diff --git "a/training_artifacts/logs/pipeline_cleaned.txt" "b/training_artifacts/logs/pipeline_cleaned.txt"
--- "a/training_artifacts/logs/pipeline_cleaned.txt"
+++ "b/training_artifacts/logs/pipeline_cleaned.txt"
@@ -48,6 +48,10 @@ Training config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__
 
 Starting distributed training with torch.distributed.run...
 
+*****************************************
+Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+*****************************************
+
 *****************************************
 Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
 *****************************************
@@ -59,19 +63,19 @@ Setting OMP_NUM_THREADS environment variable for each process to be 1 in default
   import pkg_resources
 /scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/jieba/_compat.py:18: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
   import pkg_resources
-[INFO|2025-10-22 16:01:48] llamafactory.hparams.parser:423 >> Process rank: 1, world size: 4, device: cuda:1, distributed training: True, compute dtype: torch.float16
 [INFO|2025-10-22 16:01:48] llamafactory.hparams.parser:143 >> Set `ddp_find_unused_parameters` to False in DDP training since LoRA is enabled.
-[INFO|2025-10-22 16:01:48] llamafactory.hparams.parser:423 >> Process rank: 0, world size: 4, device: cuda:0, distributed training: True, compute dtype: torch.float16
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,287 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,287 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,288 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,288 >> loading file added_tokens.json from cache at None
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,288 >> loading file special_tokens_map.json from cache at None
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,288 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,288 >> loading file chat_template.jinja from cache at None
-[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:01:48,457 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
-[INFO|configuration_utils.py:765] 2025-10-22 16:01:48,674 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
-[INFO|configuration_utils.py:839] 2025-10-22 16:01:48,676 >> Model config Qwen2Config {
+[INFO|2025-10-22 16:01:48] llamafactory.hparams.parser:423 >> Process rank: 3, world size: 4, device: cuda:1, distributed training: True, compute dtype: torch.float16
+[INFO|2025-10-22 16:01:48] llamafactory.hparams.parser:423 >> Process rank: 2, world size: 4, device: cuda:0, distributed training: True, compute dtype: torch.float16
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,643 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:01:48,814 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|configuration_utils.py:765] 2025-10-22 16:01:49,018 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:01:49,020 >> Model config Qwen2Config {
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -127,88 +131,82 @@ Setting OMP_NUM_THREADS environment variable for each process to be 1 in default
   "vocab_size": 151936
 }
 
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file added_tokens.json from cache at None
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file special_tokens_map.json from cache at None
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
-[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:48,738 >> loading file chat_template.jinja from cache at None
-[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:01:48,904 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
-[INFO|2025-10-22 16:01:48] llamafactory.data.loader:143 >> Loading dataset TAUR-dev/D-SFT_C-sft_exp_AT_pvv2__fixed-sft-data...
-/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4876: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-  warnings.warn(  # warn only once
-[rank0]:[W1022 16:01:49.085275271 ProcessGroupNCCL.cpp:5068] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
-gl064:2368555:2368555 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
-gl064:2368555:2368555 [0] NCCL INFO Bootstrap: Using ibs3:10.0.5.0<0>
-gl064:2368555:2368555 [0] NCCL INFO cudaDriverVersion 13000
-gl064:2368555:2368555 [0] NCCL INFO NCCL version 2.27.5+cuda12.9
-gl064:2368555:2368555 [0] NCCL INFO Comm config Blocking set to 1
-gl064:2368556:2368556 [1] NCCL INFO cudaDriverVersion 13000
-gl064:2368556:2368556 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
-gl064:2368556:2368556 [1] NCCL INFO Bootstrap: Using ibs3:10.0.5.0<0>
-gl064:2368556:2368556 [1] NCCL INFO NCCL version 2.27.5+cuda12.9
-gl064:2368556:2368556 [1] NCCL INFO Comm config Blocking set to 1
-gl064:2368555:2368616 [0] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
-gl064:2368555:2368616 [0] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
-gl064:2368555:2368616 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
-gl064:2368555:2368616 [0] NCCL INFO NCCL_IB_HCA set to mlx5
-gl064:2368556:2368617 [1] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
-gl064:2368556:2368617 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
-gl064:2368556:2368617 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
-gl064:2368556:2368617 [1] NCCL INFO NCCL_IB_HCA set to mlx5
-gl064:2368555:2368616 [0] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.0<0>
-gl064:2368555:2368616 [0] NCCL INFO Initialized NET plugin IB
-gl064:2368555:2368616 [0] NCCL INFO Assigned NET plugin IB to comm
-gl064:2368555:2368616 [0] NCCL INFO Using network IB
-gl064:2368555:2368616 [0] NCCL INFO ncclCommInitRankConfig comm 0x14bb0450 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x503fff15b9ea778b - Init START
-gl064:2368556:2368617 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.0<0>
-gl064:2368556:2368617 [1] NCCL INFO Initialized NET plugin IB
-gl064:2368556:2368617 [1] NCCL INFO Assigned NET plugin IB to comm
-gl064:2368556:2368617 [1] NCCL INFO Using network IB
-gl064:2368556:2368617 [1] NCCL INFO ncclCommInitRankConfig comm 0x15c0db00 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x503fff15b9ea778b - Init START
-gl064:2368555:2368616 [0] NCCL INFO RAS client listening socket at ::1<28028>
-gl064:2368556:2368617 [1] NCCL INFO RAS client listening socket at ::1<28028>
-gl064:2368555:2368616 [0] NCCL INFO Bootstrap timings total 0.321405 (create 0.000022, send 0.000239, recv 0.002956, ring 0.302954, delay 0.000000)
-gl064:2368556:2368617 [1] NCCL INFO Bootstrap timings total 0.319316 (create 0.000023, send 0.000069, recv 0.316285, ring 0.001306, delay 0.000000)
-gl064:2368555:2368616 [0] NCCL INFO Setting affinity for GPU 0 to 0-15
-gl064:2368556:2368617 [1] NCCL INFO Setting affinity for GPU 1 to 0-15
-gl064:2368556:2368617 [1] NCCL INFO comm 0x15c0db00 rank 1 nRanks 4 nNodes 2 localRanks 2 localRank 1 MNNVL 0
-gl064:2368555:2368616 [0] NCCL INFO comm 0x14bb0450 rank 0 nRanks 4 nNodes 2 localRanks 2 localRank 0 MNNVL 0
-gl064:2368555:2368616 [0] NCCL INFO Channel 00/02 : 0 1 2 3
-gl064:2368556:2368617 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0
-gl064:2368555:2368616 [0] NCCL INFO Channel 01/02 : 0 1 2 3
-gl064:2368556:2368617 [1] NCCL INFO P2P Chunksize set to 131072
-gl064:2368555:2368616 [0] NCCL INFO Trees [0] 1/2/-1->0->-1 [1] 1/-1/-1->0->2
-gl064:2368555:2368616 [0] NCCL INFO P2P Chunksize set to 131072
-gl064:2368556:2368617 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
-gl064:2368555:2368616 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
-gl064:2368555:2368616 [0] NCCL INFO Check P2P Type isAllDirectP2p 0 directMode 0
-gl064:2368555:2368623 [0] NCCL INFO [Proxy Service] Device 0 CPU core 9
-gl064:2368556:2368624 [1] NCCL INFO [Proxy Service] Device 1 CPU core 3
-gl064:2368555:2368625 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 10
-gl064:2368556:2368626 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 6
-gl064:2368556:2368617 [1] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
-gl064:2368556:2368617 [1] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
-gl064:2368555:2368616 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
-gl064:2368555:2368616 [0] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
-gl064:2368555:2368616 [0] NCCL INFO CC Off, workFifoBytes 1048576
-gl064:2368556:2368617 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
-gl064:2368556:2368617 [1] NCCL INFO ncclCommInitRankConfig comm 0x15c0db00 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x503fff15b9ea778b - Init COMPLETE
-gl064:2368556:2368617 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 4 total 0.45 (kernels 0.09, alloc 0.01, bootstrap 0.32, allgathers 0.01, topo 0.02, graphs 0.00, connections 0.00, rest 0.00)
-gl064:2368555:2368616 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
-gl064:2368555:2368616 [0] NCCL INFO ncclCommInitRankConfig comm 0x14bb0450 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x503fff15b9ea778b - Init COMPLETE
-gl064:2368555:2368616 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 4 total 0.45 (kernels 0.09, alloc 0.01, bootstrap 0.32, allgathers 0.01, topo 0.02, graphs 0.00, connections 0.00, rest 0.00)
-gl064:2368555:2368627 [0] NCCL INFO Channel 00/0 : 3[1] -> 0[0] [receive] via NET/IB/0
-gl064:2368555:2368629 [0] NCCL INFO [Proxy Progress] Device 0 CPU core 3
-gl064:2368555:2368627 [0] NCCL INFO Channel 01/0 : 3[1] -> 0[0] [receive] via NET/IB/0
-gl064:2368555:2368627 [0] NCCL INFO Channel 00 : 0[0] -> 1[1] via SHM/direct/direct
-gl064:2368555:2368627 [0] NCCL INFO Channel 01 : 0[0] -> 1[1] via SHM/direct/direct
-gl064:2368556:2368628 [1] NCCL INFO Channel 00/0 : 1[1] -> 2[0] [send] via NET/IB/0
-gl064:2368556:2368628 [1] NCCL INFO Channel 01/0 : 1[1] -> 2[0] [send] via NET/IB/0
-gl064:2368556:2368630 [1] NCCL INFO [Proxy Progress] Device 1 CPU core 7
-gl064:2368556:2368628 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
-gl064:2368555:2368627 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:01:49,085 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:01:49,252 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|2025-10-22 16:01:49] llamafactory.data.loader:143 >> Loading dataset TAUR-dev/D-SFT_C-sft_exp_AT_pvv2__fixed-sft-data...
