q3-8b-train_final_v2_nb2_mt…/slurm/9168628.0.err

./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
./slurm_setup.sh: line 48: GLOBAL_TRAIN_BATCH_SIZE / NNODES / GPUS_PER_NODE / GRADIENT_ACCUMULATION_STEPS: division by 0 (error token is "GRADIENT_ACCUMULATION_STEPS")
2026-04-13 22:59:24,062 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,062 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,062 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,062 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,066 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,066 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,066 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,066 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,066 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,067 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,068 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,068 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,068 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,068 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,079 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,079 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,079 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,079 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,079 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,079 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,080 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,080 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,085 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,085 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,086 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,086 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,086 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,087 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,086 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,087 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,088 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,088 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,089 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,090 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,089 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,090 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,094 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,094 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,105 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,106 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,106 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,106 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,108 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,108 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,114 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,114 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,117 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,117 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,124 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,124 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,125 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,125 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,125 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,125 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,131 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,131 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,136 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,136 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,139 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,139 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,147 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,147 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,148 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,148 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,154 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,153 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,154 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,153 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,157 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,157 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,158 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,158 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,159 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,159 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,164 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,164 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,165 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,165 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,168 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,168 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,177 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,177 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,178 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,178 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,180 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,180 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,188 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,188 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,191 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,191 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,191 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,191 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,196 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,196 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,214 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,214 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,275 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,275 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,304 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,304 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,304 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,304 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,304 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,304 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,312 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,312 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,312 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,312 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,346 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,346 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:24,347 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:24,347 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:24,391 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,392 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,394 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,395 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,395 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,398 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,398 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,400 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,401 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,403 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,404 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,410 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,418 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,417 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,418 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,423 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,434 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,435 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,435 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,438 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,437 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,441 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,441 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,442 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,444 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,446 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,450 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,459 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,461 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,471 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,472 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,474 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,479 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,478 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,488 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,499 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,502 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,502 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,504 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,505 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,507 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,511 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,513 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,514 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,517 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,530 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,530 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,539 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,608 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,631 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,632 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,633 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,633 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,637 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,670 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:24,671 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
2026-04-13 22:59:26,936 - INFO - Note: detected 255 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
2026-04-13 22:59:26,936 - INFO - Note: NumExpr detected 255 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
/home/ligengz/anaconda3/lib/python3.11/site-packages/pandas/core/arrays/masked.py:60: UserWarning: Pandas requires version '1.3.6' or newer of 'bottleneck' (version '1.3.5' currently installed).
  from pandas.core import (
2026-04-13 22:59:28,052 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:28,055 - INFO - PyTorch version 2.6.0 available.
2026-04-13 22:59:42,580 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,594 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,603 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,605 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,613 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,614 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,625 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,644 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,680 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,724 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,738 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,739 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_2026-04-13 22:59:42,739 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factsteps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_fac
: None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': True, 'eval_use_gather_object': False, 'average_tokens_across_devices': True, 'model_init_kwargs': None, 'chat_template_path': None, 'dataset_text_field': 'qwen_text', 'dataset_kwargs': None, 'dataset_num_proc': None, 'eos_token': None, 'pad_token': None, 'max_length': 1024, 'packing': False, 'packing_strategy': 'ffd', 'padding_free': False, 'pad_to_multiple_of': None, 'eval_packing': None, 'completion_only_loss': None, 'assistant_only_loss': False, 'activation_offloading': False, 'max_seq_length': None}
2026-04-13 22:59:42,756 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_2026-04-13 22:59:42,756 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None
steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': True, 'eval_use_gather_object': False, 'average_tokens_across_devices': True, 'model_init_kwargs': None, 'chat_template_path': None, 'dataset_text_field': 'qwen_text', 'dataset_kwargs': None, 'dataset_num_proc': None, 'eos_token': None, 'pad_token': None, 'max_length': 1024, 'packing': False, 'packing_strategy': 'ffd', 'padding_free': False, 'pad_to_multiple_of': None, 'eval_packing': None, 'completion_only_loss': None, 'assistant_only_loss': False, 'activation_offloading': False, 'max_seq_length': None}
2026-04-13 22:59:42,759 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:42,760 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-39_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,289 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,293 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,297 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,328 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,346 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,351 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,352 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,369 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,373 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,382 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,384 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,384 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,388 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,389 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,397 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,401 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:47,492 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,218 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,242 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,262 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,280 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-1089', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,284 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,284 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3038', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,304 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_2026-04-13 22:59:52,304 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None
steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': True, 'eval_use_gather_object': False, 'average_tokens_across_devices': True, 'model_init_kwargs': None, 'chat_template_path': None, 'dataset_text_field': 'qwen_text', 'dataset_kwargs': None, 'dataset_num_proc': None, 'eos_token': None, 'pad_token': None, 'max_length': 1024, 'packing': False, 'packing_strategy': 'ffd', 'padding_free': False, 'pad_to_multiple_of': None, 'eval_packing': None, 'completion_only_loss': None, 'assistant_only_loss': False, 'activation_offloading': False, 'max_seq_length': None}
2026-04-13 22:59:52,304 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3544', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,305 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,306 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3035', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,323 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,328 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,330 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,330 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,333 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,333 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-40_batch-block1-3163', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,348 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,349 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,359 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,361 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3736', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 22:59:52,367 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_22-59-41_batch-block1-3305', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:04,994 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,109 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 7, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,110 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 1, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,129 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 3, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,130 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 6, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,149 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 4, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,150 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 5, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules
2026-04-13 23:00:05,152 - INFO - Training config: {'model_name': 'runs/dev/tw-8node-resume/checkpoint-64', 'template_name': 'qwen', 'block_size': 40960, 'wandb_project': 'ThreadWeaver', 'train_file_path': './data-04-13', 'dagger': False, 'attn_implementation': 'flex_attention', 'output_dir': 'runs/dev/tw-8node-resume', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': <IntervalStrategy.NO: 'no'>, 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 2, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 1e-05, 'weight_decay': 0.0001, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 8.0, 'max_steps': -1, 'lr_scheduler_type': <SchedulerType.COSINE: 'cosine'>, 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.05, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': 'runs/dev/tw-8node-resume/runs/Apr13_23-00-03_batch-block1-3061', 'logging_strategy': <IntervalStrategy.STEPS: 'steps'>, 'logging_first_step': False, 'logging_steps': 1.0, 'logging_nan_inf_filter': True, 'save_strategy': <SaveStrategy.STEPS: 'steps'>, 'save_steps': 20, 'save_total_limit': 2, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': None, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 2, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'runs/dev/tw-8node-resume', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, 'deepspeed': 'configs/deepspeed_zero3.json', 'label_smoothing_factor': 0.0, 'optim': <OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': 'runs/dev/tw-8node-resume/checkpoint-64', 'hub_model_id': None, 'hub_strategy': <HubStrategy.EVERY_SAVE: 'every_save'>, 'hub_token': None, 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': None, '_n_gpu': 1, 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules

Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.84s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.82s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.84s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.82s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.82s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.82s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:03<00:11,  3.83s/it]
Loading checkpoint shards:  
00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
<00:00,  3.34s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.80s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.34s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  2.82s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:13<00:00,  3.37s/it]

Loading checkpoint shards:  75%|███████▌  | 3/4 [00:13<00:04,  4.57s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:15<00:00,  3.37s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:15<00:00,  3.77s/it]
2026-04-13 23:00:26,972 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,974 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,972 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,974 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,975 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,974 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,975 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,976 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,978 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,976 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,977 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,977 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,977 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,978 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,979 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,980 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,981 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,980 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,981 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,980 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,981 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,982 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,981 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,981 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,982 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,982 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,981 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,981 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,982 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,983 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,982 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,982 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,983 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,984 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,983 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,984 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,984 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,984 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,984 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,985 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,984 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,984 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,984 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,985 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,986 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,985 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,986 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,985 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,986 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,988 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,987 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,988 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,988 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,988 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,988 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,989 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,990 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,989 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,990 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,991 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,991 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,990 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,991 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,991 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,992 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,991 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,992 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,992 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,992 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,993 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,993 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,993 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,994 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,993 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,993 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,995 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,994 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,995 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,994 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,996 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,995 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,997 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,998 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,998 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,997 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:26,998 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,999 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,998 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:26,999 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,001 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,000 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,001 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,000 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,001 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,002 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,001 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,002 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,004 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,006 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,004 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,008 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,010 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,015 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,017 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,021 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,025 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,029 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,044 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,045 - INFO - Applying Liger kernels to model instance with model type: qwen3 with kwargs: {}
2026-04-13 23:00:27,053 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,055 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,056 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,058 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,058 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,058 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,058 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
2026-04-13 23:00:27,058 - INFO - Resuming trainer state from checkpoint: runs/dev/tw-8node-resume/checkpoint-64
wandb: WARNING The `run_name` is currently set to the same value as `TrainingArguments.output_dir`. If this was not intended, please specify a different run name by setting the `TrainingArguments.run_name` parameter.
wandb: Currently logged in as: ligeng-zhu to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.21.0
wandb: Run data is saved locally in runs/dev/tw-8node-resume/wandb/run-20260413_230036-tw-8node-resume
wandb: Run `wandb offline` to turn off syncing.
wandb: Resuming run runs/dev/tw-8node-resume
wandb: ⭐️ View project at https://wandb.ai/ligeng-zhu/ThreadWeaver
wandb: 🚀 View run at https://wandb.ai/ligeng-zhu/ThreadWeaver/runs/tw-8node-resume
A ConfigError was raised whilst setting the number of model parameters in Weights & Biases config.

  0%|          | 0/64 [00:00<?, ?it/s]
                                      

  0%|          | 0/64 [00:00<?, ?it/s]
  0%|          | 0/64 [00:00<?, ?it/s]