nohup: ignoring input Namespace(save_dir='/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg', self_attn_layer_to_quant='27 16 19 17 25', 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 n136-128-154:2195229:2195601 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2195229:2195601 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2195229:2195601 [0] NCCL INFO Using network IB n136-128-154:2195229:2195601 [0] NCCL INFO ncclCommInitRankConfig comm 0x11614ad0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xe72ac37abd19e19 - Init START 2025-12-09:14:52:28 WARNING [evaluator:309] Overwriting default num_fewshot of winogrande from None to 5 2025-12-09:14:52:28 INFO [api.task:434] Building contexts for winogrande on rank 1... 0%| | 0/633 [00:00 n136-128-154:2195230:2195633 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2195230:2195633 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2195230:2195633 [1] NCCL INFO Using network IB n136-128-154:2195230:2195633 [1] NCCL INFO ncclCommInitRankConfig comm 0x113f3bf0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xe72ac37abd19e19 - Init START n136-128-154:2195230:2195633 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2195229:2195601 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2195230:2195633 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2195230:2195633 [1] NCCL INFO Retrieving state for IB n136-128-154:2195230:2195633 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2195230:2195633 [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) n136-128-154:2195229:2195601 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2195229:2195601 [0] NCCL INFO Retrieving state for IB n136-128-154:2195229:2195601 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2195230:2195633 [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) n136-128-154:2195229:2195601 [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) n136-128-154:2195230:2195633 [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) n136-128-154:2195229:2195601 [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) n136-128-154:2195230:2195633 [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) n136-128-154:2195229:2195601 [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) n136-128-154:2195229:2195601 [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) n136-128-154:2195230:2195633 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195230:2195633 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195229:2195601 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195229:2195601 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195229:2195601 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2195229:2195601 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2195229:2195601 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2195229:2195601 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2195229:2195601 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2195229:2195601 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2195229:2195601 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2195229:2195601 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2195229:2195601 [0] NCCL INFO ========================================== n136-128-154:2195229:2195601 [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) n136-128-154:2195229:2195601 [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) n136-128-154:2195229:2195601 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2195230:2195633 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2195230:2195633 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2195230:2195633 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2195230:2195633 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2195230:2195633 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2195230:2195633 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2195230:2195633 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2195230:2195633 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2195230:2195633 [1] NCCL INFO ========================================== n136-128-154:2195230:2195633 [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) n136-128-154:2195230:2195633 [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) n136-128-154:2195230:2195633 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2195229:2195601 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2195229:2195601 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2195230:2195633 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2195230:2195633 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195229:2195601 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195229:2195601 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195229:2195601 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195229:2195601 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195229:2195601 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195229:2195601 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2195230:2195633 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195230:2195633 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195230:2195633 [1] NCCL INFO comm 0x113f3bf0 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2195229:2195601 [0] NCCL INFO comm 0x11614ad0 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195230:2195633 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2195230:2195633 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2195230:2195633 [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 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2195230:2195633 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2195229:2195601 [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 n136-128-154:2195229:2195601 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2195230:2195633 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2195230:2195640 [1] NCCL INFO [Proxy Service] Device 1 CPU core 102 n136-128-154:2195230:2195641 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 101 n136-128-154:2195229:2195601 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2195229:2195601 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2195229:2195642 [0] NCCL INFO [Proxy Service] Device 0 CPU core 34 n136-128-154:2195229:2195643 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 103 n136-128-154:2195230:2195633 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2195230:2195633 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2195229:2195601 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2195229:2195601 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2195229:2195601 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2195230:2195633 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2195230:2195633 [1] NCCL INFO ncclCommInitRankConfig comm 0x113f3bf0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xe72ac37abd19e19 - Init COMPLETE n136-128-154:2195230:2195633 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.42 (kernels 0.18, alloc 0.13, bootstrap 0.00, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:2195229:2195601 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2195229:2195601 [0] NCCL INFO ncclCommInitRankConfig comm 0x11614ad0 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xe72ac37abd19e19 - Init COMPLETE n136-128-154:2195229:2195601 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 7.75 (kernels 0.17, alloc 0.13, bootstrap 7.35, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:2195230:2195644 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195229:2195645 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195230:2195644 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2195229:2195645 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:14:52:29 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:14:52:29 INFO [evaluator:559] Running loglikelihood requests Running loglikelihood requests: 0%| | 0/1268 [00:00 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195230:2195672 [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 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195230:2195640 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:14:53:02 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 |↑ |0.5399|± | 0.014| [rank0]:[W1209 14:53:03.125410652 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()) n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195229:2195642 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195229:2195763 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195230:2195640 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2195230:2195684 [1] NCCL INFO comm 0x113f3bf0 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2195229:2195763 [0] NCCL INFO comm 0x11614ad0 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 winogrande 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/fg5/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`. 