+gl065:3752807:3752807 [1] NCCL INFO cudaDriverVersion 13000
+gl065:3752807:3752807 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl065:3752807:3752807 [1] NCCL INFO Bootstrap: Using ibs3:10.0.5.1<0>
+gl065:3752807:3752807 [1] NCCL INFO NCCL version 2.27.5+cuda12.9
+gl065:3752807:3752807 [1] NCCL INFO Comm config Blocking set to 1
+gl065:3752807:3752930 [1] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
+gl065:3752807:3752930 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
+gl065:3752807:3752930 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl065:3752807:3752930 [1] NCCL INFO NCCL_IB_HCA set to mlx5
+gl065:3752807:3752930 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.1<0>
+gl065:3752807:3752930 [1] NCCL INFO Initialized NET plugin IB
+gl065:3752807:3752930 [1] NCCL INFO Assigned NET plugin IB to comm
+gl065:3752807:3752930 [1] NCCL INFO Using network IB
+gl065:3752807:3752930 [1] NCCL INFO ncclCommInitRankConfig comm 0x14fa4e70 rank 3 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x503fff15b9ea778b - Init START
+[rank2]:[W1022 16:01:49.405201816 ProcessGroupNCCL.cpp:5068] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
+gl065:3752806:3752806 [0] NCCL INFO cudaDriverVersion 13000
+gl065:3752806:3752806 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl065:3752806:3752806 [0] NCCL INFO Bootstrap: Using ibs3:10.0.5.1<0>
+gl065:3752806:3752806 [0] NCCL INFO NCCL version 2.27.5+cuda12.9
+gl065:3752806:3752806 [0] NCCL INFO Comm config Blocking set to 1
+gl065:3752806:3752942 [0] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
+gl065:3752806:3752942 [0] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
+gl065:3752806:3752942 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl065:3752806:3752942 [0] NCCL INFO NCCL_IB_HCA set to mlx5
+gl065:3752806:3752942 [0] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.1<0>
+gl065:3752806:3752942 [0] NCCL INFO Initialized NET plugin IB
+gl065:3752806:3752942 [0] NCCL INFO Assigned NET plugin IB to comm
+gl065:3752806:3752942 [0] NCCL INFO Using network IB
+gl065:3752806:3752942 [0] NCCL INFO ncclCommInitRankConfig comm 0x12f77100 rank 2 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x503fff15b9ea778b - Init START
+gl065:3752807:3752930 [1] NCCL INFO RAS client listening socket at ::1<28028>
+gl065:3752806:3752942 [0] NCCL INFO RAS client listening socket at ::1<28028>
+gl065:3752807:3752930 [1] NCCL INFO Bootstrap timings total 0.320778 (create 0.000028, send 0.000429, recv 0.000956, ring 0.002613, delay 0.000000)
+gl065:3752806:3752942 [0] NCCL INFO Bootstrap timings total 0.012332 (create 0.000024, send 0.000508, recv 0.001014, ring 0.001299, delay 0.000000)
+gl065:3752807:3752930 [1] NCCL INFO Setting affinity for GPU 1 to 0-15
+gl065:3752806:3752942 [0] NCCL INFO Setting affinity for GPU 0 to 0-15
+gl065:3752806:3752942 [0] NCCL INFO comm 0x12f77100 rank 2 nRanks 4 nNodes 2 localRanks 2 localRank 0 MNNVL 0
+gl065:3752807:3752930 [1] NCCL INFO comm 0x14fa4e70 rank 3 nRanks 4 nNodes 2 localRanks 2 localRank 1 MNNVL 0
+gl065:3752806:3752942 [0] NCCL INFO Trees [0] 3/-1/-1->2->0 [1] 3/0/-1->2->-1
+gl065:3752806:3752942 [0] NCCL INFO P2P Chunksize set to 131072
+gl065:3752807:3752930 [1] NCCL INFO Trees [0] -1/-1/-1->3->2 [1] -1/-1/-1->3->2
+gl065:3752807:3752930 [1] NCCL INFO P2P Chunksize set to 131072
+gl065:3752806:3752942 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
+gl065:3752807:3752930 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
+gl065:3752807:3752949 [1] NCCL INFO [Proxy Service] Device 1 CPU core 3
+gl065:3752807:3752951 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 4
+gl065:3752806:3752950 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 13
+gl065:3752806:3752948 [0] NCCL INFO [Proxy Service] Device 0 CPU core 12
+gl065:3752807:3752930 [1] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
+gl065:3752807:3752930 [1] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
+gl065:3752806:3752942 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
+gl065:3752806:3752942 [0] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
+gl065:3752807:3752930 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
+gl065:3752807:3752930 [1] NCCL INFO ncclCommInitRankConfig comm 0x14fa4e70 rank 3 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x503fff15b9ea778b - Init COMPLETE
+gl065:3752806:3752942 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
+gl065:3752806:3752942 [0] NCCL INFO ncclCommInitRankConfig comm 0x12f77100 rank 2 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x503fff15b9ea778b - Init COMPLETE
+gl065:3752807:3752930 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 3 nranks 4 total 0.45 (kernels 0.08, alloc 0.01, bootstrap 0.32, allgathers 0.01, topo 0.02, graphs 0.00, connections 0.00, rest 0.00)
+gl065:3752806:3752942 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 2 nranks 4 total 0.13 (kernels 0.08, alloc 0.01, bootstrap 0.01, allgathers 0.00, topo 0.02, graphs 0.00, connections 0.00, rest 0.00)
+gl065:3752806:3752952 [0] NCCL INFO Channel 00/0 : 1[1] -> 2[0] [receive] via NET/IB/0
+gl065:3752806:3752952 [0] NCCL INFO Channel 01/0 : 1[1] -> 2[0] [receive] via NET/IB/0
+gl065:3752806:3752954 [0] NCCL INFO [Proxy Progress] Device 0 CPU core 15
+gl065:3752806:3752952 [0] NCCL INFO Channel 00 : 2[0] -> 3[1] via SHM/direct/direct
+gl065:3752806:3752952 [0] NCCL INFO Channel 01 : 2[0] -> 3[1] via SHM/direct/direct
+gl065:3752807:3752953 [1] NCCL INFO Channel 00/0 : 3[1] -> 0[0] [send] via NET/IB/0
+gl065:3752807:3752953 [1] NCCL INFO Channel 01/0 : 3[1] -> 0[0] [send] via NET/IB/0
+gl065:3752807:3752955 [1] NCCL INFO [Proxy Progress] Device 1 CPU core 9
+gl065:3752807:3752953 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
+gl065:3752806:3752952 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
 training example:
 input_ids:
 [33975, 25, 21806, 279, 2701, 3491, 13, 81917, 697, 32711, 3019, 553, 3019, 13, 3197, 498, 525, 8060, 11, 2968, 697, 4226, 304, 419, 3561, 25, 366, 9217, 2235, 21732, 4226, 12533, 9217, 94367, 2, 22079, 198, 16429, 279, 5109, 304, 279, 1140, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 1125, 1855, 458, 23606, 429, 16819, 220, 16, 21, 24, 13, 1446, 646, 990, 6770, 34784, 7525, 17973, 11, 85922, 11777, 608, 8, 323, 1817, 1372, 646, 1172, 387, 1483, 3055, 13, 4615, 6291, 1265, 2924, 264, 4013, 315, 7354, 330, 8304, 1599, 2974, 1380, 1817, 3019, 374, 264, 35972, 5666, 323, 279, 1590, 3019, 13653, 11508, 311, 279, 2169, 1372, 476, 432, 1265, 387, 264, 3175, 23606, 429, 3059, 304, 279, 2169, 382, 35127, 697, 4226, 304, 279, 2701, 3561, 510, 27, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 1339, 9064, 11993, 21732, 4226, 9940, 374, 279, 1140, 315, 7354, 311, 5545, 279, 2169, 1372, 476, 432, 1265, 387, 264, 3175, 23606, 429, 3059, 304, 279, 2169, 13, 4710, 2461, 3110, 510, 2679, 279, 1140, 315, 5109, 572, 508, 16, 11, 220, 17, 11, 220, 18, 60, 323, 279, 2169, 572, 220, 16, 11, 498, 1410, 3270, 510, 27, 9217, 397, 8304, 220, 16, 25, 220, 16, 488, 220, 17, 284, 220, 18, 198, 8304, 220, 17, 25, 220, 18, 608, 220, 18, 284, 220, 16, 198, 522, 9217, 1339, 269, 4710, 27, 9217, 397, 7, 16, 488, 220, 17, 8, 608, 220, 18, 198, 522, 9217, 1339, 10061, 594, 1744, 3019, 553, 3019, 13, 151643, 198, 71703, 25, 13708, 766, 397, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 8304, 220, 16, 25, 64547, 279, 5795, 481, 1205, 1184, 311, 1855, 458, 23606, 1667, 279, 5109, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 17, 25, 21144, 3204, 7525, 481, 78646, 11, 75240, 11, 46444, 11, 323, 12804, 624, 8304, 220, 18, 25, 9735, 2155, 27459, 315, 1493, 7525, 448, 279, 2661, 5109, 624, 8304, 220, 19, 25, 3393, 5257, 37906, 3080, 582, 1477, 825, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 20, 25, 9646, 582, 614, 1730, 279, 4396, 23606, 11, 3042, 432, 304, 279, 5189, 3561, 624, 6025, 2087, 1526, 678, 1493, 7354, 11, 1588, 594, 279, 1590, 23606, 429, 16819, 220, 16, 21, 24, 1447, 27, 9217, 397, 7, 21, 22, 353, 220, 17, 8, 488, 220, 22, 16, 481, 220, 18, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 488, 220, 22, 16, 284, 220, 16, 18, 23, 11, 1221, 220, 16, 18, 23, 481, 220, 18, 16, 284, 220, 16, 15, 22, 13, 1096, 1558, 537, 6144, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 10061, 752, 10146, 847, 4226, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 10253, 279, 5109, 504, 279, 2661, 738, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 311, 1352, 458, 23606, 6144, 287, 220, 16, 21, 24, 1667, 6770, 34784, 7525, 17973, 11, 85922, 11777, 608, 3593, 8304, 220, 16, 25, 71059, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 2691, 220, 22, 16, 311, 279, 1102, 504, 14822, 220, 16, 624, 8304, 220, 18, 25, 576, 1590, 1102, 686, 387, 220, 16, 21, 24, 382, 27, 9217, 397, 7, 21, 22, 353, 220, 18, 16, 8, 488, 220, 22, 16, 284, 220, 16, 21, 24, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 11, 15, 22, 22, 11, 1221, 220, 17, 11, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 11, 15, 15, 21, 13, 1096, 374, 537, 6144, 311, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 1462, 311, 3395, 847, 2033, 825, 803, 882, 624, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 785, 5795, 374, 311, 1855, 458, 23606, 1667, 279, 5109, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 429, 16819, 220, 16, 21, 24, 624, 40, 686, 1191, 553, 4460, 2155, 27459, 315, 1493, 2326, 5109, 323, 862, 3204, 7525, 17973, 11, 85922, 11777, 608, 568, 715, 12209, 358, 686, 15442, 1817, 10601, 323, 1779, 421, 807, 6144, 220, 16, 21, 24, 13, 1416, 537, 11, 358, 686, 3271, 389, 311, 279, 1790, 10601, 3080, 358, 1477, 825, 429, 4278, 624, 6025, 9271, 279, 4396, 10601, 11, 358, 686, 3042, 847, 6291, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 624, 8420, 594, 847, 11682, 3119, 1447, 16, 13, 4230, 678, 3204, 27459, 315, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 1667, 5256, 11, 75240, 11, 46444, 11, 323, 12804, 624, 17, 13, 54115, 1817, 10601, 311, 1490, 421, 432, 16819, 220, 16, 21, 24, 624, 18, 13, 1416, 902, 10601, 16819, 220, 16, 21, 24, 11, 13153, 279, 1882, 448, 501, 27459, 3080, 358, 1477, 825, 429, 4278, 624, 19, 13, 9646, 358, 614, 1730, 279, 4396, 10601, 11, 3158, 432, 438, 264, 3175, 23606, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 382, 7039, 1077, 752, 3161, 15039, 369, 279, 4396, 10601, 4894, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 16, 13, 14822, 220, 16, 25, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 15, 22, 22, 198, 17, 13, 14822, 220, 17, 25, 220, 17, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 15, 15, 21, 271, 54815, 11, 279, 1590, 1102, 374, 220, 17, 15, 15, 21, 11, 892, 1558, 537, 6144, 220, 16, 21, 24, 13, 576, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 3983, 3783, 11, 1077, 752, 1744, 911, 432, 1549, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 3838, 374, 279, 1102, 979, 498, 912, 1378, 5109, 504, 279, 1140, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 3786, 1939, 8304, 220, 16, 25, 2691, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 11778, 279, 2629, 504, 14822, 220, 16, 323, 912, 220, 22, 16, 382, 19357, 84670, 25, 320, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 271, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 25, 5512, 11, 220, 21, 22, 488, 220, 18, 16, 284, 220, 24, 23, 11, 1221, 220, 24, 23, 488, 220, 22, 16, 284, 220, 16, 21, 24, 13, 576, 4226, 374, 4396, 382, 27, 423, 8477, 397, 33092, 198, 522, 423, 8477, 397, 522, 34913, 397, 522, 26865, 10370, 39, 763, 11, 279, 4396, 4226, 374, 1447, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 29, 151643, 198]
@@ -460,8 +458,8 @@ Hence, the correct answer is:
 (67 + 31) + 71
 </answer><|endoftext|>
 
-[INFO|configuration_utils.py:765] 2025-10-22 16:01:50,484 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
-[INFO|configuration_utils.py:839] 2025-10-22 16:01:50,485 >> Model config Qwen2Config {
+[INFO|configuration_utils.py:765] 2025-10-22 16:01:50,482 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:01:50,483 >> Model config Qwen2Config {
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -518,41 +516,45 @@ Hence, the correct answer is:
 }
 
 [INFO|2025-10-22 16:01:50] llamafactory.model.model_utils.kv_cache:143 >> KV cache is disabled during training.
