05e-05,97.6,97.6,115.0,4034.71,5682.98,9170.0
+4,16800000.0,65500.0,64099.99999999999,8010.0,1020.0,13.3,8.58e-05,72.7,72.7,22.8,,,
+5,21000000.0,65600.0,63900.0,7990.0,1020.0,10.2,8.11e-05,72.5,72.5,10.4,4034.71,8570.26,9170.0
+6,25200000.0,65500.0,64000.0,8000.0,1020.0,9.36,7.63e-05,72.6,72.6,15.4,4034.71,8570.26,9170.0
+7,29400000.0,47500.0,88200.0,11000.0,1020.0,8.8,7.16e-05,100.0,100.0,8.89,4034.71,8570.26,9170.0
+8,33600000.0,46800.0,89700.0,11200.0,1020.0,8.59,6.68e-05,102.0,102.0,5.7,4034.71,8570.26,9170.0
+9,37700000.0,46600.0,89900.0,11200.0,1020.0,8.22,6.21e-05,102.0,102.0,4.83,4034.71,8570.26,9170.0
+10,41900000.0,46800.0,89600.0,11200.0,1020.0,8.03,5.74e-05,102.0,102.0,6.81,4034.71,8570.26,9170.0
+11,46100000.0,46800.0,89700.0,11200.0,1020.0,7.73,5.26e-05,102.0,102.0,4.01,4034.71,8570.26,9170.0
+12,50300000.0,46800.0,89700.0,11200.0,1020.0,7.55,4.79e-05,102.0,102.0,3.56,4034.71,8570.26,9170.0
+13,54500000.0,46600.0,89900.0,11200.0,1020.0,7.49,4.32e-05,102.0,102.0,4.42,4034.71,8570.26,9170.0
+14,58700000.0,46900.0,89500.0,11200.0,1020.0,7.36,3.84e-05,101.0,101.0,3.98,4034.71,8570.26,9170.0
+15,62900000.0,46800.0,89600.0,11200.0,1020.0,7.22,3.37e-05,102.0,102.0,2.83,4034.71,8570.26,9170.0
+16,67099999.99999999,46700.0,89700.0,11200.0,1020.0,7.12,2.89e-05,102.0,102.0,2.74,4034.71,8570.26,9170.0
+17,71300000.0,46900.0,89400.0,11200.0,1020.0,7.02,2.42e-05,101.0,101.0,2.52,4034.71,8570.26,9170.0
+18,75500000.0,47000.0,89300.0,11200.0,1020.0,6.94,1.95e-05,101.0,101.0,2.54,4034.71,8570.26,9170.0
+19,79700000.0,47500.0,88400.0,11000.0,1020.0,6.89,1.47e-05,100.0,100.0,2.97,,,
+20,83900000.0,45800.0,91600.0,11500.0,1020.0,6.84,1e-05,104.0,104.0,2.9,,,

llama-1B/8_GPUS/dp-2_tp-2_pp-2_mbz-1/profiler.csv

@@ -0,0 +1,2 @@
+forward,backward
+1ms 83μs,0ms 331μs

llama-1B/8_GPUS/dp-2_tp-2_pp-2_mbz-1/profiler/ip-26-0-163-134_3359572.1720135363501644918.pt.trace.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f43bc07563afb54801ed9852e87461dd889d6b147246fb27a9653dad4c27c2bb
+size 42800786785

llama-1B/8_GPUS/dp-2_tp-2_pp-2_mbz-1/status.txt

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