2025-12-09:14:54:30 INFO [__main__:440] Selected Tasks: ['boolq'] 2025-12-09:14:54:30 INFO [__main__:440] Selected Tasks: ['boolq'] 2025-12-09:14:54:30 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 2025-12-09:14:54:30 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:14:54:30 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 2025-12-09:14:54:30 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:14:54:31 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. 2025-12-09:14:54:32 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. 2025-12-09:14:54:32 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 n136-128-154:2195818:2196077 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2195818:2196077 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2195818:2196077 [1] NCCL INFO Using network IB n136-128-154:2195817:2196076 [0] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:2195817:2196076 [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> n136-128-154:2195817:2196076 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2195817:2196076 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2195817:2196076 [0] NCCL INFO Using network IB n136-128-154:2195818:2196077 [1] NCCL INFO ncclCommInitRankConfig comm 0x10588480 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xcf96f8fae45e0dea - Init START n136-128-154:2195817:2196076 [0] NCCL INFO ncclCommInitRankConfig comm 0xf5fa020 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xcf96f8fae45e0dea - Init START n136-128-154:2195817:2196076 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2195818:2196077 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2195818:2196077 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2195818:2196077 [1] NCCL INFO Retrieving state for IB n136-128-154:2195818:2196077 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2195818:2196077 [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) n136-128-154:2195817:2196076 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2195817:2196076 [0] NCCL INFO Retrieving state for IB n136-128-154:2195817:2196076 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2195818:2196077 [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) n136-128-154:2195817:2196076 [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) n136-128-154:2195818:2196077 [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) n136-128-154:2195817:2196076 [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) n136-128-154:2195818:2196077 [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) n136-128-154:2195817:2196076 [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) n136-128-154:2195817:2196076 [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) n136-128-154:2195818:2196077 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195817:2196076 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195818:2196077 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195817:2196076 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2195817:2196076 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2195817:2196076 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2195817:2196076 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2195817:2196076 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2195817:2196076 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2195817:2196076 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2195817:2196076 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2195817:2196076 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2195817:2196076 [0] NCCL INFO ========================================== n136-128-154:2195817:2196076 [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) n136-128-154:2195817:2196076 [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) n136-128-154:2195817:2196076 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2195818:2196077 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2195818:2196077 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2195818:2196077 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2195818:2196077 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2195818:2196077 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2195818:2196077 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2195818:2196077 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2195818:2196077 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2195818:2196077 [1] NCCL INFO ========================================== n136-128-154:2195818:2196077 [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) n136-128-154:2195818:2196077 [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) n136-128-154:2195818:2196077 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2195817:2196076 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2195817:2196076 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2195817:2196076 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195817:2196076 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195817:2196076 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195817:2196076 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195817:2196076 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195817:2196076 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195817:2196076 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2195818:2196077 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2195818:2196077 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2195818:2196077 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2195818:2196077 [1] NCCL INFO comm 0x10588480 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO comm 0xf5fa020 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195817:2196076 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2195818:2196077 [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 n136-128-154:2195818:2196077 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2195817:2196076 [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 n136-128-154:2195817:2196076 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2195818:2196077 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2195818:2196088 [1] NCCL INFO [Proxy Service] Device 1 CPU core 36 n136-128-154:2195818:2196089 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 101 n136-128-154:2195817:2196076 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2195817:2196076 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2195817:2196090 [0] NCCL INFO [Proxy Service] Device 0 CPU core 102 n136-128-154:2195817:2196091 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 103 n136-128-154:2195818:2196077 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2195818:2196077 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2195817:2196076 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2195817:2196076 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2195817:2196076 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2195818:2196077 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2195818:2196077 [1] NCCL INFO ncclCommInitRankConfig comm 0x10588480 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xcf96f8fae45e0dea - Init COMPLETE n136-128-154:2195818:2196077 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.89 (kernels 0.24, alloc 0.40, bootstrap 0.00, allgathers 0.01, topo 0.06, graphs 0.00, connections 0.06, rest 0.12) n136-128-154:2195817:2196076 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2195817:2196076 [0] NCCL INFO ncclCommInitRankConfig comm 0xf5fa020 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xcf96f8fae45e0dea - Init COMPLETE n136-128-154:2195817:2196076 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.89 (kernels 0.25, alloc 0.40, bootstrap 0.00, allgathers 0.01, topo 0.06, graphs 0.00, connections 0.06, rest 0.12) n136-128-154:2195818:2196093 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195818:2196093 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2195817:2196094 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2195818:2196093 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:14:55:02 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:14:55:02 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 n136-128-154:2195818:2196244 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2195818:2196244 [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 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195818:2196088 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:14:56:32 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 |↑ |0.6783|± |0.0082| [rank0]:[W1209 14:56:33.756217571 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()) n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195817:2196090 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2195817:2196313 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2195818:2196088 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2195818:2196249 [1] NCCL INFO comm 0x10588480 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2195817:2196313 [0] NCCL INFO comm 0xf5fa020 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/fg5/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`. 