-[WARNING|logging.py:328] 2025-10-22 16:01:50,806 >> `torch_dtype` is deprecated! Use `dtype` instead!
-[INFO|modeling_utils.py:1172] 2025-10-22 16:01:50,807 >> loading weights file model.safetensors from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/model.safetensors
-[INFO|modeling_utils.py:2341] 2025-10-22 16:01:50,808 >> Instantiating Qwen2ForCausalLM model under default dtype torch.float16.
-[INFO|configuration_utils.py:986] 2025-10-22 16:01:50,808 >> Generate config GenerationConfig {
+[WARNING|logging.py:328] 2025-10-22 16:01:50,810 >> `torch_dtype` is deprecated! Use `dtype` instead!
+[INFO|modeling_utils.py:1172] 2025-10-22 16:01:50,811 >> loading weights file model.safetensors from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/model.safetensors
+[INFO|modeling_utils.py:2341] 2025-10-22 16:01:50,812 >> Instantiating Qwen2ForCausalLM model under default dtype torch.float16.
+[INFO|configuration_utils.py:986] 2025-10-22 16:01:50,813 >> Generate config GenerationConfig {
   "bos_token_id": 151643,
   "eos_token_id": 151643,
   "use_cache": false
 }
 
 `torch_dtype` is deprecated! Use `dtype` instead!
-[INFO|configuration_utils.py:941] 2025-10-22 16:01:51,084 >> loading configuration file generation_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/generation_config.json
-[INFO|configuration_utils.py:986] 2025-10-22 16:01:51,085 >> Generate config GenerationConfig {
+[INFO|configuration_utils.py:941] 2025-10-22 16:01:51,064 >> loading configuration file generation_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/generation_config.json
+[INFO|configuration_utils.py:986] 2025-10-22 16:01:51,064 >> Generate config GenerationConfig {
   "bos_token_id": 151643,
   "eos_token_id": 151643,
   "max_new_tokens": 2048
 }
 
-[INFO|dynamic_module_utils.py:423] 2025-10-22 16:01:51,114 >> Could not locate the custom_generate/generate.py inside Qwen/Qwen2.5-0.5B.
+[INFO|dynamic_module_utils.py:423] 2025-10-22 16:01:51,095 >> Could not locate the custom_generate/generate.py inside Qwen/Qwen2.5-0.5B.
 [INFO|2025-10-22 16:01:51] llamafactory.model.model_utils.checkpointing:143 >> Gradient checkpointing enabled.
 [INFO|2025-10-22 16:01:51] llamafactory.model.model_utils.attention:143 >> Using torch SDPA for faster training and inference.
 [INFO|2025-10-22 16:01:51] llamafactory.model.adapter:143 >> Upcasting trainable params to float32.
 [INFO|2025-10-22 16:01:51] llamafactory.model.adapter:143 >> Fine-tuning method: LoRA
-[INFO|2025-10-22 16:01:51] llamafactory.model.model_utils.misc:143 >> Found linear modules: up_proj,v_proj,q_proj,down_proj,gate_proj,k_proj,o_proj
+[INFO|2025-10-22 16:01:51] llamafactory.model.model_utils.misc:143 >> Found linear modules: gate_proj,k_proj,down_proj,o_proj,up_proj,v_proj,q_proj
 [INFO|2025-10-22 16:01:51] llamafactory.model.loader:143 >> trainable params: 4,399,104 || all params: 498,431,872 || trainable%: 0.8826
-[WARNING|trainer.py:906] 2025-10-22 16:01:51,639 >> The model is already on multiple devices. Skipping the move to device specified in `args`.
-[INFO|trainer.py:699] 2025-10-22 16:01:51,642 >> max_steps is given, it will override any value given in num_train_epochs
-[INFO|trainer.py:749] 2025-10-22 16:01:51,642 >> Using auto half precision backend
-[WARNING|trainer.py:982] 2025-10-22 16:01:51,643 >> The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151643}.
+[WARNING|trainer.py:906] 2025-10-22 16:01:51,337 >> The model is already on multiple devices. Skipping the move to device specified in `args`.
+[INFO|trainer.py:699] 2025-10-22 16:01:51,339 >> max_steps is given, it will override any value given in num_train_epochs
+[INFO|trainer.py:749] 2025-10-22 16:01:51,339 >> Using auto half precision backend
+[WARNING|trainer.py:982] 2025-10-22 16:01:51,340 >> The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151643}.
 The model is already on multiple devices. Skipping the move to device specified in `args`.
 The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151643}.
-[INFO|trainer.py:2519] 2025-10-22 16:01:51,823 >> ***** Running training *****
-[INFO|trainer.py:2520] 2025-10-22 16:01:51,823 >>   Num examples = 48,600
-[INFO|trainer.py:2521] 2025-10-22 16:01:51,823 >>   Num Epochs = 1
-[INFO|trainer.py:2522] 2025-10-22 16:01:51,823 >>   Instantaneous batch size per device = 1
-[INFO|trainer.py:2525] 2025-10-22 16:01:51,823 >>   Total train batch size (w. parallel, distributed & accumulation) = 4
+[INFO|trainer.py:2519] 2025-10-22 16:01:51,824 >> ***** Running training *****
+[INFO|trainer.py:2520] 2025-10-22 16:01:51,824 >>   Num examples = 48,600
+[INFO|trainer.py:2521] 2025-10-22 16:01:51,824 >>   Num Epochs = 1
+[INFO|trainer.py:2522] 2025-10-22 16:01:51,824 >>   Instantaneous batch size per device = 1
+[INFO|trainer.py:2525] 2025-10-22 16:01:51,824 >>   Total train batch size (w. parallel, distributed & accumulation) = 4
+[INFO|trainer.py:2526] 2025-10-22 16:01:51,824 >>   Gradient Accumulation steps = 1
+[INFO|trainer.py:2527] 2025-10-22 16:01:51,824 >>   Total optimization steps = 150
+[INFO|trainer.py:2528] 2025-10-22 16:01:51,826 >>   Number of trainable parameters = 4,399,104
+tion) = 4
 [INFO|trainer.py:2526] 2025-10-22 16:01:51,823 >>   Gradient Accumulation steps = 1
 [INFO|trainer.py:2527] 2025-10-22 16:01:51,823 >>   Total optimization steps = 150
 [INFO|trainer.py:2528] 2025-10-22 16:01:51,825 >>   Number of trainable parameters = 4,399,104
@@ -630,7 +632,14 @@ wandb:  View run at https://wandb.ai/ut_nlp_deduce/llamafactory/runs/f7vqjhyf
 [INFO|tokenization_utils_base.py:2421] 2025-10-22 16:02:05,402 >> chat template saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/chat_template.jinja
 [INFO|tokenization_utils_base.py:2590] 2025-10-22 16:02:05,406 >> tokenizer config file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/tokenizer_config.json
 [INFO|tokenization_utils_base.py:2599] 2025-10-22 16:02:05,410 >> Special tokens file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/special_tokens_map.json
- 34%|      | 51/150 [00:13<00:44,  2.20it/s] 35%|      | 52/150 [00:13<00:39,  2.47it/s] 35%|      | 53/150 [00:13<00:34,  2.82it/s] 36%|      | 54/150 [00:13<00:28,  3.40it/s] 37%|      | 55/150 [00:14<00:27,  3.45it/s] 37%|      | 56/150 [00:14<00:23,  4.06it/s] 38%|      | 57/150 [00:14<00:22,  4.10it/s] 39%|      | 58/150 [00:14<00:20,  4.51it/s] 39%|      | 59/150 [00:14<00:17,  5.19it/s] 40%|      | 60/150 [00:15<00:16,  5.57it/s]                                                {'loss': 0.6139, 'grad_norm': 0.4990316331386566, 'learning_rate': 3.0333333333333337e-05, 'epoch': 0.0}
+ 34%|      | 51/150 [00:13<00:44,  2.20it/s][INFO|trainer.py:2810] 2025-10-22 16:02:29,387 >> 
+
+Training completed. Do not forget to share your model on huggingface.co/models =)
+
+
+gl065:3752807:3752807 [1] NCCL INFO comm 0x14fa4e70 rank 3 nranks 4 cudaDev 1 busId 59000 - Destroy COMPLETE
+gl065:3752806:3752806 [0] NCCL INFO comm 0x12f77100 rank 2 nranks 4 cudaDev 0 busId 47000 - Destroy COMPLETE
+      | 58/150 [00:14<00:20,  4.51it/s] 39%|      | 59/150 [00:14<00:17,  5.19it/s] 40%|      | 60/150 [00:15<00:16,  5.57it/s]                                                {'loss': 0.6139, 'grad_norm': 0.4990316331386566, 'learning_rate': 3.0333333333333337e-05, 'epoch': 0.0}
  40%|      | 60/150 [00:15<00:16,  5.57it/s] 41%|      | 61/150 [00:15<00:17,  5.19it/s] 41%|     | 62/150 [00:15<00:15,  5.74it/s] 42%|     | 63/150 [00:15<00:16,  5.17it/s] 43%|     | 64/150 [00:15<00:15,  5.45it/s] 43%|     | 65/150 [00:16<00:17,  4.97it/s] 44%|     | 66/150 [00:16<00:18,  4.59it/s] 45%|     | 67/150 [00:16<00:17,  4.86it/s] 45%|     | 68/150 [00:16<00:18,  4.54it/s] 46%|     | 69/150 [00:16<00:19,  4.15it/s] 47%|     | 70/150 [00:17<00:19,  4.10it/s]                                                {'loss': 0.597, 'grad_norm': 0.5236718058586121, 'learning_rate': 2.7000000000000002e-05, 'epoch': 0.01}
  47%|     | 70/150 [00:17<00:19,  4.10it/s] 47%|     | 71/150 [00:17<00:19,  3.97it/s] 48%|     | 72/150 [00:17<00:17,  4.48it/s] 49%|     | 73/150 [00:17<00:19,  4.00it/s] 49%|     | 74/150 [00:18<00:18,  4.19it/s] 50%|     | 75/150 [00:18<00:15,  4.70it/s] 51%|     | 76/150 [00:18<00:15,  4.73it/s] 51%|    | 77/150 [00:18<00:13,  5.23it/s] 52%|    | 78/150 [00:18<00:15,  4.68it/s] 53%|    | 79/150 [00:19<00:15,  4.55it/s] 53%|    | 80/150 [00:19<00:16,  4.27it/s]                                                {'loss': 0.6205, 'grad_norm': 0.41710713505744934, 'learning_rate': 2.3666666666666668e-05, 'epoch': 0.01}
  53%|    | 80/150 [00:19<00:16,  4.27it/s] 54%|    | 81/150 [00:19<00:14,  4.65it/s] 55%|    | 82/150 [00:19<00:16,  4.06it/s] 55%|    | 83/150 [00:20<00:15,  4.45it/s] 56%|    | 84/150 [00:20<00:15,  4.39it/s] 57%|    | 85/150 [00:20<00:14,  4.45it/s] 57%|    | 86/150 [00:20<00:12,  5.07it/s] 58%|    | 87/150 [00:20<00:12,  5.19it/s] 59%|    | 88/150 [00:21<00:12,  4.88it/s] 59%|    | 89/150 [00:21<00:13,  4.59it/s] 60%|    | 90/150 [00:21<00:11,  5.22it/s]                                                {'loss': 0.6038, 'grad_norm': 0.5673879981040955, 'learning_rate': 2.0333333333333334e-05, 'epoch': 0.01}
@@ -1048,3 +1057,1208 @@ Preparing Training Artifacts
 ========================================
 Copying configuration files...
