nohup: ignoring input Namespace(save_dir='/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg', self_attn_layer_to_quant='23 22 25 24 26', mlp_layer_to_quant='27 16 19 17 25', model_id='/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen/Qwen2.5-7B', cuda_id=6) `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1791020:1791807 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1791018:1791806 [0] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:1791018:1791806 [0] NCCL INFO NET/IB : Using [0]mlx5_0:1/RoCE [1]mlx5_1:1/RoCE [2]mlx5_2:1/RoCE [3]mlx5_3:1/RoCE [RO]; OOB eth0:fdbd:dc03:9:451::154<0> n136-128-154:1791018:1791806 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1791020:1791807 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1791020:1791807 [1] NCCL INFO Using network IB n136-128-154:1791018:1791806 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1791018:1791806 [0] NCCL INFO Using network IB n136-128-154:1791020:1791807 [1] NCCL INFO ncclCommInitRankConfig comm 0xf0a51c0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x29335eaf64f1b6d4 - Init START n136-128-154:1791018:1791806 [0] NCCL INFO ncclCommInitRankConfig comm 0xec12000 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x29335eaf64f1b6d4 - Init START n136-128-154:1791018:1791806 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1791020:1791807 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1791020:1791807 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1791020:1791807 [1] NCCL INFO Retrieving state for IB n136-128-154:1791020:1791807 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1791020:1791807 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1791020:1791807 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1791020:1791807 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1791020:1791807 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1791018:1791806 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1791018:1791806 [0] NCCL INFO Retrieving state for IB n136-128-154:1791018:1791806 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1791018:1791806 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1791018:1791806 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1791018:1791806 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1791018:1791806 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1791018:1791806 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1791020:1791807 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1791018:1791806 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1791020:1791807 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1791018:1791806 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1791018:1791806 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1791018:1791806 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1791018:1791806 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1791018:1791806 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1791018:1791806 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1791018:1791806 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1791018:1791806 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1791018:1791806 [0] NCCL INFO ========================================== n136-128-154:1791018:1791806 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1791018:1791806 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1791018:1791806 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1791020:1791807 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1791020:1791807 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1791020:1791807 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1791020:1791807 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1791020:1791807 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1791020:1791807 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1791020:1791807 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1791020:1791807 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1791020:1791807 [1] NCCL INFO ========================================== n136-128-154:1791020:1791807 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1791020:1791807 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1791020:1791807 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1791018:1791806 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1791018:1791806 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1791018:1791806 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791018:1791806 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791018:1791806 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791018:1791806 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791018:1791806 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791018:1791806 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1791018:1791806 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1791020:1791807 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1791020:1791807 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791020:1791807 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1791018:1791806 [0] NCCL INFO comm 0xec12000 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1791020:1791807 [1] NCCL INFO comm 0xf0a51c0 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791020:1791807 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1791020:1791807 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1791020:1791807 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1791020:1791807 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1791018:1791806 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1791018:1791806 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1791020:1791807 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1791020:1791819 [1] NCCL INFO [Proxy Service] Device 1 CPU core 44 n136-128-154:1791020:1791820 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 109 n136-128-154:1791018:1791806 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1791018:1791806 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1791018:1791821 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:1791018:1791822 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 100 n136-128-154:1791020:1791807 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1791020:1791807 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1791018:1791806 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1791018:1791806 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1791018:1791806 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1791020:1791807 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1791020:1791807 [1] NCCL INFO ncclCommInitRankConfig comm 0xf0a51c0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x29335eaf64f1b6d4 - Init COMPLETE n136-128-154:1791020:1791807 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.51 (kernels 0.21, alloc 0.20, bootstrap 0.00, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:1791018:1791806 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1791018:1791806 [0] NCCL INFO ncclCommInitRankConfig comm 0xec12000 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x29335eaf64f1b6d4 - Init COMPLETE n136-128-154:1791018:1791806 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.51 (kernels 0.22, alloc 0.20, bootstrap 0.00, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:1791020:1791823 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791020:1791823 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1791018:1791824 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1791020:1791823 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-06:15:46:09 INFO [evaluator:559] Running loglikelihood requests 2025-12-06:15:46:09 INFO [evaluator:559] Running loglikelihood requests Running loglikelihood requests: 0%| | 0/1268 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1791020:1791915 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1791020:1791819 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-06:15:46:42 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: auto (64) | Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr| |----------|------:|------|-----:|------|---|-----:|---|-----:| |winogrande| 1|none | 5|acc |↑ |0.5588|± | 0.014| [rank0]:[W1206 15:46:42.102750030 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1791018:1791821 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1791018:1792101 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1791020:1791819 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1791020:1791920 [1] NCCL INFO comm 0xf0a51c0 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1791018:1792101 [0] NCCL INFO comm 0xec12000 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 winogrande 评估完成! ================================================== 开始评估:任务=gsm8k | 少样本数=4 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/gsm8k.