2025-12-09:14:57:52 INFO [__main__:440] Selected Tasks: ['arc_challenge'] 2025-12-09:14:57:52 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 2025-12-09:14:57:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:14:57:52 INFO [__main__:440] Selected Tasks: ['arc_challenge'] 2025-12-09:14:57:52 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 2025-12-09:14:57:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:14:57:53 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. 2025-12-09:14:57:53 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 n136-128-154:2196430:2196729 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2196430:2196729 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2196430:2196729 [0] NCCL INFO Using network IB 76%|███████▋ | 448/586 [00:12<00:03, 36.30it/s]n136-128-154:2196430:2196729 [0] NCCL INFO ncclCommInitRankConfig comm 0x12b70520 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xf75dd042d9e31e3b - Init START 77%|███████▋ | 452/586 [00:12<00:03, 36.28it/s] 78%|███████▊ | 456/586 [00:12<00:03, 36.27it/s] 78%|███████▊ | 460/586 [00:12<00:03, 35.89it/s] 79%|███████▉ | 464/586 [00:12<00:03, 35.89it/s] 80%|███████▉ | 468/586 [00:13<00:03, 35.98it/s] 81%|████████ | 472/586 [00:13<00:03, 36.02it/s] 81%|████████ | 476/586 [00:13<00:03, 36.11it/s] 82%|████████▏ | 480/586 [00:13<00:02, 36.13it/s] 83%|████████▎ | 484/586 [00:13<00:02, 36.15it/s] 83%|████████▎ | 488/586 [00:13<00:02, 36.17it/s] 84%|████████▍ | 492/586 [00:13<00:02, 36.19it/s] 85%|████████▍ | 496/586 [00:13<00:02, 35.80it/s] 85%|████████▌ | 500/586 [00:13<00:02, 35.86it/s] 86%|████████▌ | 504/586 [00:14<00:02, 35.91it/s] 87%|████████▋ | 508/586 [00:14<00:02, 35.94it/s] 87%|████████▋ | 512/586 [00:14<00:02, 36.06it/s] 88%|████████▊ | 516/586 [00:14<00:01, 36.09it/s] 89%|████████▊ | 520/586 [00:14<00:01, 36.11it/s] 89%|████████▉ | 524/586 [00:14<00:01, 36.09it/s] 90%|█████████ | 528/586 [00:14<00:01, 36.09it/s] 91%|█████████ | 532/586 [00:14<00:01, 36.01it/s] 91%|█████████▏| 536/586 [00:14<00:01, 36.06it/s] 92%|█████████▏| 540/586 [00:15<00:01, 36.04it/s] 93%|█████████▎| 544/586 [00:15<00:01, 35.93it/s] 94%|█████████▎| 548/586 [00:15<00:01, 35.88it/s] 94%|█████████▍| 552/586 [00:15<00:00, 35.73it/s] 95%|█████████▍| 556/586 [00:15<00:00, 35.77it/s] 96%|█████████▌| 560/586 [00:15<00:00, 35.87it/s] 96%|█████████▌| 564/586 [00:15<00:00, 35.97it/s] 97%|█████████▋| 568/586 [00:15<00:00, 36.02it/s] 98%|█████████▊| 572/586 [00:15<00:00, 35.71it/s] 98%|█████████▊| 576/586 [00:16<00:00, 35.79it/s] 99%|█████████▉| 580/586 [00:16<00:00, 35.91it/s] 100%|█████████▉| 584/586 [00:16<00:00, 36.04it/s] 100%|██████████| 586/586 [00:16<00:00, 35.89it/s] n136-128-154:2196431:2196431 [1] NCCL INFO cudaDriverVersion 12040 n136-128-154:2196431:2196431 [1] NCCL INFO NCCL_SOCKET_FAMILY set by environment to AF_INET6 n136-128-154:2196431:2196431 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to =eth0 n136-128-154:2196431:2196431 [1] NCCL INFO NCCL version 2.27.5+cuda12.9 n136-128-154:2196431:2196431 [1] NCCL INFO Comm config Blocking set to 1 n136-128-154:2196431:2196737 [1] NCCL INFO NET/Plugin: Could not find: none libnccl-net-none.so. n136-128-154:2196431:2196737 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0. n136-128-154:2196431:2196737 [1] NCCL INFO NCCL_SOCKET_FAMILY set by environment to AF_INET6 n136-128-154:2196431:2196737 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to =eth0 n136-128-154:2196431:2196737 [1] NCCL INFO NCCL_IB_HCA set to mlx5 n136-128-154:2196431:2196737 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:2196431:2196737 [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> n136-128-154:2196431:2196737 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2196431:2196737 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2196431:2196737 [1] NCCL INFO Using network IB n136-128-154:2196431:2196737 [1] NCCL INFO ncclCommInitRankConfig comm 0x12d752a0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xf75dd042d9e31e3b - Init START n136-128-154:2196431:2196737 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2196430:2196729 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2196431:2196737 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2196431:2196737 [1] NCCL INFO Retrieving state for IB n136-128-154:2196431:2196737 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2196431:2196737 [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) n136-128-154:2196431:2196737 [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) n136-128-154:2196430:2196729 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2196430:2196729 [0] NCCL INFO Retrieving state for IB n136-128-154:2196430:2196729 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2196431:2196737 [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) n136-128-154:2196430:2196729 [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) n136-128-154:2196431:2196737 [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) n136-128-154:2196430:2196729 [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) n136-128-154:2196430:2196729 [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) n136-128-154:2196430:2196729 [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) n136-128-154:2196431:2196737 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2196431:2196737 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2196430:2196729 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2196430:2196729 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2196430:2196729 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2196430:2196729 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2196430:2196729 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2196430:2196729 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2196430:2196729 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2196430:2196729 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2196430:2196729 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2196430:2196729 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2196430:2196729 [0] NCCL INFO ========================================== n136-128-154:2196430:2196729 [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) n136-128-154:2196430:2196729 [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) n136-128-154:2196430:2196729 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2196431:2196737 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2196431:2196737 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2196431:2196737 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2196431:2196737 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2196431:2196737 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2196431:2196737 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2196431:2196737 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2196431:2196737 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2196431:2196737 [1] NCCL INFO ========================================== n136-128-154:2196431:2196737 [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) n136-128-154:2196431:2196737 [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) n136-128-154:2196431:2196737 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2196430:2196729 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2196430:2196729 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2196431:2196737 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2196430:2196729 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196430:2196729 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196430:2196729 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196430:2196729 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196430:2196729 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196430:2196729 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196430:2196729 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2196431:2196737 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2196431:2196737 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2196431:2196737 [1] NCCL INFO comm 0x12d752a0 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2196430:2196729 [0] NCCL INFO comm 0x12b70520 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196430:2196729 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2196431:2196737 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2196430:2196729 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2196431:2196737 [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 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2196431:2196737 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2196430:2196729 [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 n136-128-154:2196430:2196729 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2196431:2196737 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2196431:2196745 [1] NCCL INFO [Proxy Service] Device 1 CPU core 48 n136-128-154:2196431:2196746 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 47 