XPK Start: Sat Apr 18 13:12:59 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 13:13:23.946886: 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)
I0418 13:13:24.489770 138273875728192 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 13:13:33.531255 138273875728192 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 13:13:33.533557 138273875728192 distributed.py:172] Connecting to JAX distributed service on mt-05-fp8-yrrda-slice-job-0-0.mt-05-fp8-yrrda:8482
I0418 13:13:35.146423 138273875728192 max_utils.py:284] Jax distributed system initialized!
I0418 13:13:41.206727 138273875728192 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 13:13:41.206843 138273875728192 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 13:13:41.206889 138273875728192 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
I0418 13:13:41.206926 138273875728192 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 13:13:41.900604 138273875728192 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 13:13:41.900903 138273875728192 checkpointing.py:677] Setting up checkpoint logger...
I0418 13:13:41.900958 138273875728192 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 13:13:41.901002 138273875728192 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 13:13:41.901340 138273875728192 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 0x7dc1c431c440>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 13:13:45.199465 138273875728192 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 13:13:45.199703 138273875728192 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 0x7dad6c317500>}, handler_registry=None
I0418 13:13:45.199949 138273875728192 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dad6c317500>` 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`.
I0418 13:13:45.199999 138273875728192 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dad6c318f80>` 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`.
I0418 13:13:45.200038 138273875728192 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 0x7dad6c317500>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dad6c317500>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dad6c318f80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dad6c318f80>}).
I0418 13:13:45.200352 138273875728192 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 13:13:45.200420 138273875728192 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7dad346dbb00> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 13:13:46.866624 138273875728192 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints
I0418 13:13:47.744201 138273875728192 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_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7dad6c317f50>
I0418 13:13:47.744371 138273875728192 checkpointing.py:301] Checkpoint manager created!
I0418 13:13:48.828445 138273875728192 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 13:13:48.828572 138273875728192 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:13:49.216229 138273875728192 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:13:49.216317 138273875728192 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:13:49.232527 138273875728192 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:13:49.232585 138273875728192 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:13:49.284560 138273875728192 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:13:49.284638 138273875728192 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:13:49.301015 138273875728192 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:13:49.301077 138273875728192 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:13:49.317251 138273875728192 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:13:49.317308 138273875728192 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:13:49.341693 138273875728192 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 13:13:49.341760 138273875728192 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:13:49.393239 138273875728192 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 13:13:49.393319 138273875728192 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 13:13:49.856645 138273875728192 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 13:13:49.856750 138273875728192 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 13:13:51.793357 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793475 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793524 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793568 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793603 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793637 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793669 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793699 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793730 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793760 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793791 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793832 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793867 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793896 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793924 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.793959 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.793991 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794021 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794049 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794076 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794103 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794129 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794156 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794183 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794209 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794236 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794262 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794289 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794315 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794341 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794367 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794394 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794420 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794447 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794473 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794499 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794525 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794556 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794582 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794608 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794635 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_amax_history
Shape: float32[16,1024]
Logical: P()
Physical: ()
I0418 13:13:51.794661 138273875728192 maxtext_utils.py:1654] params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_scale
Shape: float32[16,1]
Logical: P()
Physical: ()
I0418 13:13:51.794710 138273875728192 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 13:13:51.794762 138273875728192 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:13:51.794799 138273875728192 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:13:51.794876 138273875728192 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 13:13:51.794931 138273875728192 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:13:51.794966 138273875728192 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:13:51.795014 138273875728192 maxtext_utils.py:1654] 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)
I0418 13:13:51.795061 138273875728192 maxtext_utils.py:1654] 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')
I0418 13:13:51.795098 138273875728192 maxtext_utils.py:1654] 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)
I0418 13:13:51.795130 138273875728192 maxtext_utils.py:1654] 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)
I0418 13:13:51.795175 138273875728192 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 13:13:51.795219 138273875728192 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 13:13:53.461672 138273875728192 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:13:53.461768 138273875728192 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:13:53.477199 138273875728192 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:13:53.477260 138273875728192 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:14:07.094121 138273875728192 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.
