XPK Start: Thu Apr 23 21:28:22 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
`rope_parameters`'s factor field must be a float >= 1, got 40
`rope_parameters`'s beta_fast field must be a float, got 32
`rope_parameters`'s beta_slow field must be a float, got 1
DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config.
2026-04-23 21:28:47.350067: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0423 21:28:47.561401 138173694007104 max_utils.py:273] Attempting to initialize the jax distributed system...
I0423 21:28:56.602909 138173694007104 distributed.py:149] Starting JAX distributed service on [::]:8482
I0423 21:28:56.605231 138173694007104 distributed.py:172] Connecting to JAX distributed service on mt-05-fp8-0swzl-slice-job-0-0.mt-05-fp8-0swzl:8482
I0423 21:28:58.241075 138173694007104 max_utils.py:284] Jax distributed system initialized!
I0423 21:29:04.692531 138173694007104 max_utils.py:800] System Information: Jax Version: 0.9.2
I0423 21:29:04.692636 138173694007104 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0423 21:29:04.692677 138173694007104 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0423 21:29:04.692712 138173694007104 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 21:29:05.388070 138173694007104 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 21:29:05.388374 138173694007104 checkpointing.py:677] Setting up checkpoint logger...
I0423 21:29:05.388429 138173694007104 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 21:29:05.388473 138173694007104 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0423 21:29:05.388826 138173694007104 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7daa7b71c110>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 21:29:08.349580 138173694007104 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0423 21:29:08.349821 138173694007104 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d96382bb590>}, handler_registry=None
I0423 21:29:08.350059 138173694007104 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d96382bb590>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0423 21:29:08.350120 138173694007104 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9638422600>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0423 21:29:08.350160 138173694007104 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d96382bb590>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d96382bb590>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9638422600>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9638422600>}).
I0423 21:29:08.350484 138173694007104 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0423 21:29:08.350554 138173694007104 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d963826b560> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0423 21:29:09.509406 138173694007104 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints
I0423 21:29:10.356676 138173694007104 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d9638420d10>
I0423 21:29:10.356848 138173694007104 checkpointing.py:301] Checkpoint manager created!
I0423 21:29:11.442291 138173694007104 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 21:29:11.442415 138173694007104 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 21:29:11.829390 138173694007104 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 21:29:11.829483 138173694007104 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 21:29:11.846060 138173694007104 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 21:29:11.846133 138173694007104 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 21:29:11.899389 138173694007104 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 21:29:11.899471 138173694007104 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 21:29:11.916150 138173694007104 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 21:29:11.916214 138173694007104 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 21:29:11.932751 138173694007104 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 21:29:11.932810 138173694007104 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 21:29:11.957737 138173694007104 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 21:29:11.957807 138173694007104 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 21:29:12.009722 138173694007104 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 21:29:12.009804 138173694007104 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 21:29:12.455907 138173694007104 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 21:29:12.456024 138173694007104 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 21:29:14.852885 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853010 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853059 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853136 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853177 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853214 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853246 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853277 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853310 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853340 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853370 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853400 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853433 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853462 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853496 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853525 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853554 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853582 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853669 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853698 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853726 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853753 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853779 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853805 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853832 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853857 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853884 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853909 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853937 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.853964 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.853990 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854016 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854043 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854069 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854105 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854135 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854161 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854187 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854214 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854240 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854268 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0423 21:29:14.854294 138173694007104 maxtext_utils.py:1668] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0423 21:29:14.854343 138173694007104 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0423 21:29:14.854397 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0423 21:29:14.854434 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0423 21:29:14.854480 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0423 21:29:14.854535 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0423 21:29:14.854568 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0423 21:29:14.854616 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0423 21:29:14.854665 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0423 21:29:14.854700 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0423 21:29:14.854733 138173694007104 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0423 21:29:14.854777 138173694007104 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0423 21:29:14.854819 138173694007104 maxtext_utils.py:1668] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0423 21:29:16.046983 138173694007104 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 21:29:16.047081 138173694007104 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 21:29:16.062841 138173694007104 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 21:29:16.062902 138173694007104 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 21:29:29.685846 138173694007104 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0423 21:29:29.686708 138173694007104 metric_logger.py:301] number parameters: 1.105 billion
I0423 21:29:44.754839 138173694007104 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0423 21:29:44.924507 138173694007104 checkpointing.py:776] Waited 0.16964387893676758 seconds for step 0 to finish before starting checkpointing.