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-06:15:48:07 INFO [__main__:440] Selected Tasks: ['gsm8k'] 2025-12-06:15:48:07 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-06:15:48:07 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-06:15:48:07 INFO [__main__:440] Selected Tasks: ['gsm8k'] 2025-12-06:15:48:07 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-06:15:48:07 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-06:15:48:09 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-06:15:48:09 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00', '<|im_end|>'], 'do_sample': False, 'temperature': 0.0} 2025-12-06:15:48:32 WARNING [evaluator:309] Overwriting default num_fewshot of gsm8k from 5 to 4 2025-12-06:15:48:32 INFO [api.task:434] Building contexts for gsm8k on rank 1... 0%| | 0/659 [00:00', '<|im_end|>'], 'do_sample': False, 'temperature': 0.0} 2025-12-06:15:48:35 WARNING [evaluator:309] Overwriting default num_fewshot of gsm8k from 5 to 4 2025-12-06:15:48:35 INFO [api.task:434] Building contexts for gsm8k on rank 0... 0%| | 0/660 [00:00 n136-128-154:1792223:1792658 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1792224:1792659 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:1792224:1792659 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/RoCE [1]mlx5_1:1/RoCE [2]mlx5_2:1/RoCE [3]mlx5_3:1/RoCE [RO]; OOB eth0:fdbd:dc03:9:451::154<0> n136-128-154:1792224:1792659 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1792223:1792658 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1792223:1792658 [0] NCCL INFO Using network IB n136-128-154:1792224:1792659 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1792224:1792659 [1] NCCL INFO Using network IB n136-128-154:1792224:1792659 [1] NCCL INFO ncclCommInitRankConfig comm 0x10b76070 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x380a53995dbc77d1 - Init START n136-128-154:1792223:1792658 [0] NCCL INFO ncclCommInitRankConfig comm 0xf8345c0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x380a53995dbc77d1 - Init START n136-128-154:1792224:1792659 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1792223:1792658 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1792224:1792659 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1792224:1792659 [1] NCCL INFO Retrieving state for IB n136-128-154:1792224:1792659 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1792224:1792659 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1792223:1792658 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1792223:1792658 [0] NCCL INFO Retrieving state for IB n136-128-154:1792223:1792658 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1792224:1792659 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1792223:1792658 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1792224:1792659 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1792223:1792658 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1792224:1792659 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1792223:1792658 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1792223:1792658 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1792224:1792659 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1792224:1792659 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1792223:1792658 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1792223:1792658 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1792223:1792658 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1792223:1792658 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1792223:1792658 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1792223:1792658 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1792223:1792658 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1792223:1792658 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1792223:1792658 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1792223:1792658 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1792223:1792658 [0] NCCL INFO ========================================== n136-128-154:1792223:1792658 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1792223:1792658 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1792223:1792658 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1792224:1792659 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1792224:1792659 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1792224:1792659 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1792224:1792659 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1792224:1792659 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1792224:1792659 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1792224:1792659 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1792224:1792659 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1792224:1792659 [1] NCCL INFO ========================================== n136-128-154:1792224:1792659 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1792224:1792659 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1792224:1792659 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1792223:1792658 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1792223:1792658 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1792223:1792658 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792223:1792658 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1792224:1792659 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1792224:1792659 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1792224:1792659 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792224:1792659 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1792223:1792658 [0] NCCL INFO comm 0xf8345c0 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1792224:1792659 [1] NCCL INFO comm 0x10b76070 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1792223:1792658 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1792223:1792658 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1792223:1792658 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1792224:1792659 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1792224:1792659 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1792224:1792659 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1792223:1792658 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1792223:1792658 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1792223:1792673 [0] NCCL INFO [Proxy Service] Device 0 CPU core 99 n136-128-154:1792223:1792674 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 36 n136-128-154:1792224:1792659 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1792224:1792675 [1] NCCL INFO [Proxy Service] Device 1 CPU core 108 n136-128-154:1792224:1792676 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 45 n136-128-154:1792223:1792658 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1792223:1792658 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1792223:1792658 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1792224:1792659 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1792224:1792659 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1792223:1792658 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1792223:1792658 [0] NCCL INFO ncclCommInitRankConfig comm 0xf8345c0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x380a53995dbc77d1 - Init COMPLETE n136-128-154:1792223:1792658 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.57 (kernels 0.24, alloc 0.21, bootstrap 0.00, allgathers 0.01, topo 0.07, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:1792224:1792659 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1792224:1792659 [1] NCCL INFO ncclCommInitRankConfig comm 0x10b76070 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x380a53995dbc77d1 - Init COMPLETE n136-128-154:1792224:1792659 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.56 (kernels 0.24, alloc 0.21, bootstrap 0.00, allgathers 0.01, topo 