I0418 13:14:07.095052 138273875728192 metric_logger.py:301] number parameters: 1.105 billion
I0418 13:14:22.420875 138273875728192 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 13:14:22.589376 138273875728192 checkpointing.py:776] Waited 0.16847681999206543 seconds for step 0 to finish before starting checkpointing.
I0418 13:14:22.591954 138273875728192 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:14:22.593702 138273875728192 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0418 13:14:22.595248 138273875728192 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/0.
I0418 13:14:23.766569 138273875728192 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 13:14:23.767466 138273875728192 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0418 13:14:23.767526 138273875728192 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
I0418 13:14:24.181702 138273875728192 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.415254s
I0418 13:14:24.181901 138273875728192 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 3.639 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.42395901679992676 s) (per-host)
I0418 13:14:24.181959 138273875728192 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.424028s (batch_requests_ready=0.003821s, total_serialization_initiated=0.420130s, others=0.000077s)
I0418 13:14:24.182067 138273875728192 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.428252s (all_items=0.000039s, per_item={'items': '0.00003862'}, temp_paths=0.428213)
I0418 13:14:24.182897 138273875728192 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.59 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/0.
I0418 13:14:24.183268 138144066303744 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:34:24.183227
I0418 13:14:24.191621 138273875728192 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:14:24.191946 138142992561920 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:14:24.192103 138273875728192 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776518062.5919356, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776518062.593744, 'checkpointer_blocking_duration_secs': 1.5896327495574951, 'get_old_steps_start_time': 1776518064.1834018, 'get_old_steps_duration_secs': 2.7894973754882812e-05, 'checkpoint_manager_blocking_start_time': 1776518062.5900972, 'checkpoint_manager_blocking_duration_secs': 1.6019682884216309}
I0418 13:14:24.192223 138273875728192 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 13:14:24.192273 138273875728192 max_utils.py:750]
Memstats: After params initialized:
I0418 13:14:24.192324 138273875728192 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_24(process=6,(0,6,0,0))
I0418 13:14:24.192357 138273875728192 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_25(process=6,(1,6,0,0))
I0418 13:14:24.192384 138273875728192 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_28(process=6,(0,7,0,0))
I0418 13:14:24.192407 138273875728192 max_utils.py:756] Using (GB) 0.45 / 31.25 (1.440000%) on TPU_29(process=6,(1,7,0,0))
I0418 13:14:24.506314 138273875728192 metric_logger.py:196] completed step: 0, seconds: 15.326, TFLOP/s/device: 0.887, Tokens/s/device: 133.632, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52776.805
I0418 13:14:24.692399 138273875728192 metric_logger.py:196] completed step: 1, seconds: 2.083, TFLOP/s/device: 6.522, Tokens/s/device: 983.017, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52761.840
I0418 13:14:25.115298 138273875728192 metric_logger.py:196] completed step: 2, seconds: 0.028, TFLOP/s/device: 479.907, Tokens/s/device: 72336.818, total_weights: 65536, loss: 10.816, lm_loss: 10.816, perplexity: 49832.398
I0418 13:14:25.273783 138273875728192 metric_logger.py:196] completed step: 3, seconds: 0.423, TFLOP/s/device: 32.098, Tokens/s/device: 4838.188, total_weights: 65536, loss: 10.431, lm_loss: 10.431, perplexity: 33901.820
I0418 13:14:25.591882 138273875728192 metric_logger.py:196] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.698, Tokens/s/device: 12465.155, total_weights: 65536, loss: 9.992, lm_loss: 9.992, perplexity: 21847.639
I0418 13:14:25.598522 138273875728192 metric_logger.py:196] completed step: 5, seconds: 0.158, TFLOP/s/device: 85.774, Tokens/s/device: 12928.803, total_weights: 65536, loss: 9.549, lm_loss: 9.549, perplexity: 14035.006
I0418 13:14:29.481098 2838 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 13:14:32.312694 138144011302656 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/0/items/array_metadatas/process_6
I0418 13:14:46.513342 138144066303744 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 69.424 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 22.75536060333252 s) (per-host)
I0418 13:14:46.513468 138144066303744 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 22.329521s.