 Copying and cleaning training logs...
+Training artifacts prepared in: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/training_artifacts
+Contents:
+Log files:
+
+========================================
+STAGE 3: Uploading to HuggingFace Hub
+Repository: TAUR-dev/testing_llamafactory_helper_quick_test__interactive
+Start Time: Wed Oct 22 04:02:45 PM EDT 2025
+========================================
+Uploading contents of: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged
+Directory structure:
+
+Executing: huggingface-cli upload TAUR-dev/testing_llamafactory_helper_quick_test__interactive /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged .
+Start hashing 17 files.
+Finished hashing 17 files.
+[33m  Warning: 'huggingface-cli upload' is deprecated. Use 'hf upload' instead.[0m
+Processing Files (0 / 0)      : |          |  0.00B /  0.00B            
+New Data Upload               : |          |  0.00B /  0.00B            [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  11%|         |  109MB /  988MB            [A[A[A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  11%|         |  109MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  12%|        |  120MB / 1.00GB,   ???B/s  
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  23%|       |  226MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  24%|       |  238MB / 1.00GB,  587MB/s  
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  28%|       |  272MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  28%|       |  284MB / 1.00GB,  408MB/s  
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  28%|       |  272MB /  988MB            [A[A[A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  33%|      |  327MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  34%|      |  339MB / 1.00GB,  273MB/s  
+New Data Upload               :  41%|      | 55.1MB /  134MB, 68.8MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  38%|      |  379MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  39%|      |  390MB / 1.00GB,  270MB/s  
+New Data Upload               :  53%|    |  107MB /  201MB,  107MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  45%|     |  441MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  45%|     |  452MB / 1.00GB,  277MB/s  
+New Data Upload               :  63%|   |  169MB /  268MB,  141MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  51%|     |  506MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  52%|    |  517MB / 1.00GB,  284MB/s  
+New Data Upload               :  70%|   |  234MB /  335MB,  167MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  56%|    |  558MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  57%|    |  569MB / 1.00GB,  280MB/s  
+New Data Upload               :  85%| |  285MB /  335MB,  178MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  63%|   |  622MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  63%|   |  634MB / 1.00GB,  285MB/s  
+New Data Upload               :  87%| |  350MB /  402MB,  195MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  69%|   |  678MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  69%|   |  689MB / 1.00GB,  284MB/s  
+New Data Upload               :  86%| |  406MB /  469MB,  203MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  75%|  |  741MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  75%|  |  752MB / 1.00GB,  287MB/s  
+New Data Upload               :  87%| |  469MB /  537MB,  213MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  80%|  |  790MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  80%|  |  802MB / 1.00GB,  284MB/s  
+New Data Upload               :  86%| |  518MB /  604MB,  216MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  86%| |  848MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  86%| |  859MB / 1.00GB,  284MB/s  
+New Data Upload               :  86%| |  575MB /  671MB,  221MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors:  93%||  922MB /  988MB            [A[A[AProcessing Files (1 / 2)      :  93%||  933MB / 1.00GB,  290MB/s  
+New Data Upload               :  91%| |  650MB /  716MB,  232MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  987MB /  988MB            [A[A[AProcessing Files (1 / 2)      : 100%||  999MB / 1.00GB,  293MB/s  
+New Data Upload               : 100%||  715MB /  716MB,  238MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  987MB /  988MB            [A[A[A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[AProcessing Files (1 / 2)      : 100%||  999MB / 1.00GB,  258MB/s  
+New Data Upload               : 100%||  715MB /  716MB,  210MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[AProcessing Files (1 / 2)      : 100%||  999MB / 1.00GB,  244MB/s  
+New Data Upload               : 100%||  716MB /  716MB,  199MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[AProcessing Files (2 / 2)      : 100%|| 1.00GB / 1.00GB,  231MB/s  
+New Data Upload               : 100%||  716MB /  716MB,  188MB/s  [A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[A
+
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            [A[A
+
+
+  .../merged/model.safetensors: 100%||  988MB /  988MB            [A[A[AProcessing Files (2 / 2)      : 100%|| 1.00GB / 1.00GB,  209MB/s  
+New Data Upload               : 100%||  716MB /  716MB,  170MB/s  
+  ...ive/merged/tokenizer.json: 100%|| 11.4MB / 11.4MB            
+  .../merged/model.safetensors: 100%||  988MB /  988MB            
+Removing 10 file(s) from commit that have not changed.
+https://huggingface.co/TAUR-dev/testing_llamafactory_helper_quick_test__interactive/tree/main/.
+
+========================================
+Upload completed successfully
+Model and training artifacts uploaded to: TAUR-dev/testing_llamafactory_helper_quick_test__interactive
+End Time: Wed Oct 22 04:02:53 PM EDT 2025
+========================================
+
+========================================
+STAGE 4: Cleanup
+========================================
+Keeping checkpoints in: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
+Keeping merged model in: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged
+
+========================================
+PIPELINE COMPLETED SUCCESSFULLY
+End Time: Wed Oct 22 04:02:53 PM EDT 2025
+========================================
+
+========================================
+Cleaning up LlamaFactory processes
+========================================
+Cleaned up processes on gl064.hpc.nyu.edu
+Cleaning up processes on worker node: gl065
+Process cleanup complete
+========================================
+Job Name: lf_torch_test__interactive
+Hostname: gl064.hpc.nyu.edu
+Number of nodes: 2
+GPUs per node: 2
+Start Time: Wed Oct 22 04:04:47 PM EDT 2025
+Log file: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/logs/pipeline.log
+========================================
+Sourcing secrets from: /scratch/zrs2020/LlamaFactoryHelper/secrets.env
+
+========================================
+Configuration Paths
+========================================
+Train Config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/train_config.yaml
+Merge Config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/merge_config.yaml
+Dataset Info: 
+Output Dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
+Export Dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged
+HF Repo ID: TAUR-dev/testing_llamafactory_helper_quick_test__interactive
+
+
+========================================
+Multi-Node Coordination
+========================================
+This is the master node - coordinating worker nodes...
+Master node: gl064
+Master port: 29500
+World size: 2
+
+Launching on worker node 1: gl065
+All worker nodes launched successfully
+Master node (this node) will now join training as rank 0
+
+
+========================================
+STAGE 1: Training Model
+Start Time: Wed Oct 22 04:04:50 PM EDT 2025
+========================================
+Multi-node training detected
+Nodes: 2, GPUs per node: 2
+Master address: gl064
+Master port: 29500
+Node rank: 0
+World size: 2
+CUDA_VISIBLE_DEVICES: 0,1
+LLaMA-Factory path: /scratch/zrs2020/LlamaFactoryHelper/LLaMA-Factory
+Training config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/train_config.yaml
+
+Starting distributed training with torch.distributed.run...
+
+*****************************************
+Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+*****************************************
+==================================
+Sourcing secrets from: /scratch/zrs2020/LlamaFactoryHelper/secrets.env
+
+========================================
+Configuration Paths
+========================================
+Train Config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/train_config.yaml
+Merge Config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/merge_config.yaml
+Dataset Info: 
+Output Dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
+Export Dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged
+HF Repo ID: TAUR-dev/testing_llamafactory_helper_quick_test__interactive
+
+
+========================================
+STAGE 1: Training Model
+Start Time: Wed Oct 22 04:04:54 PM EDT 2025
+========================================
+Multi-node training detected
+Nodes: 2, GPUs per node: 2
+Master address: gl064
+Master port: 29500
+Node rank: 1
+World size: 2
+CUDA_VISIBLE_DEVICES: 0,1
+LLaMA-Factory path: /scratch/zrs2020/LlamaFactoryHelper/LLaMA-Factory
+Training config: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/train_config.yaml
+
+Starting distributed training with torch.distributed.run...
+
+*****************************************
+Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+*****************************************
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/transformers/utils/hub.py:110: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/y:18: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
+  import pkg_resources
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/jieba/_compat.py:18: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
+  import pkg_resources
+[INFO|2025-10-22 16:05:08] llamafactory.hparams.parser:143 >> Set `ddp_find_unused_parameters` to False in DDP training since LoRA is enabled.
+[INFO|2025-10-22 16:05:08] llamafactory.hparams.parser:423 >> Process rank: 0, world size: 4, device: cuda:0, distributed training: True, compute dtype: torch.float16
+[INFO|2025-10-22 16:05:08] llamafactory.hparams.parser:423 >> Process rank: 1, world size: 4, device: cuda:1, distributed training: True, compute dtype: torch.float16
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,654 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,654 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,654 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,654 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,654 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,655 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:08,655 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:05:08,827 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:09,061 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:09,063 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:09,127 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:05:09,295 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|2025-10-22 16:05:09] llamafactory.data.loader:143 >> Loading dataset TAUR-dev/D-SFT_C-sft_exp_AT_pvv2__fixed-sft-data...
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4876: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
+  warnings.warn(  # warn only once
+[rank0]:[W1022 16:05:09.469177712 ProcessGroupNCCL.cpp:5068] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
+gl064:2370172:2370172 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl064:2370172:2370172 [0] NCCL INFO Bootstrap: Using ibs3:10.0.5.0<0>
+gl064:2370172:2370172 [0] NCCL INFO cudaDriverVersion 13000
+gl064:2370172:2370172 [0] NCCL INFO NCCL version 2.27.5+cuda12.9
+gl064:2370172:2370172 [0] NCCL INFO Comm config Blocking set to 1
+gl064:2370173:2370173 [1] NCCL INFO cudaDriverVersion 13000
+gl064:2370173:2370173 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl064:2370173:2370173 [1] NCCL INFO Bootstrap: Using ibs3:10.0.5.0<0>
+gl064:2370173:2370173 [1] NCCL INFO NCCL version 2.27.5+cuda12.9
+gl064:2370173:2370173 [1] NCCL INFO Comm config Blocking set to 1
+gl064:2370172:2370248 [0] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
+gl064:2370172:2370248 [0] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
+gl064:2370172:2370248 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl064:2370172:2370248 [0] NCCL INFO NCCL_IB_HCA set to mlx5
+gl064:2370173:2370249 [1] NCCL INFO NET/Plugin: Could not find: libnccl-net.so. 
+gl064:2370173:2370249 [1] NCCL INFO NCCL_IB_DISABLE set by environment to 0.