n136-128-154:2196430:2196729 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2196430:2196729 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2196430:2196747 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:2196430:2196748 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 37 n136-128-154:2196431:2196737 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2196431:2196737 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2196430:2196729 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2196430:2196729 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2196430:2196729 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2196431:2196737 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2196431:2196737 [1] NCCL INFO ncclCommInitRankConfig comm 0x12d752a0 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xf75dd042d9e31e3b - Init COMPLETE n136-128-154:2196431:2196737 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.47 (kernels 0.12, alloc 0.16, bootstrap 0.00, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.01, rest 0.12) n136-128-154:2196430:2196729 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2196430:2196729 [0] NCCL INFO ncclCommInitRankConfig comm 0x12b70520 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xf75dd042d9e31e3b - Init COMPLETE n136-128-154:2196430:2196729 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 4.57 (kernels 0.13, alloc 0.16, bootstrap 4.09, allgathers 0.00, topo 0.06, graphs 0.00, connections 0.02, rest 0.11) n136-128-154:2196431:2196749 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196431:2196749 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2196430:2196750 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2196431:2196749 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:14:58:33 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:14:58:33 INFO [evaluator:559] Running loglikelihood requests Passed argument batch_size = auto:1. Detecting largest batch size Running loglikelihood requests: 0%| | 0/2344 [00:00 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2196431:2197004 [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 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2196431:2196745 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:15:02:23 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 |↑ |0.3677|± |0.0141| | | |none | 25|acc_norm|↑ |0.4155|± |0.0144| [rank0]:[W1209 15:02:23.092941084 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()) n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2196430:2196747 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2196431:2196745 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2196430:2197059 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2196431:2197009 [1] NCCL INFO comm 0x12d752a0 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2196430:2197059 [0] NCCL INFO comm 0x12b70520 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/fg5/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`. 2025-12-09:15:03:52 INFO [__main__:440] Selected Tasks: ['truthfulqa_mc1'] 2025-12-09:15:03:52 INFO [__main__:440] Selected Tasks: ['truthfulqa_mc1'] 2025-12-09:15:03:52 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 2025-12-09:15:03:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:03:52 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 2025-12-09:15:03:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:03:53 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. 2025-12-09:15:03:53 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. 2025-12-09:15:03:53 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 n136-128-154:2197115:2197344 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2197115:2197344 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2197115:2197344 [0] NCCL INFO Using network IB n136-128-154:2197116:2197345 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:2197116:2197345 [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> n136-128-154:2197116:2197345 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2197116:2197345 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2197116:2197345 [1] NCCL INFO Using network IB n136-128-154:2197115:2197344 [0] NCCL INFO ncclCommInitRankConfig comm 0x1029d480 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xa98f936c31921541 - Init START n136-128-154:2197116:2197345 [1] NCCL INFO ncclCommInitRankConfig comm 0xea51340 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xa98f936c31921541 - Init START n136-128-154:2197115:2197344 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2197116:2197345 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2197116:2197345 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2197116:2197345 [1] NCCL INFO Retrieving state for IB n136-128-154:2197116:2197345 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2197116:2197345 [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) n136-128-154:2197115:2197344 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2197116:2197345 [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) n136-128-154:2197116:2197345 [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) n136-128-154:2197116:2197345 [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) n136-128-154:2197115:2197344 [0] NCCL INFO Retrieving state for IB n136-128-154:2197115:2197344 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197116:2197345 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197115:2197344 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197116:2197345 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197115:2197344 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2197115:2197344 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2197115:2197344 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2197115:2197344 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2197115:2197344 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2197115:2197344 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2197115:2197344 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2197115:2197344 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2197115:2197344 [0] NCCL INFO ========================================== n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [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) n136-128-154:2197115:2197344 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2197116:2197345 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2197116:2197345 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2197116:2197345 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2197116:2197345 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2197116:2197345 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2197116:2197345 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2197116:2197345 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2197116:2197345 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2197116:2197345 [1] NCCL INFO ========================================== n136-128-154:2197116:2197345 [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) n136-128-154:2197116:2197345 [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) n136-128-154:2197116:2197345 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2197115:2197344 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2197115:2197344 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2197115:2197344 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197115:2197344 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2197116:2197345 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2197116:2197345 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197116:2197345 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197116:2197345 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197115:2197344 [0] NCCL INFO comm 0x1029d480 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2197116:2197345 [1] NCCL INFO comm 0xea51340 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197116:2197345 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2197116:2197345 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2197116:2197345 [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 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2197116:2197345 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2197115:2197344 [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 n136-128-154:2197115:2197344 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2197115:2197344 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2197115:2197344 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2197115:2197358 [0] NCCL INFO [Proxy Service] Device 0 CPU core 98 n136-128-154:2197115:2197359 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 35 n136-128-154:2197116:2197345 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2197116:2197360 [1] NCCL INFO [Proxy Service] Device 1 CPU core 115 n136-128-154:2197116:2197361 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 53 n136-128-154:2197115:2197344 