I0423 21:29:44.927078 138173694007104 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 21:29:44.929061 138173694007104 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0423 21:29:44.930549 138173694007104 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/0.
I0423 21:29:45.253431 138173694007104 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 21:29:45.254482 138173694007104 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0423 21:29:45.254546 138173694007104 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0423 21:29:45.684858 138173694007104 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.431548s
I0423 21:29:45.685034 138173694007104 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 3.507 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.4398825168609619 s) (per-host)
I0423 21:29:45.685087 138173694007104 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.439946s (batch_requests_ready=0.003626s, total_serialization_initiated=0.436248s, others=0.000072s)
I0423 21:29:45.685212 138173694007104 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.444055s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.444036)
I0423 21:29:45.686043 138173694007104 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/0.
I0423 21:29:45.686434 138044519413504 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 21:49:45.686394
I0423 21:29:45.688560 138173694007104 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 21:29:45.688837 138044006987520 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 21:29:45.688996 138173694007104 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776979784.9270597, 'wait_for_prev_duration_secs': 7.581710815429688e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776979784.9291258, 'checkpointer_blocking_duration_secs': 0.7574687004089355, 'get_old_steps_start_time': 1776979785.686624, 'get_old_steps_duration_secs': 3.4332275390625e-05, 'checkpoint_manager_blocking_start_time': 1776979784.9251847, 'checkpoint_manager_blocking_duration_secs': 0.7637679576873779}
I0423 21:29:45.689139 138173694007104 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0423 21:29:45.689203 138173694007104 max_utils.py:750]
Memstats: After params initialized:
I0423 21:29:45.689265 138173694007104 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_24(process=6,(0,6,0,0))
I0423 21:29:45.689319 138173694007104 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_25(process=6,(1,6,0,0))
I0423 21:29:45.689361 138173694007104 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_28(process=6,(0,7,0,0))
I0423 21:29:45.689394 138173694007104 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_29(process=6,(1,7,0,0))
I0423 21:29:46.006466 138173694007104 metric_logger.py:196] completed step: 0, seconds: 15.068, TFLOP/s/device: 0.902, Tokens/s/device: 135.917, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52776.805
I0423 21:29:46.191496 138173694007104 metric_logger.py:196] completed step: 1, seconds: 1.250, TFLOP/s/device: 10.873, Tokens/s/device: 1638.968, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52761.840
I0423 21:29:46.622367 138173694007104 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 494.311, Tokens/s/device: 74507.949, total_weights: 65536, loss: 10.816, lm_loss: 10.816, perplexity: 49832.398
I0423 21:29:46.780795 138173694007104 metric_logger.py:196] completed step: 3, seconds: 0.431, TFLOP/s/device: 31.522, Tokens/s/device: 4751.266, total_weights: 65536, loss: 10.431, lm_loss: 10.431, perplexity: 33901.820
I0423 21:29:47.097927 138173694007104 metric_logger.py:196] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.781, Tokens/s/device: 12477.686, total_weights: 65536, loss: 9.992, lm_loss: 9.992, perplexity: 21847.639
I0423 21:29:47.105695 138173694007104 metric_logger.py:196] completed step: 5, seconds: 0.158, TFLOP/s/device: 85.812, Tokens/s/device: 12934.519, total_weights: 65536, loss: 9.549, lm_loss: 9.549, perplexity: 14035.006
I0423 21:29:49.350770 2818 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 21:29:51.887428 138044502628096 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/0/items/array_metadatas/process_6
I0423 21:30:06.024908 138044519413504 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 76.024 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 20.779733180999756 s) (per-host)
I0423 21:30:06.025018 138044519413504 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 20.338470s.
I0423 21:30:11.657563 138173694007104 metric_logger.py:196] completed step: 6, seconds: 0.319, TFLOP/s/device: 42.627, Tokens/s/device: 6425.259, total_weights: 65536, loss: 9.111, lm_loss: 9.111, perplexity: 9058.688
I0423 21:30:11.815990 138173694007104 metric_logger.py:196] completed step: 7, seconds: 24.395, TFLOP/s/device: 0.557, Tokens/s/device: 83.952, total_weights: 65536, loss: 8.685, lm_loss: 8.685, perplexity: 5914.720
I0423 21:30:11.974451 138173694007104 metric_logger.py:196] completed step: 8, seconds: 0.163, TFLOP/s/device: 83.105, Tokens/s/device: 12526.454, total_weights: 65536, loss: 8.281, lm_loss: 8.281, perplexity: 3950.010
I0423 21:30:11.979514 138173694007104 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0423 21:30:12.290930 138173694007104 checkpointing.py:776] Waited 0.31137919425964355 seconds for step 10 to finish before starting checkpointing.