+gl064:2370173:2370249 [1] NCCL INFO NCCL_SOCKET_IFNAME set by environment to ibs
+gl064:2370173:2370249 [1] NCCL INFO NCCL_IB_HCA set to mlx5
+gl064:2370172:2370248 [0] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.0<0>
+gl064:2370172:2370248 [0] NCCL INFO Initialized NET plugin IB
+gl064:2370172:2370248 [0] NCCL INFO Assigned NET plugin IB to comm
+gl064:2370172:2370248 [0] NCCL INFO Using network IB
+gl064:2370172:2370248 [0] NCCL INFO ncclCommInitRankConfig comm 0x14055870 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x781b8128280447d8 - Init START
+gl064:2370173:2370249 [1] NCCL INFO NET/IB : Using [0]mlx5_0:1/IB [RO]; OOB ibs3:10.0.5.0<0>
+gl064:2370173:2370249 [1] NCCL INFO Initialized NET plugin IB
+gl064:2370173:2370249 [1] NCCL INFO Assigned NET plugin IB to comm
+gl064:2370173:2370249 [1] NCCL INFO Using network IB
+gl064:2370173:2370249 [1] NCCL INFO ncclCommInitRankConfig comm 0x16d00aa0 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x781b8128280447d8 - Init START
+gl064:2370173:2370249 [1] NCCL INFO RAS client listening socket at ::1<28028>
+gl064:2370172:2370248 [0] NCCL INFO RAS client listening socket at ::1<28028>
+gl064:2370173:2370249 [1] NCCL INFO Bootstrap timings total 0.007859 (create 0.000023, send 0.000070, recv 0.006268, ring 0.000905, delay 0.000000)
+gl064:2370172:2370248 [0] NCCL INFO Bootstrap timings total 0.020201 (create 0.000022, send 0.000182, recv 0.001361, ring 0.000479, delay 0.000000)
+gl064:2370173:2370249 [1] NCCL INFO Setting affinity for GPU 1 to 0-15
+gl064:2370172:2370248 [0] NCCL INFO Setting affinity for GPU 0 to 0-15
+gl064:2370172:2370248 [0] NCCL INFO comm 0x14055870 rank 0 nRanks 4 nNodes 2 localRanks 2 localRank 0 MNNVL 0
+gl064:2370173:2370249 [1] NCCL INFO comm 0x16d00aa0 rank 1 nRanks 4 nNodes 2 localRanks 2 localRank 1 MNNVL 0
+gl064:2370172:2370248 [0] NCCL INFO Channel 00/02 : 0 1 2 3
+gl064:2370173:2370249 [1] NCCL INFO Trees [0] -1/-1/-1->1->0 [1] -1/-1/-1->1->0
+gl064:2370172:2370248 [0] NCCL INFO Channel 01/02 : 0 1 2 3
+gl064:2370173:2370249 [1] NCCL INFO P2P Chunksize set to 131072
+gl064:2370172:2370248 [0] NCCL INFO Trees [0] 1/2/-1->0->-1 [1] 1/-1/-1->0->2
+gl064:2370172:2370248 [0] NCCL INFO P2P Chunksize set to 131072
+gl064:2370173:2370249 [1] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
+gl064:2370172:2370248 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
+gl064:2370172:2370248 [0] NCCL INFO Check P2P Type isAllDirectP2p 0 directMode 0
+gl064:2370173:2370254 [1] NCCL INFO [Proxy Service] Device 1 CPU core 7
+gl064:2370173:2370256 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 14
+gl064:2370172:2370255 [0] NCCL INFO [Proxy Service] Device 0 CPU core 12
+gl064:2370172:2370257 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 15
+gl064:2370173:2370249 [1] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
+gl064:2370173:2370249 [1] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
+gl064:2370172:2370248 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
+gl064:2370172:2370248 [0] NCCL INFO 2 coll channels, 2 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
+gl064:2370172:2370248 [0] NCCL INFO CC Off, workFifoBytes 1048576
+gl064:2370173:2370249 [1] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
+gl064:2370173:2370249 [1] NCCL INFO ncclCommInitRankConfig comm 0x16d00aa0 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 59000 commId 0x781b8128280447d8 - Init COMPLETE
+gl064:2370173:2370249 [1] NCCL INFO Init timings - ncclCommInitRankConfig: rank 1 nranks 4 total 0.13 (kernels 0.08, alloc 0.01, bootstrap 0.01, allgathers 0.02, topo 0.01, graphs 0.00, connections 0.00, rest 0.00)
+gl064:2370172:2370248 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
+gl064:2370172:2370248 [0] NCCL INFO ncclCommInitRankConfig comm 0x14055870 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId 47000 commId 0x781b8128280447d8 - Init COMPLETE
+gl064:2370172:2370248 [0] NCCL INFO Init timings - ncclCommInitRankConfig: rank 0 nranks 4 total 0.14 (kernels 0.09, alloc 0.01, bootstrap 0.02, allgathers 0.01, topo 0.01, graphs 0.00, connections 0.00, rest 0.00)
+gl064:2370172:2370258 [0] NCCL INFO Channel 00/0 : 3[1] -> 0[0] [receive] via NET/IB/0
+gl064:2370172:2370258 [0] NCCL INFO Channel 01/0 : 3[1] -> 0[0] [receive] via NET/IB/0
+gl064:2370172:2370260 [0] NCCL INFO [Proxy Progress] Device 0 CPU core 1
+gl064:2370172:2370258 [0] NCCL INFO Channel 00 : 0[0] -> 1[1] via SHM/direct/direct
+gl064:2370172:2370258 [0] NCCL INFO Channel 01 : 0[0] -> 1[1] via SHM/direct/direct
+gl064:2370173:2370259 [1] NCCL INFO Channel 00/0 : 1[1] -> 2[0] [send] via NET/IB/0
+gl064:2370173:2370259 [1] NCCL INFO Channel 01/0 : 1[1] -> 2[0] [send] via NET/IB/0
+gl064:2370173:2370261 [1] NCCL INFO [Proxy Progress] Device 1 CPU core 13
+gl064:2370173:2370259 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
+gl064:2370172:2370258 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 0
+training example:
+input_ids:
+[33975, 25, 21806, 279, 2701, 3491, 13, 81917, 697, 32711, 3019, 553, 3019, 13, 3197, 498, 525, 8060, 11, 2968, 697, 4226, 304, 419, 3561, 25, 366, 9217, 2235, 21732, 4226, 12533, 9217, 94367, 2, 22079, 198, 16429, 279, 5109, 304, 279, 1140, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 1125, 1855, 458, 23606, 429, 16819, 220, 16, 21, 24, 13, 1446, 646, 990, 6770, 34784, 7525, 17973, 11, 85922, 11777, 608, 8, 323, 1817, 1372, 646, 1172, 387, 1483, 3055, 13, 4615, 6291, 1265, 2924, 264, 4013, 315, 7354, 330, 8304, 1599, 2974, 1380, 1817, 3019, 374, 264, 35972, 5666, 323, 279, 1590, 3019, 13653, 11508, 311, 279, 2169, 1372, 476, 432, 1265, 387, 264, 3175, 23606, 429, 3059, 304, 279, 2169, 382, 35127, 697, 4226, 304, 279, 2701, 3561, 510, 27, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 1339, 9064, 11993, 21732, 4226, 9940, 374, 279, 1140, 315, 7354, 311, 5545, 279, 2169, 1372, 476, 432, 1265, 387, 264, 3175, 23606, 429, 3059, 304, 279, 2169, 13, 4710, 2461, 3110, 510, 2679, 279, 1140, 315, 5109, 572, 508, 16, 11, 220, 17, 11, 220, 18, 60, 323, 279, 2169, 572, 220, 16, 11, 498, 1410, 3270, 510, 27, 9217, 397, 8304, 220, 16, 25, 220, 16, 488, 220, 17, 284, 220, 18, 198, 8304, 220, 17, 25, 220, 18, 608, 220, 18, 284, 220, 16, 198, 522, 9217, 1339, 269, 4710, 27, 9217, 397, 7, 16, 488, 220, 17, 8, 608, 220, 18, 198, 522, 9217, 1339, 10061, 594, 1744, 3019, 553, 3019, 13, 151643, 198, 71703, 25, 13708, 766, 397, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 8304, 220, 16, 25, 64547, 279, 5795, 481, 1205, 1184, 311, 1855, 458, 23606, 1667, 279, 5109, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 17, 25, 21144, 3204, 7525, 481, 78646, 11, 75240, 11, 46444, 11, 323, 12804, 624, 8304, 220, 18, 25, 9735, 2155, 27459, 315, 1493, 7525, 448, 279, 2661, 5109, 624, 8304, 220, 19, 25, 3393, 5257, 37906, 3080, 582, 1477, 825, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 20, 25, 9646, 582, 614, 1730, 279, 4396, 23606, 11, 3042, 432, 304, 279, 5189, 3561, 624, 6025, 2087, 1526, 678, 1493, 7354, 11, 1588, 594, 279, 1590, 23606, 429, 16819, 220, 16, 21, 24, 1447, 27, 9217, 397, 7, 21, 22, 353, 220, 17, 8, 488, 220, 22, 16, 481, 220, 18, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 488, 220, 22, 16, 284, 220, 16, 18, 23, 11, 1221, 220, 16, 18, 23, 481, 220, 18, 16, 284, 220, 16, 15, 22, 13, 1096, 1558, 537, 6144, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 10061, 752, 10146, 847, 4226, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 10253, 279, 5109, 504, 279, 2661, 738, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 311, 1352, 458, 23606, 6144, 287, 220, 16, 21, 24, 1667, 6770, 34784, 7525, 17973, 11, 85922, 11777, 608, 3593, 8304, 220, 16, 25, 71059, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 2691, 220, 22, 16, 311, 279, 1102, 504, 14822, 220, 16, 624, 8304, 220, 18, 25, 576, 1590, 1102, 686, 387, 220, 16, 21, 24, 382, 27, 9217, 397, 7, 21, 22, 353, 220, 18, 16, 8, 488, 220, 22, 16, 284, 220, 16, 21, 24, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 11, 15, 22, 22, 11, 1221, 220, 17, 11, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 11, 15, 15, 21, 13, 1096, 374, 537, 6144, 311, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 1462, 311, 3395, 847, 2033, 825, 803, 882, 624, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 785, 5795, 374, 311, 1855, 458, 23606, 1667, 279, 5109, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 429, 16819, 220, 16, 21, 24, 624, 40, 686, 1191, 553, 4460, 2155, 27459, 315, 1493, 2326, 5109, 323, 862, 3204, 7525, 17973, 11, 85922, 11777, 608, 568, 715, 12209, 358, 686, 15442, 1817, 10601, 323, 1779, 421, 807, 6144, 220, 16, 21, 24, 13, 1416, 537, 11, 358, 686, 3271, 389, 311, 279, 1790, 10601, 3080, 358, 1477, 825, 429, 4278, 624, 6025, 9271, 279, 4396, 10601, 11, 358, 686, 3042, 847, 6291, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 624, 8420, 594, 847, 11682, 3119, 1447, 16, 13, 4230, 678, 3204, 27459, 315, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 1667, 5256, 11, 75240, 11, 46444, 11, 323, 12804, 624, 17, 13, 54115, 1817, 10601, 311, 1490, 421, 432, 16819, 220, 16, 21, 24, 624, 18, 13, 1416, 902, 10601, 16819, 220, 16, 21, 24, 11, 13153, 279, 1882, 448, 501, 27459, 3080, 358, 1477, 825, 429, 4278, 624, 19, 13, 9646, 358, 614, 1730, 279, 4396, 10601, 11, 3158, 432, 438, 264, 3175, 23606, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 382, 7039, 1077, 752, 3161, 15039, 369, 279, 4396, 10601, 4894, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 16, 13, 14822, 220, 16, 25, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 15, 22, 22, 198, 17, 13, 14822, 220, 17, 25, 220, 17, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 15, 15, 21, 271, 54815, 11, 279, 1590, 1102, 374, 220, 17, 15, 15, 21, 11, 892, 1558, 537, 6144, 220, 16, 21, 24, 13, 576, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 3983, 3783, 11, 1077, 752, 1744, 911, 432, 1549, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 3838, 374, 279, 1102, 979, 498, 912, 1378, 5109, 504, 279, 1140, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 3786, 1939, 8304, 220, 16, 25, 2691, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 11778, 279, 2629, 504, 14822, 220, 16, 323, 912, 220, 22, 16, 382, 19357, 84670, 25, 320, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 271, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 25, 5512, 11, 220, 21, 22, 488, 220, 18, 16, 284, 220, 24, 23, 11, 1221, 220, 24, 23, 488, 220, 22, 16, 284, 220, 16, 21, 24, 13, 576, 4226, 374, 4396, 382, 27, 423, 8477, 397, 33092, 198, 522, 423, 8477, 397, 522, 34913, 397, 522, 26865, 10370, 39, 763, 11, 279, 4396, 4226, 374, 1447, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 29, 151643, 198]
+inputs:
+Human: Answer the following problem. Explain your reasoning step by step. When you are finished, give your answer in this format: <answer>(your answer)</answer>.