0.07, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:1792223:1792678 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1792224:1792677 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792223:1792678 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1792224:1792677 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-06:15:48:37 INFO [evaluator:559] Running generate_until requests 2025-12-06:15:48:37 INFO [evaluator:559] Running generate_until requests Passed argument batch_size = auto. Detecting largest batch size Running generate_until requests: 0%| | 0/660 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1792224:1860544 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1792224:1792675 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:00:37:11 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 4, batch_size: auto |Tasks|Version| Filter |n-shot| Metric | |Value | |Stderr| |-----|------:|----------------|-----:|-----------|---|-----:|---|-----:| |gsm8k| 3|flexible-extract| 4|exact_match|↑ |0.0038|± |0.0017| | | |strict-match | 4|exact_match|↑ |0.0000|± |0.0000| [rank0]:[W1207 00:37:12.466506526 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1792223:1792673 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1792223:1860601 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1792224:1792675 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1792224:1860550 [1] NCCL INFO comm 0x10b76070 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1792223:1860601 [0] NCCL INFO comm 0xf8345c0 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 gsm8k 评估完成! ================================================== 开始评估:任务=boolq | 少样本数=0 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/boolq.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-07:00:38:37 INFO [__main__:440] Selected Tasks: ['boolq'] 2025-12-07:00:38:37 INFO [__main__:440] Selected Tasks: ['boolq'] 2025-12-07:00:38:37 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:38:37 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:38:37 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:38:37 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:38:38 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:38:39 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:38:39 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} `torch_dtype` is deprecated! Use `dtype` instead! `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1860653:1860854 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1860653:1860854 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1860653:1860854 [0] NCCL INFO Using network IB n136-128-154:1860653:1860854 [0] NCCL INFO ncclCommInitRankConfig comm 0xfdcd020 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x724ed52b33f18117 - Init START 2025-12-07:00:39:03 WARNING [evaluator:309] Overwriting default num_fewshot of boolq from None to 0 2025-12-07:00:39:03 INFO [api.task:434] Building contexts for boolq on rank 1... 0%| | 0/1635 [00:00 n136-128-154:1860654:1860868 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1860654:1860868 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1860654:1860868 [1] NCCL INFO Using network IB n136-128-154:1860654:1860868 [1] NCCL INFO ncclCommInitRankConfig comm 0xe916fa0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x724ed52b33f18117 - Init START n136-128-154:1860654:1860868 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1860653:1860854 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1860654:1860868 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1860654:1860868 [1] NCCL INFO Retrieving state for IB n136-128-154:1860654:1860868 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1860654:1860868 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1860654:1860868 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1860653:1860854 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1860653:1860854 [0] NCCL INFO Retrieving state for IB n136-128-154:1860653:1860854 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1860654:1860868 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1860653:1860854 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1860654:1860868 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1860653:1860854 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1860653:1860854 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1860653:1860854 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1860654:1860868 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1860653:1860854 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1860654:1860868 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1860653:1860854 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1860653:1860854 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1860653:1860854 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1860653:1860854 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1860653:1860854 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1860653:1860854 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1860653:1860854 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1860653:1860854 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1860653:1860854 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1860653:1860854 [0] NCCL INFO ========================================== n136-128-154:1860653:1860854 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1860653:1860854 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1860653:1860854 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1860654:1860868 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1860654:1860868 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1860654:1860868 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1860654:1860868 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1860654:1860868 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1860654:1860868 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1860654:1860868 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1860654:1860868 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1860654:1860868 [1] NCCL INFO ========================================== n136-128-154:1860654:1860868 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1860654:1860868 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1860654:1860868 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1860653:1860854 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1860653:1860854 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1860653:1860854 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860653:1860854 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1860654:1860868 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1860654:1860868 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1860654:1860868 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860654:1860868 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1860653:1860854 [0] NCCL INFO comm 0xfdcd020 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1860654:1860868 [1] NCCL INFO comm 0xe916fa0 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1860654:1860868 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1860653:1860854 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1860654:1860868 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1860654:1860868 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1860653:1860854 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1860653:1860854 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1860654:1860868 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1860653:1860854 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1860653:1860854 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1860653:1860876 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 35 n136-128-154:1860653:1860875 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:1860654:1860868 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1860654:1860877 [1] NCCL INFO [Proxy Service] Device 1 CPU core 47 n136-128-154:1860654:1860878 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 112 n136-128-154:1860653:1860854 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1860653:1860854 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1860653:1860854 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1860654:1860868 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1860654:1860868 