I0423 21:30:12.293783 138173694007104 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 21:30:13.938523 138044519413504 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 28.251959s.
I0423 21:30:13.938820 138044006987520 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 21:30:13.938941 138044006987520 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 21:30:13.938991 138044006987520 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 21:30:13.940731 138044006987520 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 21:30:13.940909 138173694007104 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0423 21:30:13.941038 138173694007104 checkpoint_manager.py:1452] Waiting for previous save to complete took 1.647255 seconds. If this number is high, consider checkpointing less frequently.
I0423 21:30:13.943023 138173694007104 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 10
I0423 21:30:13.944974 138173694007104 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/10.
I0423 21:30:14.656936 138173694007104 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0423 21:30:14.657030 138173694007104 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0423 21:30:14.709826 138173694007104 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.053796s
I0423 21:30:14.710003 138173694007104 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 25.665 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.06011033058166504 s) (per-host)
I0423 21:30:14.710053 138173694007104 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.060172s (batch_requests_ready=0.003163s, total_serialization_initiated=0.056940s, others=0.000069s)
I0423 21:30:14.710168 138173694007104 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.064327s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.064312)
I0423 21:30:14.710937 138173694007104 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/10.
I0423 21:30:14.711256 138040743798528 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 21:50:14.711220
I0423 21:30:14.713367 138173694007104 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 21:30:14.713721 138044006987520 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 21:30:14.713880 138173694007104 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776979812.2937527, 'wait_for_prev_duration_secs': 1.6472549438476562, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776979813.943064, 'checkpointer_blocking_duration_secs': 0.7683379650115967, 'get_old_steps_start_time': 1776979814.7114267, 'get_old_steps_duration_secs': 3.6716461181640625e-05, 'checkpoint_manager_blocking_start_time': 1776979812.2917056, 'checkpoint_manager_blocking_duration_secs': 2.42214035987854}
I0423 21:30:14.713999 138173694007104 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0423 21:30:14.714619 138173694007104 metric_logger.py:196] completed step: 9, seconds: 0.158, TFLOP/s/device: 85.893, Tokens/s/device: 12946.702, total_weights: 65536, loss: 7.908, lm_loss: 7.908, perplexity: 2720.116
I0423 21:30:14.721850 138173694007104 metric_logger.py:196] completed step: 10, seconds: 0.159, TFLOP/s/device: 85.672, Tokens/s/device: 12913.396, total_weights: 65536, loss: 7.568, lm_loss: 7.568, perplexity: 1936.198
I0423 21:30:14.879928 138173694007104 metric_logger.py:196] completed step: 11, seconds: 2.740, TFLOP/s/device: 4.959, Tokens/s/device: 747.410, total_weights: 65536, loss: 7.288, lm_loss: 7.288, perplexity: 1463.272
I0423 21:30:15.450479 138173694007104 metric_logger.py:196] completed step: 12, seconds: 0.007, TFLOP/s/device: 2073.738, Tokens/s/device: 312576.313, total_weights: 65536, loss: 7.032, lm_loss: 7.032, perplexity: 1132.061
I0423 21:30:15.608800 138173694007104 metric_logger.py:196] completed step: 13, seconds: 0.566, TFLOP/s/device: 24.002, Tokens/s/device: 3617.895, total_weights: 65536, loss: 6.815, lm_loss: 6.815, perplexity: 911.031
I0423 21:30:15.767167 138173694007104 metric_logger.py:196] completed step: 14, seconds: 0.164, TFLOP/s/device: 83.088, Tokens/s/device: 12524.002, total_weights: 65536, loss: 6.635, lm_loss: 6.635, perplexity: 761.080
I0423 21:30:15.925462 138173694007104 metric_logger.py:196] completed step: 15, seconds: 0.158, TFLOP/s/device: 85.982, Tokens/s/device: 12960.139, total_weights: 65536, loss: 6.492, lm_loss: 6.492, perplexity: 659.915
I0423 21:30:16.083724 138173694007104 metric_logger.py:196] completed step: 16, seconds: 0.159, TFLOP/s/device: 85.687, Tokens/s/device: 12915.676, total_weights: 65536, loss: 6.380, lm_loss: 6.380, perplexity: 589.948
I0423 21:30:16.241898 138173694007104 metric_logger.py:196] completed step: 17, seconds: 0.158, TFLOP/s/device: 85.988, Tokens/s/device: 12961.041, total_weights: 65536, loss: 6.293, lm_loss: 6.293, perplexity: 540.941
I0423 21:30:16.400222 138173694007104 metric_logger.py:196] completed step: 18, seconds: 0.158, TFLOP/s/device: 85.834, Tokens/s/device: 12937.869, total_weights: 65536, loss: 6.223, lm_loss: 6.223, perplexity: 504.108
I0423 21:30:16.557904 138173694007104 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0423 21:30:16.558923 138173694007104 checkpointing.py:776] Waited 0.0010390281677246094 seconds for step 19 to finish before starting checkpointing.