+
+# Problem
+Using the numbers in the list [67, 71, 31], create an equation that equals 169. You can use basic arithmetic operations (+, -, *, /) and each number can only be used once. Your solution should include a series of steps "Step X:" where each step is a mathematical operation and the final step ultimately leads to the target number or it should be a single equation that results in the target.
+
+Give your answer in the following format:
+<answer>
+(your answer)
+</answer>
+
+Where "(your answer)" is the list of steps to reach the target number or it should be a single equation that results in the target. 
+
+For example:
+If the list of numbers was [1, 2, 3] and the target was 1, you could write:
+<answer>
+Step 1: 1 + 2 = 3
+Step 2: 3 / 3 = 1
+</answer>
+
+or 
+
+<answer>
+(1 + 2) / 3
+</answer>
+
+Let's think step by step.<|endoftext|>
+Assistant:<think>
+<sample>
+First, I'll try to understand the problem better by writing out a plan and go really deep into detail about how I should solve this.
+Step 1: Identify the goal - We need to create an equation using the numbers [67, 71, 31] that equals 169.
+Step 2: Consider possible operations - Addition, subtraction, multiplication, and division.
+Step 3: Try different combinations of these operations with the given numbers.
+Step 4: Test various equations until we find one that equals 169.
+Step 5: Once we have found the correct equation, present it in the specified format.
+After going through all these steps, here's the final equation that equals 169:
+
+<answer>
+(67 * 2) + 71 - 31
+</answer>
+</sample>
+<reflect>
+Let us verify this answer:
+
+First, 67 + 71 = 138, then 138 - 31 = 107. This does not equal 169, so the answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+Let me verify my answer.
+<sample>
+Rewritten Question:
+
+Use the numbers from the given set [67, 71, 31] to form an equation equaling 169 using basic arithmetic operations (+, -, *, /).
+
+Step 1: Multiply 67 and 31.
+Step 2: Add 71 to the result from Step 1.
+Step 3: The final result will be 169.
+
+<answer>
+(67 * 31) + 71 = 169
+</answer>
+</sample>
+<reflect>
+Let us verify this answer:
+
+First, 67 * 31 = 2,077, then 2,077 - 71 = 2,006. This is not equal to 169, so the answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+Time to review my response one more time.
+<sample>
+First, I'll try to understand the problem better by writing out a plan and go really deep into detail about how I should solve this.
+The goal is to create an equation using the numbers 67, 71, and 31 that equals 169.
+I will start by trying different combinations of these three numbers and their possible operations (+, -, *, /). 
+Then I will evaluate each combination and check if they equal 169. If not, I will move on to the next combination until I find one that works.
+After finding the correct combination, I will present my solution in the "<answer>
+(your answer)
+</answer>" format.
+Here's my detailed plan:
+
+1. Create all possible combinations of 67, 71, and 31 using addition, subtraction, multiplication, and division.
+2. Evaluate each combination to see if it equals 169.
+3. If no combination equals 169, repeat the process with new combinations until I find one that works.
+4. Once I have found the correct combination, express it as a single equation in the "<answer>
+(your answer)
+</answer>" format.
+
+Now let me begin searching for the correct combination!
+</sample>
+<reflect>
+Let us verify this answer:
+
+1. Step 1: 67 * 31 = 2077
+2. Step 2: 2077 - 71 = 2006
+
+Therefore, the final result is 2006, which does not equal 169. The answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+But wait, let me think about it again.
+<sample>
+Rewritten Question:
+
+What is the result when you add two numbers from the list [67, 71, 31] together?
+
+Step 1: Add 67 and 31.
+Step 2: Take the sum from Step 1 and add 71.
+
+Final Equation: (67 + 31) + 71
+
+<answer>
+(67 + 31) + 71
+</answer>
+</sample>
+<reflect>
+Let us verify this answer: First, 67 + 31 = 98, then 98 + 71 = 169. The answer is correct.
+
+<verdict>
+Correct
+</verdict>
+</reflect>
+</think>
+
+
+Hence, the correct answer is:
+
+<answer>
+(67 + 31) + 71
+</answer><|endoftext|>
+
+label_ids:
+[-100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, 13708, 766, 397, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 8304, 220, 16, 25, 64547, 279, 5795, 481, 1205, 1184, 311, 1855, 458, 23606, 1667, 279, 5109, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 17, 25, 21144, 3204, 7525, 481, 78646, 11, 75240, 11, 46444, 11, 323, 12804, 624, 8304, 220, 18, 25, 9735, 2155, 27459, 315, 1493, 7525, 448, 279, 2661, 5109, 624, 8304, 220, 19, 25, 3393, 5257, 37906, 3080, 582, 1477, 825, 429, 16819, 220, 16, 21, 24, 624, 8304, 220, 20, 25, 9646, 582, 614, 1730, 279, 4396, 23606, 11, 3042, 432, 304, 279, 5189, 3561, 624, 6025, 2087, 1526, 678, 1493, 7354, 11, 1588, 594, 279, 1590, 23606, 429, 16819, 220, 16, 21, 24, 1447, 27, 9217, 397, 7, 21, 22, 353, 220, 17, 8, 488, 220, 22, 16, 481, 220, 18, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 488, 220, 22, 16, 284, 220, 16, 18, 23, 11, 1221, 220, 16, 18, 23, 481, 220, 18, 16, 284, 220, 16, 15, 22, 13, 1096, 1558, 537, 6144, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 10061, 752, 10146, 847, 4226, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 10253, 279, 5109, 504, 279, 2661, 738, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 311, 1352, 458, 23606, 6144, 287, 220, 16, 21, 24, 1667, 6770, 34784, 7525, 17973, 11, 85922, 11777, 608, 3593, 8304, 220, 16, 25, 71059, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 2691, 220, 22, 16, 311, 279, 1102, 504, 14822, 220, 16, 624, 8304, 220, 18, 25, 576, 1590, 1102, 686, 387, 220, 16, 21, 24, 382, 27, 9217, 397, 7, 21, 22, 353, 220, 18, 16, 8, 488, 220, 22, 16, 284, 220, 16, 21, 24, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 5338, 11, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 11, 15, 22, 22, 11, 1221, 220, 17, 11, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 11, 15, 15, 21, 13, 1096, 374, 537, 6144, 311, 220, 16, 21, 24, 11, 773, 279, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 1462, 311, 3395, 847, 2033, 825, 803, 882, 624, 27, 13611, 397, 5338, 11, 358, 3278, 1430, 311, 3535, 279, 3491, 2664, 553, 4378, 700, 264, 3119, 323, 728, 2167, 5538, 1119, 7716, 911, 1246, 358, 1265, 11625, 419, 624, 785, 5795, 374, 311, 1855, 458, 23606, 1667, 279, 5109, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 429, 16819, 220, 16, 21, 24, 624, 40, 686, 1191, 553, 4460, 2155, 27459, 315, 1493, 2326, 5109, 323, 862, 3204, 7525, 17973, 11, 85922, 11777, 608, 568, 715, 12209, 358, 686, 15442, 1817, 10601, 323, 1779, 421, 807, 6144, 220, 16, 21, 24, 13, 1416, 537, 11, 358, 686, 3271, 389, 311, 279, 1790, 10601, 3080, 358, 1477, 825, 429, 4278, 624, 6025, 9271, 279, 4396, 10601, 11, 358, 686, 3042, 847, 6291, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 624, 8420, 594, 847, 11682, 3119, 1447, 16, 13, 4230, 678, 3204, 27459, 315, 220, 21, 22, 11, 220, 22, 16, 11, 323, 220, 18, 16, 1667, 5256, 11, 75240, 11, 46444, 11, 323, 12804, 624, 17, 13, 54115, 1817, 10601, 311, 1490, 421, 432, 16819, 220, 16, 21, 24, 624, 18, 13, 1416, 902, 10601, 16819, 220, 16, 21, 24, 11, 13153, 279, 1882, 448, 501, 27459, 3080, 358, 1477, 825, 429, 4278, 624, 19, 13, 9646, 358, 614, 1730, 279, 4396, 10601, 11, 3158, 432, 438, 264, 3175, 23606, 304, 279, 4055, 9217, 397, 7021, 413, 4226, 340, 522, 9217, 9877, 3561, 382, 7039, 1077, 752, 3161, 15039, 369, 279, 4396, 10601, 4894, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 1447, 16, 13, 14822, 220, 16, 25, 220, 21, 22, 353, 220, 18, 16, 284, 220, 17, 15, 22, 22, 198, 17, 13, 14822, 220, 17, 25, 220, 17, 15, 22, 22, 481, 220, 22, 16, 284, 220, 17, 15, 15, 21, 271, 54815, 11, 279, 1590, 1102, 374, 220, 17, 15, 15, 21, 11, 892, 1558, 537, 6144, 220, 16, 21, 24, 13, 576, 4226, 374, 15114, 382, 27, 423, 8477, 397, 40468, 198, 522, 423, 8477, 397, 522, 34913, 397, 3983, 3783, 11, 1077, 752, 1744, 911, 432, 1549, 624, 27, 13611, 397, 58465, 12153, 15846, 1447, 3838, 374, 279, 1102, 979, 498, 912, 1378, 5109, 504, 279, 1140, 508, 21, 22, 11, 220, 22, 16, 11, 220, 18, 16, 60, 3786, 1939, 8304, 220, 16, 25, 2691, 220, 21, 22, 323, 220, 18, 16, 624, 8304, 220, 17, 25, 11778, 279, 2629, 504, 14822, 220, 16, 323, 912, 220, 22, 16, 382, 19357, 84670, 25, 320, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 271, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 397, 522, 13611, 397, 27, 34913, 397, 10061, 601, 10146, 419, 4226, 25, 5512, 11, 220, 21, 22, 488, 220, 18, 16, 284, 220, 24, 23, 11, 1221, 220, 24, 23, 488, 220, 22, 16, 284, 220, 16, 21, 24, 13, 576, 4226, 374, 4396, 382, 27, 423, 8477, 397, 33092, 198, 522, 423, 8477, 397, 522, 34913, 397, 522, 26865, 10370, 39, 763, 11, 279, 4396, 4226, 374, 1447, 27, 9217, 397, 7, 21, 22, 488, 220, 18, 16, 8, 488, 220, 22, 16, 198, 522, 9217, 29, 151643, 198]
+labels:
+<think>
+<sample>
+First, I'll try to understand the problem better by writing out a plan and go really deep into detail about how I should solve this.
+Step 1: Identify the goal - We need to create an equation using the numbers [67, 71, 31] that equals 169.
+Step 2: Consider possible operations - Addition, subtraction, multiplication, and division.
+Step 3: Try different combinations of these operations with the given numbers.
+Step 4: Test various equations until we find one that equals 169.
+Step 5: Once we have found the correct equation, present it in the specified format.
+After going through all these steps, here's the final equation that equals 169:
+
+<answer>
+(67 * 2) + 71 - 31
+</answer>
+</sample>
+<reflect>
+Let us verify this answer:
+
+First, 67 + 71 = 138, then 138 - 31 = 107. This does not equal 169, so the answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+Let me verify my answer.
+<sample>
+Rewritten Question:
+
+Use the numbers from the given set [67, 71, 31] to form an equation equaling 169 using basic arithmetic operations (+, -, *, /).
+
+Step 1: Multiply 67 and 31.
+Step 2: Add 71 to the result from Step 1.
+Step 3: The final result will be 169.
+
+<answer>
+(67 * 31) + 71 = 169
+</answer>
+</sample>
+<reflect>
+Let us verify this answer:
+
+First, 67 * 31 = 2,077, then 2,077 - 71 = 2,006. This is not equal to 169, so the answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+Time to review my response one more time.
+<sample>
+First, I'll try to understand the problem better by writing out a plan and go really deep into detail about how I should solve this.
+The goal is to create an equation using the numbers 67, 71, and 31 that equals 169.