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1860653:1860854 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1860653:1860854 [0] NCCL INFO ncclCommInitRankConfig comm 0xfdcd020 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x724ed52b33f18117 - Init COMPLETE n136-128-154:1860653:1860854 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 2.09 (kernels 0.17, alloc 0.23, bootstrap 1.52, allgathers 0.00, topo 0.05, graphs 0.00, connections 0.03, rest 0.08) n136-128-154:1860654:1860868 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1860654:1860868 [1] NCCL INFO ncclCommInitRankConfig comm 0xe916fa0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x724ed52b33f18117 - Init COMPLETE n136-128-154:1860654:1860868 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.48 (kernels 0.11, alloc 0.19, bootstrap 0.00, allgathers 0.00, topo 0.05, graphs 0.00, connections 0.04, rest 0.07) n136-128-154:1860653:1860879 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860653:1860879 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1860654:1860880 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1860653:1860879 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-07:00:39:05 INFO [evaluator:559] Running loglikelihood requests 2025-12-07:00:39:05 INFO [evaluator:559] Running loglikelihood requests Passed argument batch_size = auto:1. Detecting largest batch size Running loglikelihood requests: 0%| | 0/3270 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1860654:1860963 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1860654:1860877 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:00:40:36 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 0, batch_size: auto (45) |Tasks|Version|Filter|n-shot|Metric| |Value | |Stderr| |-----|------:|------|-----:|------|---|-----:|---|-----:| |boolq| 2|none | 0|acc |↑ |0.6373|± |0.0084| [rank0]:[W1207 00:40:37.217866339 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1860653:1860875 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1860653:1861018 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1860654:1860877 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1860654:1860968 [1] NCCL INFO comm 0xe916fa0 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1860653:1861018 [0] NCCL INFO comm 0xfdcd020 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 boolq 评估完成! ================================================== 开始评估:任务=arc_challenge | 少样本数=25 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/arc_challenge.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-07:00:42:00 INFO [__main__:440] Selected Tasks: ['arc_challenge'] 2025-12-07:00:42:00 INFO [__main__:440] Selected Tasks: ['arc_challenge'] 2025-12-07:00:42:00 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:42:00 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:42:00 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:42:00 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:42:01 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:42:01 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:42:01 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} `torch_dtype` is deprecated! Use `dtype` instead! `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1861092:1861317 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1861092:1861317 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1861092:1861317 [0] NCCL INFO Using network IB n136-128-154:1861092:1861317 [0] NCCL INFO ncclCommInitRankConfig comm 0x106c2ac0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xd0b62efd4b604bc3 - Init START n136-128-154:1861093:1861320 [1] NCCL INFO NET/Plugin: Could not find: none libnccl-net-none.so. n136-128-154:1861093:1861320 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0. n136-128-154:1861093:1861320 [1] NCCL INFO NCCL_SOCKET_FAMILY set by environment to AF_INET6 n136-128-154:1861093:1861320 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to =eth0 n136-128-154:1861093:1861320 [1] NCCL INFO NCCL_IB_HCA set to mlx5 n136-128-154:1861093:1861320 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:1861093:1861320 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/RoCE [1]mlx5_1:1/RoCE [2]mlx5_2:1/RoCE [3]mlx5_3:1/RoCE [RO]; OOB eth0:fdbd:dc03:9:451::154<0> n136-128-154:1861093:1861320 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1861093:1861320 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1861093:1861320 [1] NCCL INFO Using network IB n136-128-154:1861093:1861320 [1] NCCL INFO ncclCommInitRankConfig comm 0x10ab4400 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xd0b62efd4b604bc3 - Init START n136-128-154:1861093:1861320 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1861092:1861317 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1861093:1861320 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1861093:1861320 [1] NCCL INFO Retrieving state for IB n136-128-154:1861093:1861320 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1861093:1861320 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1861093:1861320 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1861093:1861320 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1861093:1861320 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1861092:1861317 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1861092:1861317 [0] NCCL INFO Retrieving state for IB n136-128-154:1861092:1861317 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1861092:1861317 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1861092:1861317 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1861092:1861317 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1861092:1861317 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1861092:1861317 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861092:1861317 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861093:1861320 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861093:1861320 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861092:1861317 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1861093:1861320 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1861092:1861317 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1861093:1861320 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1861092:1861317 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861093:1861320 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1861092:1861317 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861093:1861320 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1861092:1861317 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1861093:1861320 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1861092:1861317 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1861093:1861320 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1861092:1861317 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1861093:1861320 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1861092:1861317 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1861093:1861320 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1861092:1861317 [0] NCCL INFO ========================================== n136-128-154:1861093:1861320 [1] NCCL INFO ========================================== n136-128-154:1861092:1861317 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861093:1861320 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861092:1861317 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861093:1861320 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861092:1861317 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1861093:1861320 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1861093:1861320 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1861093:1861320 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1861093:1861320 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1861093:1861320 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1861093:1861320 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861093:1861320 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861092:1861317 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861092:1861317 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861093:1861320 [1] NCCL INFO comm 0x10ab4400 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1861092:1861317 [0] NCCL INFO comm 0x106c2ac0 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861093:1861320 