I0423 21:30:16.561089 138173694007104 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 21:30:20.908891 138040240498432 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/10/items/array_metadatas/process_6
I0423 21:30:57.875553 138040743798528 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 36.547 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.22562026977539 s) (per-host)
I0423 21:30:57.875664 138040743798528 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.164290s.
I0423 21:31:06.955408 138040743798528 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.244018s.
I0423 21:31:06.955696 138044006987520 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 21:31:06.955819 138044006987520 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 21:31:06.955869 138044006987520 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0423 21:31:06.957647 138044006987520 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0423 21:31:06.957885 138173694007104 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0423 21:31:06.958011 138173694007104 checkpoint_manager.py:1452] Waiting for previous save to complete took 50.396928 seconds. If this number is high, consider checkpointing less frequently.
I0423 21:31:06.959654 138173694007104 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 19
I0423 21:31:06.961662 138173694007104 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/19.
I0423 21:31:07.265873 138173694007104 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0423 21:31:07.265966 138173694007104 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0423 21:31:07.318536 138173694007104 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.053699s
I0423 21:31:07.318706 138173694007104 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 25.808 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.05977678298950195 s) (per-host)
I0423 21:31:07.318754 138173694007104 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.059835s (batch_requests_ready=0.002991s, total_serialization_initiated=0.056781s, others=0.000063s)
I0423 21:31:07.318854 138173694007104 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.063975s (all_items=0.000011s, per_item={'items': '0.00001121'}, temp_paths=0.063964)
I0423 21:31:07.319558 138173694007104 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.36 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/19.
I0423 21:31:07.319900 138049375696640 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 21:51:07.319858
I0423 21:31:07.321921 138173694007104 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 21:31:07.322300 138044006987520 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 21:31:07.322465 138173694007104 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776979816.5610585, 'wait_for_prev_duration_secs': 50.39692759513855, 'time_between_consecutive_saves_sec': 2.6202900409698486, 'checkpointer_blocking_start_time': 1776979866.9596927, 'checkpointer_blocking_duration_secs': 0.3603546619415283, 'get_old_steps_start_time': 1776979867.3200653, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776979816.5592132, 'checkpoint_manager_blocking_duration_secs': 50.76322078704834}
I0423 21:31:07.322584 138173694007104 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0423 21:31:07.322626 138173694007104 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 21:31:13.792618 138042908067584 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_05_fp8/checkpoints/19/items/array_metadatas/process_6
I0423 21:31:48.674048 138049375696640 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.145 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.41507029533386 s) (per-host)
I0423 21:31:48.674201 138049375696640 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.354186s.
I0423 21:31:56.258189 138049375696640 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 48.938157s.
I0423 21:31:56.258532 138044006987520 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 21:31:56.258680 138044006987520 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 21:31:56.258723 138044006987520 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0423 21:31:56.260475 138044006987520 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0423 21:31:56.260659 138173694007104 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0423 21:31:56.260808 138173694007104 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 21:31:56.261746 138173694007104 metric_logger.py:196] completed step: 19, seconds: 0.158, TFLOP/s/device: 85.843, Tokens/s/device: 12939.177, total_weights: 65536, loss: 6.167, lm_loss: 6.167, perplexity: 476.947
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Thu Apr 23 21:32:07 UTC 2026
EXIT_CODE=0