+I will start by trying different combinations of these three numbers and their possible operations (+, -, *, /). 
+Then I will evaluate each combination and check if they equal 169. If not, I will move on to the next combination until I find one that works.
+After finding the correct combination, I will present my solution in the "<answer>
+(your answer)
+</answer>" format.
+Here's my detailed plan:
+
+1. Create all possible combinations of 67, 71, and 31 using addition, subtraction, multiplication, and division.
+2. Evaluate each combination to see if it equals 169.
+3. If no combination equals 169, repeat the process with new combinations until I find one that works.
+4. Once I have found the correct combination, express it as a single equation in the "<answer>
+(your answer)
+</answer>" format.
+
+Now let me begin searching for the correct combination!
+</sample>
+<reflect>
+Let us verify this answer:
+
+1. Step 1: 67 * 31 = 2077
+2. Step 2: 2077 - 71 = 2006
+
+Therefore, the final result is 2006, which does not equal 169. The answer is incorrect.
+
+<verdict>
+Incorrect
+</verdict>
+</reflect>
+But wait, let me think about it again.
+<sample>
+Rewritten Question:
+
+What is the result when you add two numbers from the list [67, 71, 31] together?
+
+Step 1: Add 67 and 31.
+Step 2: Take the sum from Step 1 and add 71.
+
+Final Equation: (67 + 31) + 71
+
+<answer>
+(67 + 31) + 71
+</answer>
+</sample>
+<reflect>
+Let us verify this answer: First, 67 + 31 = 98, then 98 + 71 = 169. The answer is correct.
+
+<verdict>
+Correct
+</verdict>
+</reflect>
+</think>
+
+
+Hence, the correct answer is:
+
+<answer>
+(67 + 31) + 71
+</answer><|endoftext|>
+
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:10,670 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:10,671 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|2025-10-22 16:05:10] llamafactory.model.model_utils.kv_cache:143 >> KV cache is disabled during training.
+[WARNING|logging.py:328] 2025-10-22 16:05:11,006 >> `torch_dtype` is deprecated! Use `dtype` instead!
+[INFO|modeling_utils.py:1172] 2025-10-22 16:05:11,007 >> loading weights file model.safetensors from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/model.safetensors
+[INFO|modeling_utils.py:2341] 2025-10-22 16:05:11,009 >> Instantiating Qwen2ForCausalLM model under default dtype torch.float16.
+[INFO|configuration_utils.py:986] 2025-10-22 16:05:11,009 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "use_cache": false
+}
+
+`torch_dtype` is deprecated! Use `dtype` instead!
+[INFO|configuration_utils.py:941] 2025-10-22 16:05:11,287 >> loading configuration file generation_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/generation_config.json
+[INFO|configuration_utils.py:986] 2025-10-22 16:05:11,287 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048
+}
+
+[INFO|dynamic_module_utils.py:423] 2025-10-22 16:05:11,318 >> Could not locate the custom_generate/generate.py inside Qwen/Qwen2.5-0.5B.
+[INFO|2025-10-22 16:05:11] llamafactory.model.model_utils.checkpointing:143 >> Gradient checkpointing enabled.
+[INFO|2025-10-22 16:05:11] llamafactory.model.model_utils.attention:143 >> Using torch SDPA for faster training and inference.
+[INFO|2025-10-22 16:05:11] llamafactory.model.adapter:143 >> Upcasting trainable params to float32.
+[INFO|2025-10-22 16:05:11] llamafactory.model.adapter:143 >> Fine-tuning method: LoRA
+[INFO|2025-10-22 16:05:11] llamafactory.model.model_utils.misc:143 >> Found linear modules: gate_proj,k_proj,o_proj,v_proj,up_proj,q_proj,down_proj
+[INFO|2025-10-22 16:05:11] llamafactory.model.loader:143 >> trainable params: 4,399,104 || all params: 498,431,872 || trainable%: 0.8826
+[WARNING|trainer.py:906] 2025-10-22 16:05:11,559 >> The model is already on multiple devices. Skipping the move to device specified in `args`.
+[INFO|trainer.py:699] 2025-10-22 16:05:11,561 >> max_steps is given, it will override any value given in num_train_epochs
+[INFO|trainer.py:749] 2025-10-22 16:05:11,561 >> Using auto half precision backend
+[WARNING|2025-10-22 16:05:11] llamafactory.train.callbacks:154 >> Previous trainer log in this folder will be deleted.
+[WARNING|trainer.py:982] 2025-10-22 16:05:11,564 >> The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151643}.
+The model is already on multiple devices. Skipping the move to device specified in `args`.
+The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151643}.
+[INFO|trainer.py:2519] 2025-10-22 16:05:12,165 >> ***** Running training *****
+[INFO|trainer.py:2520] 2025-10-22 16:05:12,165 >>   Num examples = 48,600
+[INFO|trainer.py:2521] 2025-10-22 16:05:12,165 >>   Num Epochs = 1
+[INFO|trainer.py:2522] 2025-10-22 16:05:12,165 >>   Instantaneous batch size per device = 1
+[INFO|trainer.py:2525] 2025-10-22 16:05:12,165 >>   Total train batch size (w. parallel, distributed & accumulation) = 4
+[INFO|trainer.py:2526] 2025-10-22 16:05:12,165 >>   Gradient Accumulation steps = 1
+[INFO|trainer.py:2527] 2025-10-22 16:05:12,165 >>   Total optimization steps = 100
+[INFO|trainer.py:2528] 2025-10-22 16:05:12,167 >>   Number of trainable parameters = 4,399,104
+[INFO|integration_utils.py:867] 2025-10-22 16:05:12,178 >> Automatic Weights & Biases logging enabled, to disable set os.environ["WANDB_DISABLED"] = "true"
+wandb: Currently logged in as: zsprague (ut_nlp_deduce) to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
+wandb: Tracking run with wandb version 0.22.2
+wandb: Run data is saved locally in /scratch/zrs2020/LlamaFactoryHelper/wandb/run-20251022_160512-dppinxzz
+wandb: Run `wandb offline` to turn off syncing.
+wandb: Syncing run interactive_test
+wandb:  View project at https://wandb.ai/ut_nlp_deduce/llamafactory
+wandb:  View run at https://wandb.ai/ut_nlp_deduce/llamafactory/runs/dppinxzz
+  0%|          | 0/100 [00:00<?, ?it/s]  1%|          | 1/100 [00:00<00:59,  1.66it/s]  2%|         | 2/100 [00:00<00:34,  2.87it/s]  3%|         | 3/100 [00:00<00:25,  3.81it/s]  4%|         | 4/100 [00:01<00:23,  4.04it/s]  5%|         | 5/100 [00:01<00:34,  2.74it/s]  6%|         | 6/100 [00:01<00:30,  3.07it/s]  7%|         | 7/100 [00:02<00:28,  3.31it/s]  8%|         | 8/100 [00:02<00:24,  3.68it/s]  9%|         | 9/100 [00:02<00:23,  3.95it/s] 10%|         | 10/100 [00:02<00:20,  4.34it/s]                                                {'loss': 0.8095, 'grad_norm': 0.4080905020236969, 'learning_rate': 4.55e-05, 'epoch': 0.0}
+ 10%|         | 10/100 [00:02<00:20,  4.34it/s] 11%|         | 11/100 [00:03<00:20,  4.41it/s] 12%|        | 12/100 [00:03<00:26,  3.30it/s] 13%|        | 13/100 [00:03<00:22,  3.89it/s] 14%|        | 14/100 [00:03<00:19,  4.43it/s] 15%|        | 15/100 [00:04<00:17,  4.78it/s] 16%|        | 16/100 [00:04<00:23,  3.63it/s] 17%|        | 17/100 [00:04<00:22,  3.65it/s] 18%|        | 18/100 [00:04<00:22,  3.71it/s] 19%|        | 19/100 [00:05<00:21,  3.70it/s] 20%|        | 20/100 [00:05<00:20,  3.98it/s]                                                {'loss': 0.7526, 'grad_norm': 0.3977513313293457, 'learning_rate': 4.05e-05, 'epoch': 0.0}
+ 20%|        | 20/100 [00:05<00:20,  3.98it/s] 21%|        | 21/100 [00:05<00:20,  3.94it/s] 22%|       | 22/100 [00:05<00:18,  4.18it/s] 23%|       | 23/100 [00:06<00:18,  4.07it/s] 24%|       | 24/100 [00:06<00:17,  4.40it/s] 25%|       | 25/100 [00:06<00:18,  4.02it/s] 26%|       | 26/100 [00:06<00:17,  4.17it/s] 27%|       | 27/100 [00:07<00:17,  4.09it/s] 28%|       | 28/100 [00:07<00:19,  3.61it/s] 29%|       | 29/100 [00:07<00:19,  3.72it/s] 30%|       | 30/100 [00:07<00:18,  3.83it/s]                                                {'loss': 0.7384, 'grad_norm': 0.4652378559112549, 'learning_rate': 3.55e-05, 'epoch': 0.0}
+ 30%|       | 30/100 [00:07<00:18,  3.83it/s] 31%|       | 31/100 [00:08<00:18,  3.82it/s] 32%|      | 32/100 [00:08<00:16,  4.04it/s] 33%|      | 33/100 [00:08<00:15,  4.44it/s] 34%|      | 34/100 [00:08<00:13,  4.88it/s] 35%|      | 35/100 [00:09<00:14,  4.58it/s] 36%|      | 36/100 [00:09<00:12,  5.04it/s] 37%|      | 37/100 [00:09<00:13,  4.71it/s] 38%|      | 38/100 [00:09<00:13,  4.67it/s] 39%|      | 39/100 [00:09<00:12,  4.98it/s] 40%|      | 40/100 [00:10<00:13,  4.59it/s]                                                {'loss': 0.7139, 'grad_norm': 0.37474513053894043, 'learning_rate': 3.05e-05, 'epoch': 0.0}
+ 40%|      | 40/100 [00:10<00:13,  4.59it/s] 41%|      | 41/100 [00:10<00:14,  4.10it/s] 42%|     | 42/100 [00:10<00:14,  4.07it/s] 43%|     | 43/100 [00:10<00:12,  4.58it/s] 44%|     | 44/100 [00:10<00:11,  4.90it/s] 45%|     | 45/100 [00:11<00:10,  5.33it/s] 46%|     | 46/100 [00:11<00:10,  4.97it/s] 47%|     | 47/100 [00:11<00:10,  4.88it/s] 48%|     | 48/100 [00:11<00:10,  5.13it/s] 49%|     | 49/100 [00:12<00:11,  4.50it/s] 50%|     | 50/100 [00:12<00:11,  4.48it/s]                                                {'loss': 0.6497, 'grad_norm': 0.5904101133346558, 'learning_rate': 2.5500000000000003e-05, 'epoch': 0.0}
+ 50%|     | 50/100 [00:12<00:11,  4.48it/s][INFO|trainer.py:4309] 2025-10-22 16:05:25,446 >> Saving model checkpoint to /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:25,628 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:25,629 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|tokenization_utils_base.py:2421] 2025-10-22 16:05:25,782 >> chat template saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/chat_template.jinja
+[INFO|tokenization_utils_base.py:2590] 2025-10-22 16:05:25,804 >> tokenizer config file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2599] 2025-10-22 16:05:25,809 >> Special tokens file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-50/special_tokens_map.json
+ 51%|     | 51/100 [00:13<00:24,  2.01it/s] 52%|    | 52/100 [00:13<00:20,  2.30it/s] 53%|    | 53/100 [00:13<00:17,  2.66it/s] 54%|    | 54/100 [00:14<00:14,  3.23it/s] 55%|    | 55/100 [00:14<00:13,  3.33it/s] 56%|    | 56/100 [00:14<00:11,  3.94it/s] 57%|    | 57/100 [00:14<00:10,  4.02it/s] 58%|    | 58/100 [00:14<00:09,  4.43it/s] 59%|    | 59/100 [00:15<00:08,  5.12it/s] 60%|    | 60/100 [00:15<00:07,  5.52it/s]                                                {'loss': 0.6288, 'grad_norm': 0.4911618232727051, 'learning_rate': 2.05e-05, 'epoch': 0.0}
+ 60%|    | 60/100 [00:15<00:07,  5.52it/s] 61%|    | 61/100 [00:15<00:07,  5.14it/s] 62%|   | 62/100 [00:15<00:06,  5.71it/s] 63%|   | 63/100 [00:15<00:07,  5.15it/s] 64%|   | 64/100 [00:15<00:06,  5.44it/s] 65%|   | 65/100 [00:16<00:07,  4.97it/s] 66%|   | 66/100 [00:16<00:07,  4.59it/s] 67%|   | 67/100 [00:16<00:06,  4.86it/s] 68%|   | 68/100 [00:16<00:07,  4.54it/s] 69%|   | 69/100 [00:17<00:07,  4.15it/s] 70%|   | 70/100 [00:17<00:07,  4.10it/s]                                                {'loss': 0.6135, 'grad_norm': 0.5213523507118225, 'learning_rate': 1.55e-05, 'epoch': 0.01}