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2197115:2197344 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2197115:2197344 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2197116:2197345 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2197116:2197345 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2197115:2197344 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2197115:2197344 [0] NCCL INFO ncclCommInitRankConfig comm 0x1029d480 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0xa98f936c31921541 - Init COMPLETE n136-128-154:2197115:2197344 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 0.86 (kernels 0.22, alloc 0.44, bootstrap 0.00, allgathers 0.01, topo 0.05, graphs 0.00, connections 0.01, rest 0.12) n136-128-154:2197116:2197345 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2197116:2197345 [1] NCCL INFO ncclCommInitRankConfig comm 0xea51340 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0xa98f936c31921541 - Init COMPLETE n136-128-154:2197116:2197345 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.85 (kernels 0.22, alloc 0.44, bootstrap 0.00, allgathers 0.00, topo 0.05, graphs 0.00, connections 0.01, rest 0.12) n136-128-154:2197115:2197362 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197363 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197115:2197362 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2197116:2197363 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:15:04:17 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:15:04:17 INFO [evaluator:559] Running loglikelihood requests Passed argument batch_size = auto:1. Detecting largest batch size Running loglikelihood requests: 0%| | 0/2066 [00:00 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197116:2197410 [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 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197116:2197360 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:15:05:10 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 |↑ |0.2827|± |0.0158| [rank0]:[W1209 15:05:11.708249790 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()) n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197115:2197358 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197116:2197360 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197115:2197467 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197116:2197415 [1] NCCL INFO comm 0xea51340 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2197115:2197467 [0] NCCL INFO comm 0x1029d480 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/fg5/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`. 2025-12-09:15:06:33 INFO [__main__:440] Selected Tasks: ['piqa'] 2025-12-09:15:06:33 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 2025-12-09:15:06:33 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:06:33 INFO [__main__:440] Selected Tasks: ['piqa'] 2025-12-09:15:06:33 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 2025-12-09:15:06:33 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:06:34 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. 2025-12-09:15:06:34 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! 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. 2025-12-09:15:06:34 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! Loading checkpoint shards: 0%| | 0/4 [00:00 n136-128-154:2197525:2197936 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2197525:2197936 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2197525:2197936 [0] NCCL INFO Using network IB n136-128-154:2197525:2197936 [0] NCCL INFO ncclCommInitRankConfig comm 0x10631150 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x2ac4a3924f989c1 - Init START 2025-12-09:15:06:59 WARNING [evaluator:309] Overwriting default num_fewshot of piqa from None to 0 2025-12-09:15:06:59 INFO [api.task:434] Building contexts for piqa on rank 1... 0%| | 0/919 [00:00 n136-128-154:2197526:2197979 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2197526:2197979 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2197526:2197979 [1] NCCL INFO Using network IB n136-128-154:2197526:2197979 [1] NCCL INFO ncclCommInitRankConfig comm 0xffc5770 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x2ac4a3924f989c1 - Init START n136-128-154:2197526:2197979 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2197525:2197936 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2197526:2197979 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2197526:2197979 [1] NCCL INFO Retrieving state for IB n136-128-154:2197526:2197979 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2197526:2197979 [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) n136-128-154:2197526:2197979 [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) n136-128-154:2197526:2197979 [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) n136-128-154:2197526:2197979 [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) n136-128-154:2197525:2197936 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2197525:2197936 [0] NCCL INFO Retrieving state for IB n136-128-154:2197525:2197936 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2197525:2197936 [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) n136-128-154:2197525:2197936 [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) n136-128-154:2197525:2197936 [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) n136-128-154:2197525:2197936 [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) n136-128-154:2197525:2197936 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197525:2197936 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197526:2197979 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197526:2197979 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2197525:2197936 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2197525:2197936 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2197526:2197979 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2197525:2197936 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197526:2197979 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2197525:2197936 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2197525:2197936 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2197525:2197936 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2197526:2197979 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2197526:2197979 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2197525:2197936 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2197525:2197936 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2197525:2197936 [0] NCCL INFO ========================================== n136-128-154:2197526:2197979 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2197526:2197979 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2197525:2197936 [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) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2197525:2197936 [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) n136-128-154:2197526:2197979 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2197526:2197979 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2197526:2197979 [1] NCCL INFO ========================================== n136-128-154:2197525:2197936 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2197526:2197979 [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) n136-128-154:2197526:2197979 [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) n136-128-154:2197526:2197979 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2197525:2197936 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2197525:2197936 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2197525:2197936 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2197526:2197979 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197526:2197979 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2197525:2197936 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2197525:2197936 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197525:2197936 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2197526:2197979 [1] NCCL INFO comm 0xffc5770 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2197525:2197936 [0] NCCL INFO comm 0x10631150 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197526:2197979 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2197526:2197979 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2197526:2197979 [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 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2197526:2197979 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2197525:2197936 [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 n136-128-154:2197525:2197936 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2197526:2197979 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2197526:2197986 [1] NCCL INFO [Proxy Service] Device 1 CPU core 38 n136-128-154:2197526:2197987 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 40 n136-128-154:2197525:2197936 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2197525:2197936 