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1861093:1861320 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1861093:1861320 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1861093:1861320 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1861092:1861317 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1861092:1861317 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1861093:1861320 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1861093:1861331 [1] NCCL INFO [Proxy Service] Device 1 CPU core 117 n136-128-154:1861093:1861332 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 54 n136-128-154:1861092:1861317 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1861092:1861317 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1861092:1861333 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:1861092:1861334 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 35 n136-128-154:1861093:1861320 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1861093:1861320 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1861092:1861317 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1861092:1861317 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1861092:1861317 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1861093:1861320 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1861093:1861320 [1] NCCL INFO ncclCommInitRankConfig comm 0x10ab4400 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xd0b62efd4b604bc3 - Init COMPLETE n136-128-154:1861093:1861320 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.55 (kernels 0.19, alloc 0.18, bootstrap 0.00, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.11) n136-128-154:1861092:1861317 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1861092:1861317 [0] NCCL INFO ncclCommInitRankConfig comm 0x106c2ac0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xd0b62efd4b604bc3 - Init COMPLETE n136-128-154:1861092:1861317 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.84 (kernels 0.21, alloc 0.32, bootstrap 0.13, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.11) n136-128-154:1861093:1861335 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861335 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861092:1861336 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1861093:1861335 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-07:00:42:36 INFO [evaluator:559] Running loglikelihood requests 2025-12-07:00:42:36 INFO [evaluator:559] Running loglikelihood requests Running loglikelihood requests: 0%| | 0/2344 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861093:1861588 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861093:1861331 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:00:46:24 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 25, batch_size: auto (45) | Tasks |Version|Filter|n-shot| Metric | |Value | |Stderr| |-------------|------:|------|-----:|--------|---|-----:|---|-----:| |arc_challenge| 1|none | 25|acc |↑ |0.3959|± |0.0143| | | |none | 25|acc_norm|↑ |0.4283|± |0.0145| [rank0]:[W1207 00:46:25.993834409 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861092:1861333 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861093:1861331 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861092:1861643 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861093:1861593 [1] NCCL INFO comm 0x10ab4400 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1861092:1861643 [0] NCCL INFO comm 0x106c2ac0 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 arc_challenge 评估完成! ================================================== 开始评估:任务=truthfulqa_mc1 | 少样本数=0 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/truthfulqa_mc1.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-07:00:47:48 INFO [__main__:440] Selected Tasks: ['truthfulqa_mc1'] 2025-12-07:00:47:48 INFO [__main__:440] Selected Tasks: ['truthfulqa_mc1'] 2025-12-07:00:47:48 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:47:48 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:47:48 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:47:48 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:47:49 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:47:50 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:47:50 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} `torch_dtype` is deprecated! Use `dtype` instead! `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1861701:1861893 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1861701:1861893 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1861701:1861893 [0] NCCL INFO Using network IB n136-128-154:1861701:1861893 [0] NCCL INFO ncclCommInitRankConfig comm 0xf5433f0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x405acd16b64ca290 - Init START 2025-12-07:00:48:17 INFO [evaluator:305] num_fewshot has been set to 0 for truthfulqa_mc1 in its config. Manual configuration will be ignored. 2025-12-07:00:48:17 INFO [api.task:434] Building contexts for truthfulqa_mc1 on rank 1... 0%| | 0/408 [00:00 n136-128-154:1861702:1861908 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1861702:1861908 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1861702:1861908 [1] NCCL INFO Using network IB n136-128-154:1861702:1861908 [1] NCCL INFO ncclCommInitRankConfig comm 0xf847890 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x405acd16b64ca290 - Init START n136-128-154:1861702:1861908 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1861701:1861893 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1861701:1861893 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1861701:1861893 [0] NCCL INFO Retrieving state for IB n136-128-154:1861701:1861893 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1861701:1861893 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1861702:1861908 [1] NCCL INFO Retrieving state for IB n136-128-154:1861702:1861908 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1861701:1861893 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1861701:1861893 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1861701:1861893 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1861702:1861908 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861702:1861908 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861701:1861893 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861701:1861893 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1861701:1861893 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1861701:1861893 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1861701:1861893 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1861701:1861893 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1861701:1861893 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1861701:1861893 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1861701:1861893 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1861701:1861893 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1861701:1861893 [0] NCCL INFO ========================================== n136-128-154:1861701:1861893 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861701:1861893 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861701:1861893 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1861702:1861908 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1861702:1861908 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1861702:1861908 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1861702:1861908 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1861702:1861908 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1861702:1861908 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1861702:1861908 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1861702:1861908 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1861702:1861908 [1] NCCL INFO ========================================== n136-128-154:1861702:1861908 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861702:1861908 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1861702:1861908 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1861701:1861893 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1861701:1861893 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1861701:1861893 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861701:1861893 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861701:1861893 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861701:1861893 