+ 70%|   | 70/100 [00:17<00:07,  4.10it/s] 71%|   | 71/100 [00:17<00:07,  3.97it/s] 72%|  | 72/100 [00:17<00:06,  4.49it/s] 73%|  | 73/100 [00:18<00:06,  4.00it/s] 74%|  | 74/100 [00:18<00:06,  4.19it/s] 75%|  | 75/100 [00:18<00:05,  4.70it/s] 76%|  | 76/100 [00:18<00:05,  4.73it/s] 77%|  | 77/100 [00:18<00:04,  5.24it/s] 78%|  | 78/100 [00:19<00:04,  4.69it/s] 79%|  | 79/100 [00:19<00:04,  4.55it/s] 80%|  | 80/100 [00:19<00:04,  4.27it/s]                                                {'loss': 0.6435, 'grad_norm': 0.4015622138977051, 'learning_rate': 1.05e-05, 'epoch': 0.01}
+ 80%|  | 80/100 [00:19<00:04,  4.27it/s] 81%|  | 81/100 [00:19<00:04,  4.66it/s] 82%| | 82/100 [00:20<00:04,  4.06it/s] 83%| | 83/100 [00:20<00:03,  4.45it/s] 84%| | 84/100 [00:20<00:03,  4.40it/s] 85%| | 85/100 [00:20<00:03,  4.45it/s] 86%| | 86/100 [00:20<00:02,  5.05it/s] 87%| | 87/100 [00:21<00:02,  5.18it/s] 88%| | 88/100 [00:21<00:02,  4.87it/s] 89%| | 89/100 [00:21<00:02,  4.59it/s] 90%| | 90/100 [00:21<00:01,  5.20it/s]                                                {'loss': 0.6313, 'grad_norm': 0.5440771579742432, 'learning_rate': 5.500000000000001e-06, 'epoch': 0.01}
+ 90%| | 90/100 [00:21<00:01,  5.20it/s] 91%| | 91/100 [00:21<00:01,  4.63it/s] 92%|| 92/100 [00:22<00:01,  4.52it/s] 93%|| 93/100 [00:22<00:01,  4.74it/s] 94%|| 94/100 [00:22<00:01,  4.69it/s] 95%|| 95/100 [00:22<00:01,  4.78it/s] 96%|| 96/100 [00:23<00:00,  4.42it/s] 97%|| 97/100 [00:23<00:00,  3.84it/s] 98%|| 98/100 [00:23<00:00,  4.26it/s] 99%|| 99/100 [00:23<00:00,  4.53it/s]100%|| 100/100 [00:23<00:00,  4.30it/s]                                                 {'loss': 0.6241, 'grad_norm': 0.43957775831222534, 'learning_rate': 5.000000000000001e-07, 'epoch': 0.01}
+100%|| 100/100 [00:23<00:00,  4.30it/s][INFO|trainer.py:4309] 2025-10-22 16:05:37,200 >> Saving model checkpoint to /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-100
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:37,345 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:37,346 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|tokenization_utils_base.py:2421] 2025-10-22 16:05:37,515 >> chat template saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-100/chat_template.jinja
+[INFO|tokenization_utils_base.py:2590] 2025-10-22 16:05:37,547 >> tokenizer config file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-100/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2599] 2025-10-22 16:05:37,565 >> Special tokens file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-100/special_tokens_map.json
+[INFO|trainer.py:2810] 2025-10-22 16:05:38,041 >> 
+
+Training completed. Do not forget to share your model on huggingface.co/models =)
+
+
+                                                 {'train_runtime': 25.8748, 'train_samples_per_second': 15.459, 'train_steps_per_second': 3.865, 'train_loss': 0.6805182361602783, 'epoch': 0.01}
+100%|| 100/100 [00:24<00:00,  4.30it/s]100%|| 100/100 [00:24<00:00,  4.02it/s]
+[INFO|trainer.py:4309] 2025-10-22 16:05:38,051 >> Saving model checkpoint to /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:38,140 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:38,141 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|tokenization_utils_base.py:2421] 2025-10-22 16:05:38,274 >> chat template saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/chat_template.jinja
+[INFO|tokenization_utils_base.py:2590] 2025-10-22 16:05:38,280 >> tokenizer config file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2599] 2025-10-22 16:05:38,283 >> Special tokens file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/special_tokens_map.json
+***** train metrics *****
+  epoch                    =     0.0082
+  total_flos               =  1473847GF
+  train_loss               =     0.6805
+  train_runtime            = 0:00:25.87
+  train_samples_per_second =     15.459
+  train_steps_per_second   =      3.865
+[INFO|modelcard.py:456] 2025-10-22 16:05:38,469 >> Dropping the following result as it does not have all the necessary fields:
+{'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}}
+gl064:2370173:2370173 [1] NCCL INFO comm 0x16d00aa0 rank 1 nranks 4 cudaDev 1 busId 59000 - Destroy COMPLETE
+gl064:2370172:2370172 [0] NCCL INFO comm 0x14055870 rank 0 nranks 4 cudaDev 0 busId 47000 - Destroy COMPLETE
+[1;34mwandb[0m: 
+[1;34mwandb[0m:  View run [33minteractive_test[0m at: [34m[0m
+[1;34mwandb[0m: Find logs at: [1;35mwandb/run-20251022_160512-dppinxzz/logs[0m
+
+========================================
+Training completed successfully
+End Time: Wed Oct 22 04:05:41 PM EDT 2025
+========================================
+
+========================================
+STAGE 2: Merging/Exporting Model
+Start Time: Wed Oct 22 04:05:41 PM EDT 2025
+========================================
+Looking for checkpoints in: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
+Found most recent checkpoint: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-150
+Checkpoint details:
+  Path: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-150
+  Last modified: 2025-10-22 16:02:30.204175325 -0400
+======================================
+PIPELINE COMPLETED SUCCESSFULLY
+End Time: Wed Oct 22 04:05:41 PM EDT 2025
+========================================
+Successfully updated merge config
+=======
+Cleaning up LlamaFactory processes
+========================================
+Cleaned up processes on gl065.hpc.nyu.edu
+Process cleanup complete
+Updated merge config to use: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-150
+
+Merge config contents:
+  model_name_or_path: Qwen/Qwen2.5-0.5B
+  finetuning_type: lora
+  trust_remote_code: true
+  adapter_name_or_path: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-150
+  template: default
+  export_dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged
+
+Executing command: llamafactory-cli export /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/configs/merge_config.yaml
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/transformers/utils/hub.py:110: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
+  warnings.warn(
+/scratch/zrs2020/miniconda/miniconda3/envs/llamafactory/lib/python3.12/site-packages/jieba/_compat.py:18: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
+  import pkg_resources
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,842 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,842 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,842 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,842 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,843 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,843 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:49,843 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:05:50,013 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:50,242 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:50,244 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file vocab.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/vocab.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file merges.txt from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/merges.txt
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file tokenizer.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file added_tokens.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file special_tokens_map.json from cache at None
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file tokenizer_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2095] 2025-10-22 16:05:50,336 >> loading file chat_template.jinja from cache at None
+[INFO|tokenization_utils_base.py:2364] 2025-10-22 16:05:50,499 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+[INFO|configuration_utils.py:765] 2025-10-22 16:05:50,543 >> loading configuration file config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/config.json
+[INFO|configuration_utils.py:839] 2025-10-22 16:05:50,544 >> Model config Qwen2Config {
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+[WARNING|logging.py:328] 2025-10-22 16:05:50,544 >> `torch_dtype` is deprecated! Use `dtype` instead!
+[INFO|2025-10-22 16:05:50] llamafactory.model.model_utils.kv_cache:143 >> KV cache is enabled for faster generation.
+[WARNING|logging.py:328] 2025-10-22 16:05:50,858 >> `torch_dtype` is deprecated! Use `dtype` instead!
+[INFO|modeling_utils.py:1172] 2025-10-22 16:05:50,859 >> loading weights file model.safetensors from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/model.safetensors
+[INFO|modeling_utils.py:2341] 2025-10-22 16:05:50,859 >> Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
+[INFO|configuration_utils.py:986] 2025-10-22 16:05:50,860 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643
+}
+
+[INFO|configuration_utils.py:941] 2025-10-22 16:05:50,951 >> loading configuration file generation_config.json from cache at /scratch/zrs2020/.cache/hf_cache/home/hub/models--Qwen--Qwen2.5-0.5B/snapshots/060db6499f32faf8b98477b0a26969ef7d8b9987/generation_config.json
+[INFO|configuration_utils.py:986] 2025-10-22 16:05:50,952 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048
+}
+
+[INFO|dynamic_module_utils.py:423] 2025-10-22 16:05:50,980 >> Could not locate the custom_generate/generate.py inside Qwen/Qwen2.5-0.5B.
+[INFO|2025-10-22 16:05:50] llamafactory.model.model_utils.attention:143 >> Using torch SDPA for faster training and inference.
+[INFO|2025-10-22 16:05:52] llamafactory.model.adapter:143 >> Merged 1 adapter(s).
+[INFO|2025-10-22 16:05:52] llamafactory.model.adapter:143 >> Loaded adapter(s): /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints/checkpoint-150
+[INFO|2025-10-22 16:05:52] llamafactory.model.loader:143 >> all params: 494,032,768
+[INFO|2025-10-22 16:05:52] llamafactory.train.tuner:143 >> Convert model dtype to: torch.bfloat16.
+[INFO|configuration_utils.py:491] 2025-10-22 16:05:52,089 >> Configuration saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/config.json
+[INFO|configuration_utils.py:757] 2025-10-22 16:05:52,094 >> Configuration saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/generation_config.json
+[INFO|modeling_utils.py:4181] 2025-10-22 16:05:53,693 >> Model weights saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/model.safetensors
+[INFO|tokenization_utils_base.py:2421] 2025-10-22 16:05:53,698 >> chat template saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/chat_template.jinja
+[INFO|tokenization_utils_base.py:2590] 2025-10-22 16:05:53,702 >> tokenizer config file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2599] 2025-10-22 16:05:53,706 >> Special tokens file saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/special_tokens_map.json
+[INFO|2025-10-22 16:05:53] llamafactory.train.tuner:143 >> Ollama modelfile saved in /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/merged/Modelfile
+
+========================================
+Merge/Export completed successfully
+End Time: Wed Oct 22 04:05:54 PM EDT 2025
+========================================
+
+========================================
+Preparing Training Artifacts
+========================================
+Copying configuration files...
+Copying and cleaning training logs...