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2197525:2197988 [0] NCCL INFO [Proxy Service] Device 0 CPU core 100 n136-128-154:2197525:2197989 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 105 n136-128-154:2197525:2197936 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2197525:2197936 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2197525:2197936 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2197526:2197979 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2197526:2197979 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2197526:2197979 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2197526:2197979 [1] NCCL INFO ncclCommInitRankConfig comm 0xffc5770 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x2ac4a3924f989c1 - Init COMPLETE n136-128-154:2197526:2197979 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.37 (kernels 0.14, alloc 0.14, bootstrap 0.00, allgathers 0.00, topo 0.05, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:2197525:2197936 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2197525:2197936 [0] NCCL INFO ncclCommInitRankConfig comm 0x10631150 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x2ac4a3924f989c1 - Init COMPLETE n136-128-154:2197525:2197936 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 4.96 (kernels 0.17, alloc 0.11, bootstrap 4.58, allgathers 0.00, topo 0.05, graphs 0.00, connections 0.01, rest 0.03) n136-128-154:2197526:2197990 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197526:2197990 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2197525:2197991 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2197526:2197990 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:15:07:00 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:15:07:00 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 n136-128-154:2197526:2198033 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2197526:2198033 [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 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197526:2197986 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:15:07:30 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 |↑ |0.6752|± |0.0109| | | |none | 0|acc_norm|↑ |0.6567|± |0.0111| [rank0]:[W1209 15:07:30.153193927 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()) n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197525:2197988 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2197526:2197986 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2197525:2198091 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2197526:2198037 [1] NCCL INFO comm 0xffc5770 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2197525:2198091 [0] NCCL INFO comm 0x10631150 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/fg5/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`. 2025-12-09:15:08:56 INFO [__main__:440] Selected Tasks: ['hellaswag'] 2025-12-09:15:08:56 INFO [__main__:440] Selected Tasks: ['hellaswag'] 2025-12-09:15:08:56 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 2025-12-09:15:08:56 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:08:56 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 2025-12-09:15:08:56 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen2.5-7B-quantization-fg'} 2025-12-09:15:08:57 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. 2025-12-09:15:08:58 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. 2025-12-09:15:08:58 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 n136-128-154:2198157:2198515 [0] NCCL INFO Initialized NET plugin IB n136-128-154:2198157:2198515 [0] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2198157:2198515 [0] NCCL INFO Using network IB 94%|█████████▍| 4736/5021 [00:24<00:01, 186.32it/s]n136-128-154:2198157:2198515 [0] NCCL INFO ncclCommInitRankConfig comm 0x13b13560 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x8dc6d77cdbe28b76 - Init START 95%|█████████▍| 4755/5021 [00:24<00:01, 183.07it/s] 95%|█████████▌| 4774/5021 [00:24<00:01, 181.06it/s] 95%|█████████▌| 4793/5021 [00:25<00:01, 182.01it/s] 96%|█████████▌| 4812/5021 [00:25<00:01, 183.60it/s] 96%|█████████▌| 4832/5021 [00:25<00:01, 185.65it/s] 97%|█████████▋| 4851/5021 [00:25<00:00, 185.11it/s] 97%|█████████▋| 4871/5021 [00:25<00:00, 186.86it/s] 97%|█████████▋| 4891/5021 [00:25<00:00, 187.87it/s] 98%|█████████▊| 4910/5021 [00:25<00:00, 188.37it/s] 98%|█████████▊| 4929/5021 [00:25<00:00, 188.46it/s] 99%|█████████▊| 4948/5021 [00:25<00:00, 188.68it/s] 99%|█████████▉| 4968/5021 [00:25<00:00, 189.49it/s] 99%|█████████▉| 4987/5021 [00:26<00:00, 189.51it/s] 100%|█████████▉| 5006/5021 [00:26<00:00, 189.30it/s] 100%|██████████| 5021/5021 [00:26<00:00, 191.47it/s] n136-128-154:2198158:2198158 [1] NCCL INFO cudaDriverVersion 12040 n136-128-154:2198158:2198158 [1] NCCL INFO NCCL_SOCKET_FAMILY set by environment to AF_INET6 n136-128-154:2198158:2198158 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to =eth0 n136-128-154:2198158:2198158 [1] NCCL INFO NCCL version 2.27.5+cuda12.9 n136-128-154:2198158:2198158 [1] NCCL INFO Comm config Blocking set to 1 n136-128-154:2198158:2198521 [1] NCCL INFO NET/Plugin: Could not find: none libnccl-net-none.so. n136-128-154:2198158:2198521 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0. n136-128-154:2198158:2198521 [1] NCCL INFO NCCL_SOCKET_FAMILY set by environment to AF_INET6 n136-128-154:2198158:2198521 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to =eth0 n136-128-154:2198158:2198521 [1] NCCL INFO NCCL_IB_HCA set to mlx5 n136-128-154:2198158:2198521 [1] NCCL INFO NCCL_IB_PCI_RELAXED_ORDERING set by environment to 1. n136-128-154:2198158:2198521 [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> n136-128-154:2198158:2198521 [1] NCCL INFO Initialized NET plugin IB n136-128-154:2198158:2198521 [1] NCCL INFO Assigned NET plugin IB to comm n136-128-154:2198158:2198521 [1] NCCL INFO Using network IB n136-128-154:2198158:2198521 [1] NCCL INFO ncclCommInitRankConfig comm 0xf821860 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x8dc6d77cdbe28b76 - Init START n136-128-154:2198158:2198521 [1] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2198157:2198515 [0] NCCL INFO RAS client listening socket at ::1<28028> n136-128-154:2198157:2198515 [0] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2198157:2198515 [0] NCCL INFO Retrieving state for IB n136-128-154:2198157:2198515 [0] NCCL INFO Initialized state 0 for IB n136-128-154:2198157:2198515 [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) n136-128-154:2198158:2198521 [1] NCCL INFO TOPO/NET : Importing network plugins to topology n136-128-154:2198158:2198521 [1] NCCL INFO Retrieving state for IB n136-128-154:2198158:2198521 [1] NCCL INFO Initialized state 0 for IB n136-128-154:2198157:2198515 [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) n136-128-154:2198158:2198521 [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) n136-128-154:2198158:2198521 [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) n136-128-154:2198157:2198515 [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) n136-128-154:2198157:2198515 [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) n136-128-154:2198158:2198521 [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) n136-128-154:2198158:2198521 [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) n136-128-154:2198158:2198521 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2198158:2198521 [1] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2198157:2198515 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 0 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2198157:2198515 [0] NCCL INFO GPU Direct RDMA Enabled for GPU 1 / HCA 3 (distance 5 <= 5), read 0 mode Default n136-128-154:2198157:2198515 [0] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2198157:2198515 [0] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2198157:2198515 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2198157:2198515 [0] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2198157:2198515 [0] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2198157:2198515 [0] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2198157:2198515 [0] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2198157:2198515 [0] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2198157:2198515 [0] NCCL INFO ========================================== n136-128-154:2198157:2198515 [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) n136-128-154:2198157:2198515 [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) n136-128-154:2198157:2198515 [0] NCCL INFO Setting affinity for GPU 6 to 33-62,97-126 n136-128-154:2198158:2198521 [1] NCCL INFO === System : maxBw 240.0 totalBw 240.0 === n136-128-154:2198158:2198521 [1] NCCL INFO CPU/0-1 (1/1/2) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-83000 (1000c0101000ffff) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - NIC/0-95000 n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-bf000 (1000c0101000ffff) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-c3000 (1000c01010de13b8) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - GPU/0-c5000 (0) n136-128-154:2198158:2198521 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-c7000 (1000c01010de13b8) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - GPU/0-c9000 (1) n136-128-154:2198158:2198521 [1] NCCL INFO + NVL[240.0] - NVS/0-0 n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - NIC/0-cc000 n136-128-154:2198158:2198521 [1] NCCL INFO + SYS[10.0] - CPU/0-0 n136-128-154:2198158:2198521 [1] NCCL INFO CPU/0-0 (1/1/2) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-a000 (1000c0101000ffff) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - NIC/0-1d000 n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - PCI/0-44000 (1000c0101000ffff) n136-128-154:2198158:2198521 [1] NCCL INFO + PCI[24.0] - NIC/0-60000 n136-128-154:2198158:2198521 [1] NCCL INFO + SYS[10.0] - CPU/0-1 n136-128-154:2198158:2198521 [1] NCCL INFO ========================================== n136-128-154:2198158:2198521 [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) n136-128-154:2198158:2198521 [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) n136-128-154:2198158:2198521 [1] NCCL INFO Setting affinity for GPU 7 to 33-62,97-126 n136-128-154:2198158:2198521 [1] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2198158:2198521 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO Pattern 4, crossNic 0, nChannels 12, bw 20.000000/20.000000, type NVL/PIX, sameChannels 1 n136-128-154:2198158:2198521 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 6 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 7 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 8 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 9 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 10 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 11 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2198158:2198521 [1] NCCL INFO Pattern 1, crossNic 0, nChannels 12, bw 40.000000/40.000000, type NVL/PIX, sameChannels 0 n136-128-154:2198157:2198515 [0] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 0 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 1 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 2 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 3 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 4 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198158:2198521 [1] NCCL INFO 5 : GPU/0-c5000 GPU/0-c9000 n136-128-154:2198157:2198515 [0] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 6 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 7 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 8 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 9 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 10 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198158:2198521 [1] NCCL INFO 11 : GPU/0-c9000 GPU/0-c5000 n136-128-154:2198157:2198515 [0] NCCL INFO comm 0x13b13560 rank 0 nRanks 2 nNodes 1 localRanks 2 localRank 0 MNNVL 0 n136-128-154:2198158:2198521 [1] NCCL INFO comm 0xf821860 rank 1 nRanks 2 nNodes 1 localRanks 2 localRank 1 MNNVL 0 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 0 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 12 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 0 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 1 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 12 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 13 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 1 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 2 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 13 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 14 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 2 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 3 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 14 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 15 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 3 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 4 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 15 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 16 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 4 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 5 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 16 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 17 : -1 -> 0 -> 1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 5 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 6 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 17 : 0 -> 1 -> -1/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 18 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 6 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 7 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 18 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 19 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 7 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 8 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 19 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 20 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 8 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 9 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 20 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 21 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 9 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 10 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 21 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 22 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 10 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 11 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 22 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Tree 23 : 1 -> 0 -> -1/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 11 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198158:2198521 [1] NCCL INFO Tree 23 : -1 -> 1 -> 0/-1/-1 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 00/24 : 0 1 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 01/24 : 0 1 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 02/24 : 0 1 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 03/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 00 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 04/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 01 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 05/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 02 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 06/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 03 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 07/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 04 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 08/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 05 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 09/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 06 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 10/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 07 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 11/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 08 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 12/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 09 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 13/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 10 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 14/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 11 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 15/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 12 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 16/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 13 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 17/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 14 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 18/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 15 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 19/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 16 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 20/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 17 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 21/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 18 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 22/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 19 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Channel 23/24 : 0 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 20 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 00 : 1 -> 0 -> 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 21 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 01 : 1 -> 0 -> 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 22 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 02 : 1 -> 0 -> 1 n136-128-154:2198158:2198521 [1] NCCL INFO Ring 23 : 0 -> 1 -> 0 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 03 : 1 -> 0 -> 1 n136-128-154:2198158:2198521 [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 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 04 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 05 : 1 -> 0 -> 1 n136-128-154:2198158:2198521 [1] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 06 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 07 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 08 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 09 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 10 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 11 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 12 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 13 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 14 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 15 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 16 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 17 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 18 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 19 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 20 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 21 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 22 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [0] NCCL INFO Ring 23 : 1 -> 0 -> 1 n136-128-154:2198157:2198515 [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 n136-128-154:2198157:2198515 [0] NCCL INFO P2P Chunksize set to 524288 n136-128-154:2198158:2198521 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2198158:2198528 [1] NCCL INFO [Proxy Service] Device 1 CPU core 101 n136-128-154:2198158:2198529 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 42 n136-128-154:2198157:2198515 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. n136-128-154:2198157:2198515 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0 n136-128-154:2198157:2198530 [0] NCCL INFO [Proxy Service] Device 0 CPU core 44 n136-128-154:2198157:2198531 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 109 n136-128-154:2198158:2198521 [1] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2198158:2198521 [1] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2198157:2198515 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 n136-128-154:2198157:2198515 [0] NCCL INFO 24 coll channels, 24 collnet channels, 0 nvls channels, 32 p2p channels, 32 p2p channels per peer n136-128-154:2198157:2198515 [0] NCCL INFO CC Off, workFifoBytes 1048576 n136-128-154:2198157:2198515 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2198157:2198515 [0] NCCL INFO ncclCommInitRankConfig comm 0x13b13560 rank 0 nranks 2 cudaDev 0 nvmlDev 6 busId c5000 commId 0x8dc6d77cdbe28b76 - Init COMPLETE n136-128-154:2198157:2198515 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 2 total 2.30 (kernels 0.21, alloc 0.16, bootstrap 1.76, allgathers 0.01, topo 0.08, graphs 0.00, connections 0.03, rest 0.06) n136-128-154:2198158:2198521 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin. n136-128-154:2198158:2198521 [1] NCCL INFO ncclCommInitRankConfig comm 0xf821860 rank 1 nranks 2 cudaDev 1 nvmlDev 7 busId c9000 commId 0x8dc6d77cdbe28b76 - Init COMPLETE n136-128-154:2198158:2198521 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 2 total 0.43 (kernels 0.12, alloc 0.13, bootstrap 0.00, allgathers 0.01, topo 0.08, graphs 0.00, connections 0.03, rest 0.06) n136-128-154:2198157:2198532 [0] NCCL INFO Channel 00/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 01/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 02/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 03/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 04/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 05/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 06/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 07/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 08/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 09/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 10/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 11/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 12/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 13/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 14/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 15/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 16/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 17/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 18/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 19/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 20/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 21/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 22/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Channel 23/0 : 0[6] -> 1[7] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 00/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 01/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 02/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 03/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 04/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 05/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 06/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 07/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 08/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 09/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 10/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 11/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 12/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 13/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 14/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 15/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 16/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 17/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 18/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 19/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 20/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 21/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 22/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2198533 [1] NCCL INFO Channel 23/0 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198157:2198532 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 n136-128-154:2198158:2198533 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1 2025-12-09:15:09:47 INFO [evaluator:559] Running loglikelihood requests 2025-12-09:15:09:47 INFO [evaluator:559] Running loglikelihood requests Running loglikelihood requests: 0%| | 0/20084 [00:00 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 01/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 02/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 03/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 04/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 05/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 06/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 07/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 08/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 09/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 10/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 11/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 12/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 13/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 14/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 15/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 16/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 17/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 18/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 19/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 20/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 21/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 22/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 23/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 24/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 25/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 26/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 27/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 28/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 29/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [1] NCCL INFO Channel 30/1 : 1[7] -> 0[6] via P2P/CUMEM/read n136-128-154:2198158:2206208 [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 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2198158:2198528 [1] NCCL INFO misc/socket.cc:915 -> 3 2025-12-09:15:33:33 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 |↑ |0.4218|± |0.0049| | | |none | 10|acc_norm|↑ |0.5710|± |0.0049| [rank0]:[W1209 15:33:35.615863487 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()) n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2198157:2198530 [0] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:64 -> 3 n136-128-154:2198158:2198528 [1] NCCL INFO misc/socket.cc:915 -> 3 n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:81 -> 3 n136-128-154:2198157:2206266 [0] NCCL INFO misc/socket.cc:863 -> 3 n136-128-154:2198158:2206213 [1] NCCL INFO comm 0xf821860 rank 1 nranks 2 cudaDev 1 busId c9000 - Abort COMPLETE n136-128-154:2198157:2206266 [0] NCCL INFO comm 0x13b13560 rank 0 nranks 2 cudaDev 0 busId c5000 - Abort COMPLETE 任务 hellaswag 评估完成! Execution time: 2885.37 seconds