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861701:1861893 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861701:1861893 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861701:1861893 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1861702:1861908 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1861702:1861908 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1861702:1861908 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1861702:1861908 [1] NCCL INFO comm 0xf847890 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1861701:1861893 [0] NCCL INFO comm 0xf5433f0 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861702:1861908 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1861702:1861908 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1861702:1861908 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1861702:1861908 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1861701:1861893 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1861701:1861893 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1861701:1861893 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1861701:1861893 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1861701:1861915 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:1861701:1861916 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 38 n136-128-154:1861702:1861908 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1861702:1861917 [1] NCCL INFO [Proxy Service] Device 1 CPU core 103 n136-128-154:1861702:1861918 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 104 n136-128-154:1861701:1861893 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1861701:1861893 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1861701:1861893 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1861702:1861908 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1861702:1861908 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1861701:1861893 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1861701:1861893 [0] NCCL INFO ncclCommInitRankConfig comm 0xf5433f0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x405acd16b64ca290 - Init COMPLETE n136-128-154:1861701:1861893 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 7.54 (kernels 0.20, alloc 0.19, bootstrap 7.00, allgathers 0.00, topo 0.07, graphs 0.00, connections 0.01, rest 0.07) n136-128-154:1861702:1861908 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1861702:1861908 [1] NCCL INFO ncclCommInitRankConfig comm 0xf847890 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x405acd16b64ca290 - Init COMPLETE n136-128-154:1861702:1861908 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.46 (kernels 0.12, alloc 0.18, bootstrap 0.00, allgathers 0.00, topo 0.07, graphs 0.00, connections 0.01, rest 0.07) n136-128-154:1861701:1861919 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861920 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861701:1861919 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1861702:1861920 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-07:00:48:18 INFO [evaluator:559] Running loglikelihood requests 2025-12-07:00:48:18 INFO [evaluator:559] Running loglikelihood requests Running loglikelihood requests: 0%| | 0/2066 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1861702:1861962 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861702:1861917 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:00:49:10 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 0, batch_size: auto (64) | Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr| |--------------|------:|------|-----:|------|---|-----:|---|-----:| |truthfulqa_mc1| 2|none | 0|acc |↑ |0.2974|± | 0.016| [rank0]:[W1207 00:49:11.872260977 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861701:1861915 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1861702:1861917 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1861701:1862020 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1861702:1861968 [1] NCCL INFO comm 0xf847890 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1861701:1862020 [0] NCCL INFO comm 0xf5433f0 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 truthfulqa_mc1 评估完成! ================================================== 开始评估:任务=piqa | 少样本数=0 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/piqa.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-07:00:50:37 INFO [__main__:440] Selected Tasks: ['piqa'] 2025-12-07:00:50:37 INFO [__main__:440] Selected Tasks: ['piqa'] 2025-12-07:00:50:37 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:50:37 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:50:37 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:50:37 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:50:38 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:50:38 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} `torch_dtype` is deprecated! Use `dtype` instead! The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:50:38 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1862074:1862295 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1862074:1862295 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1862074:1862295 [1] NCCL INFO Using network IB n136-128-154:1862073:1862294 [0] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:1862073:1862294 [0] NCCL INFO NET/IB : Using [0]mlx5_0:1/RoCE [1]mlx5_1:1/RoCE [2]mlx5_2:1/RoCE [3]mlx5_3:1/RoCE [RO]; OOB eth0:fdbd:dc03:9:451::154<0> n136-128-154:1862073:1862294 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1862073:1862294 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1862073:1862294 [0] NCCL INFO Using network IB n136-128-154:1862074:1862295 [1] NCCL INFO ncclCommInitRankConfig comm 0x10ebbab0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x632767f26687259c - Init START n136-128-154:1862073:1862294 [0] NCCL INFO ncclCommInitRankConfig comm 0x100cb200 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x632767f26687259c - Init START n136-128-154:1862074:1862295 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1862073:1862294 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1862073:1862294 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1862073:1862294 [0] NCCL INFO Retrieving state for IB n136-128-154:1862073:1862294 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1862073:1862294 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1862074:1862295 [1] NCCL INFO Retrieving state for IB n136-128-154:1862074:1862295 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1862073:1862294 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1862073:1862294 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1862073:1862294 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1862074:1862295 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862074:1862295 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862073:1862294 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862073:1862294 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862074:1862295 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1862074:1862295 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1862074:1862295 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1862074:1862295 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1862074:1862295 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1862074:1862295 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1862074:1862295 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1862074:1862295 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1862074:1862295 [1] NCCL INFO ========================================== n136-128-154:1862074:1862295 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862074:1862295 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862074:1862295 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1862073:1862294 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1862073:1862294 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1862073:1862294 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1862073:1862294 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1862073:1862294 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1862073:1862294 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1862073:1862294 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1862073:1862294 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1862073:1862294 [0] NCCL INFO ========================================== n136-128-154:1862073:1862294 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862073:1862294 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862073:1862294 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1862074:1862295 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1862074:1862295 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1862074:1862295 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1862074:1862295 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862074:1862295 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1862074:1862295 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862074:1862295 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862073:1862294 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862073:1862294 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862074:1862295 [1] NCCL INFO comm 0x10ebbab0 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1862073:1862294 [0] NCCL INFO comm 0x100cb200 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862074:1862295 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1862073:1862294 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1862074:1862295 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1862074:1862295 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1862074:1862295 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1862073:1862294 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1862073:1862294 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1862074:1862295 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1862073:1862294 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1862073:1862294 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1862073:1862307 [0] NCCL INFO [Proxy Service] Device 0 CPU core 38 n136-128-154:1862074:1862306 [1] NCCL INFO [Proxy Service] Device 1 CPU core 34 n136-128-154:1862073:1862308 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 103 n136-128-154:1862074:1862309 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 40 n136-128-154:1862074:1862295 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1862074:1862295 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1862073:1862294 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1862073:1862294 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1862073:1862294 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1862074:1862295 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1862074:1862295 [1] NCCL INFO ncclCommInitRankConfig comm 0x10ebbab0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x632767f26687259c - Init COMPLETE n136-128-154:1862074:1862295 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.64 (kernels 0.26, alloc 0.17, bootstrap 0.02, allgathers 0.00, topo 0.08, graphs 0.00, connections 0.01, rest 0.09) n136-128-154:1862073:1862294 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1862073:1862294 [0] NCCL INFO ncclCommInitRankConfig comm 0x100cb200 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x632767f26687259c - Init COMPLETE n136-128-154:1862073:1862294 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.64 (kernels 0.26, alloc 0.19, bootstrap 0.00, allgathers 0.00, topo 0.08, graphs 0.00, connections 0.01, rest 0.09) n136-128-154:1862074:1862310 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862074:1862310 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862073:1862311 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1862074:1862310 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-07:00:50:56 INFO [evaluator:559] Running loglikelihood requests 2025-12-07:00:50:56 INFO [evaluator:559] Running loglikelihood requests Passed argument batch_size = auto:1. Detecting largest batch size Running loglikelihood requests: 0%| | 0/1838 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862074:1862332 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862074:1862306 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:00:51:26 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 0, batch_size: auto (64) |Tasks|Version|Filter|n-shot| Metric | |Value | |Stderr| |-----|------:|------|-----:|--------|---|-----:|---|-----:| |piqa | 1|none | 0|acc |↑ |0.6774|± |0.0109| | | |none | 0|acc_norm|↑ |0.6817|± |0.0109| [rank0]:[W1207 00:51:27.186925054 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862073:1862307 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862073:1862392 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862074:1862306 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1862074:1862337 [1] NCCL INFO comm 0x10ebbab0 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1862073:1862392 [0] NCCL INFO comm 0x100cb200 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 piqa 评估完成! ================================================== 开始评估:任务=hellaswag | 少样本数=10 | 模型=Qwen2.5-7B-quantization-fg 输出路径:results2/Qwen2.5-7B-quantization-fg/fg2/hellaswag.json ================================================== The following values were not passed to `accelerate launch` and had defaults used instead: More than one GPU was found, enabling multi-GPU training. If this was unintended please pass in `--num_processes=1`. `--num_machines` was set to a value of `1` `--mixed_precision` was set to a value of `'no'` `--dynamo_backend` was set to a value of `'no'` To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`. 2025-12-07:00:52:52 INFO [__main__:440] Selected Tasks: ['hellaswag'] 2025-12-07:00:52:52 INFO [__main__:440] Selected Tasks: ['hellaswag'] 2025-12-07:00:52:52 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:52:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:52:52 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-07:00:52:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-07:00:52:53 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:52:53 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:1'} The tokenizer you are loading from '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue. 2025-12-07:00:52:53 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} `torch_dtype` is deprecated! Use `dtype` instead! `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:1862446:1862683 [0] NCCL INFO Initialized NET plugin IB n136-128-154:1862446:1862683 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1862446:1862683 [0] NCCL INFO Using network IB n136-128-154:1862447:1862684 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:1862447:1862684 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/RoCE [1]mlx5_1:1/RoCE [2]mlx5_2:1/RoCE [3]mlx5_3:1/RoCE [RO]; OOB eth0:fdbd:dc03:9:451::154<0> n136-128-154:1862447:1862684 [1] NCCL INFO Initialized NET plugin IB n136-128-154:1862447:1862684 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:1862447:1862684 [1] NCCL INFO Using network IB n136-128-154:1862446:1862683 [0] NCCL INFO ncclCommInitRankConfig comm 0xe17bf30 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xd9077272157001ae - Init START n136-128-154:1862447:1862684 [1] NCCL INFO ncclCommInitRankConfig comm 0xdd95b90 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xd9077272157001ae - Init START n136-128-154:1862446:1862683 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1862447:1862684 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:1862446:1862683 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1862446:1862683 [0] NCCL INFO Retrieving state for IB n136-128-154:1862446:1862683 [0] NCCL INFO Initialized state 0 for IB n136-128-154:1862446:1862683 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1862446:1862683 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1862446:1862683 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1862447:1862684 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:1862447:1862684 [1] NCCL INFO Retrieving state for IB n136-128-154:1862447:1862684 [1] NCCL INFO Initialized state 0 for IB n136-128-154:1862446:1862683 [0] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1862447:1862684 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_0 in topo with pciPath=/sys/devices/pci0000:09/0000:09:02.0/0000:0a:00.0/0000:0b:08.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0 keep=1 coll=(null) n136-128-154:1862447:1862684 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_1 in topo with pciPath=/sys/devices/pci0000:43/0000:43:02.0/0000:44:00.0/0000:45:08.0/0000:5e:00.0/0000:5f:00.0/0000:60:00.0 keep=1 coll=(null) n136-128-154:1862447:1862684 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_2 in topo with pciPath=/sys/devices/pci0000:82/0000:82:02.0/0000:83:00.0/0000:84:08.0/0000:93:00.0/0000:94:00.0/0000:95:00.0 keep=1 coll=(null) n136-128-154:1862447:1862684 [1] NCCL INFO ncclTopoPopulateNics : Filled mlx5_3 in topo with pciPath=/sys/devices/pci0000:be/0000:be:02.0/0000:bf:00.0/0000:c0:04.0/0000:ca:00.0/0000:cb:10.0/0000:cc:00.0 keep=1 coll=(null) n136-128-154:1862446:1862683 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862447:1862684 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862446:1862683 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862447:1862684 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:1862446:1862683 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1862446:1862683 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1862446:1862683 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1862446:1862683 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1862446:1862683 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1862446:1862683 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1862446:1862683 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1862446:1862683 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1862446:1862683 [0] NCCL INFO ========================================== n136-128-154:1862446:1862683 [0] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862446:1862683 [0] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862446:1862683 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:1862447:1862684 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:1862447:1862684 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:1862447:1862684 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:1862447:1862684 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:1862447:1862684 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:1862447:1862684 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:1862447:1862684 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:1862447:1862684 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:1862447:1862684 [1] NCCL INFO ========================================== n136-128-154:1862447:1862684 [1] NCCL INFO GPU/0-c5000 :GPU/0-c5000 (0/5000.0/LOC) GPU/0-c9000 (2/240.0/NVL) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862447:1862684 [1] NCCL INFO GPU/0-c9000 :GPU/0-c5000 (2/240.0/NVL) GPU/0-c9000 (0/5000.0/LOC) NVS/0-0 (1/240.0/NVL) CPU/0-1 (3/24.0/PHB) CPU/0-0 (4/10.0/SYS) n136-128-154:1862446:1862683 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1862447:1862684 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:1862446:1862683 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:1862447:1862684 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1862447:1862684 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862447:1862684 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:1862446:1862683 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:1862446:1862683 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862446:1862683 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:1862447:1862684 [1] NCCL INFO comm 0xdd95b90 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:1862446:1862683 [0] NCCL INFO comm 0xe17bf30 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862447:1862684 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:1862447:1862684 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:1862447:1862684 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0 [2] -1/-1/-1->1->0 [3] -1/-1/-1->1->0 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 0/-1/-1->1->-1 [7] 0/-1/-1->1->-1 [8] 0/-1/-1->1->-1 [9] 0/-1/-1->1->-1 [10] 0/-1/-1->1->-1 [11] 0/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] -1/-1/-1->1->0 [15] -1/-1/-1->1->0 [16] -1/-1/-1->1->0 [17] -1/-1/-1->1->0 [18] 0/-1/-1->1->-1 [19] 0/-1/-1->1->-1 [20] 0/-1/-1->1->-1 [21] 0/-1/-1->1->-1 [22] 0/-1/-1->1->-1 [23] 0/-1/-1->1->-1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:1862447:1862684 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:1862446:1862683 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] 1/-1/-1->0->-1 [3] 1/-1/-1->0->-1 [4] 1/-1/-1->0->-1 [5] 1/-1/-1->0->-1 [6] -1/-1/-1->0->1 [7] -1/-1/-1->0->1 [8] -1/-1/-1->0->1 [9] -1/-1/-1->0->1 [10] -1/-1/-1->0->1 [11] -1/-1/-1->0->1 [12] 1/-1/-1->0->-1 [13] 1/-1/-1->0->-1 [14] 1/-1/-1->0->-1 [15] 1/-1/-1->0->-1 [16] 1/-1/-1->0->-1 [17] 1/-1/-1->0->-1 [18] -1/-1/-1->0->1 [19] -1/-1/-1->0->1 [20] -1/-1/-1->0->1 [21] -1/-1/-1->0->1 [22] -1/-1/-1->0->1 [23] -1/-1/-1->0->1 n136-128-154:1862446:1862683 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:1862446:1862683 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1862446:1862683 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:1862446:1862697 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 108 n136-128-154:1862446:1862696 [0] NCCL INFO [Proxy Service] Device 0 CPU core 107 n136-128-154:1862447:1862684 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:1862447:1862698 [1] NCCL INFO [Proxy Service] Device 1 CPU core 34 n136-128-154:1862447:1862699 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 100 n136-128-154:1862446:1862683 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1862446:1862683 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1862447:1862684 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:1862447:1862684 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:1862446:1862683 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:1862447:1862684 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1862447:1862684 [1] NCCL INFO ncclCommInitRankConfig comm 0xdd95b90 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xd9077272157001ae - Init COMPLETE n136-128-154:1862447:1862684 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.86 (kernels 0.23, alloc 0.40, bootstrap 0.00, allgathers 0.01, topo 0.08, graphs 0.00, connections 0.02, rest 0.12) n136-128-154:1862446:1862683 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:1862446:1862683 [0] NCCL INFO ncclCommInitRankConfig comm 0xe17bf30 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xd9077272157001ae - Init COMPLETE n136-128-154:1862446:1862683 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.87 (kernels 0.23, alloc 0.40, bootstrap 0.01, allgathers 0.00, topo 0.08, graphs 0.00, connections 0.02, rest 0.12) n136-128-154:1862446:1862700 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862446:1862700 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1862701 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:1862446:1862700 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-07:00:53:44 INFO [evaluator:559] Running loglikelihood requests 2025-12-07:00:53:44 INFO [evaluator:559] Running loglikelihood requests Passed argument batch_size = auto:1. Detecting largest batch size Running loglikelihood requests: 0%| | 0/20084 [00:00 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:1862447:1865958 [1] NCCL INFO Channel 31/1 : 1[7] -> 0[6] via P2P/CUMEM/read fatal: detected dubious ownership in repository at '/mnt/bn/life-mllm/users/cxr/quantization' To add an exception for this directory, call: git config --global --add safe.directory /mnt/bn/life-mllm/users/cxr/quantization n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862447:1862698 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-07:01:17:28 INFO [loggers.evaluation_tracker:209] Saving results aggregated hf (pretrained=/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg), gen_kwargs: (None), limit: None, num_fewshot: 10, batch_size: auto (57) | Tasks |Version|Filter|n-shot| Metric | |Value | |Stderr| |---------|------:|------|-----:|--------|---|-----:|---|-----:| |hellaswag| 1|none | 10|acc |↑ |0.4351|± |0.0049| | | |none | 10|acc_norm|↑ |0.5978|± |0.0049| [rank0]:[W1207 01:17:29.546832091 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator()) n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862446:1862696 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1862447:1862698 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:1862446:1866014 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:1862447:1865964 [1] NCCL INFO comm 0xdd95b90 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:1862446:1866014 [0] NCCL INFO comm 0xe17bf30 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 hellaswag 评估完成! Execution time: 34699.1 seconds