I0418 13:14:52.553646 138273875728192 metric_logger.py:196] completed step: 6, seconds: 0.319, TFLOP/s/device: 42.638, Tokens/s/device: 6426.913, total_weights: 65536, loss: 9.111, lm_loss: 9.111, perplexity: 9058.688
I0418 13:14:52.712431 138273875728192 metric_logger.py:196] completed step: 7, seconds: 26.798, TFLOP/s/device: 0.507, Tokens/s/device: 76.424, total_weights: 65536, loss: 8.685, lm_loss: 8.685, perplexity: 5914.720
I0418 13:14:52.870689 138273875728192 metric_logger.py:196] completed step: 8, seconds: 0.163, TFLOP/s/device: 83.163, Tokens/s/device: 12535.271, total_weights: 65536, loss: 8.281, lm_loss: 8.281, perplexity: 3950.010
I0418 13:14:52.875923 138273875728192 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 13:14:53.187147 138273875728192 checkpointing.py:776] Waited 0.3111894130706787 seconds for step 10 to finish before starting checkpointing.
I0418 13:14:53.189806 138273875728192 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:14:55.773468 138144066303744 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 31.589505s.
I0418 13:14:55.773745 138142992561920 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:14:55.773879 138142992561920 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:14:55.773932 138142992561920 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:14:55.775779 138142992561920 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 13:14:55.775898 138273875728192 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.
W0418 13:14:55.775971 138273875728192 checkpoint_manager.py:1452] Waiting for previous save to complete took 2.586178 seconds. If this number is high, consider checkpointing less frequently.
I0418 13:14:55.777802 138273875728192 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 10
I0418 13:14:55.780105 138273875728192 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/10.
I0418 13:14:56.494869 138273875728192 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0418 13:14:56.494969 138273875728192 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
I0418 13:14:56.549887 138273875728192 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.055936s
I0418 13:14:56.550051 138273875728192 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 24.716 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.06241869926452637 s) (per-host)
I0418 13:14:56.550103 138273875728192 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.062479s (batch_requests_ready=0.003371s, total_serialization_initiated=0.059041s, others=0.000068s)
I0418 13:14:56.550211 138273875728192 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.066545s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.066529)
I0418 13:14:56.550964 138273875728192 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/10.
I0418 13:14:56.551245 138142992561920 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:34:56.551221
I0418 13:14:56.553230 138273875728192 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:14:56.553504 138140299814656 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:14:56.553638 138273875728192 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776518093.1897686, 'wait_for_prev_duration_secs': 2.5861780643463135, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776518095.777853, 'checkpointer_blocking_duration_secs': 0.773496150970459, 'get_old_steps_start_time': 1776518096.5513833, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776518093.1879911, 'checkpoint_manager_blocking_duration_secs': 3.3656158447265625}
I0418 13:14:56.553830 138273875728192 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 13:14:56.554397 138273875728192 metric_logger.py:196] completed step: 9, seconds: 0.159, TFLOP/s/device: 85.588, Tokens/s/device: 12900.787, total_weights: 65536, loss: 7.908, lm_loss: 7.908, perplexity: 2720.116
I0418 13:14:56.561575 138273875728192 metric_logger.py:196] completed step: 10, seconds: 0.158, TFLOP/s/device: 85.788, Tokens/s/device: 12930.926, total_weights: 65536, loss: 7.568, lm_loss: 7.568, perplexity: 1936.198
I0418 13:14:56.719796 138273875728192 metric_logger.py:196] completed step: 11, seconds: 3.683, TFLOP/s/device: 3.689, Tokens/s/device: 555.999, total_weights: 65536, loss: 7.288, lm_loss: 7.288, perplexity: 1463.272
I0418 13:14:56.878200 138273875728192 metric_logger.py:196] completed step: 12, seconds: 0.006, TFLOP/s/device: 2183.373, Tokens/s/device: 329101.719, total_weights: 65536, loss: 7.032, lm_loss: 7.032, perplexity: 1132.061
I0418 13:14:57.445882 138273875728192 metric_logger.py:196] completed step: 13, seconds: 0.159, TFLOP/s/device: 85.235, Tokens/s/device: 12847.536, total_weights: 65536, loss: 6.815, lm_loss: 6.815, perplexity: 911.031
I0418 13:14:57.604293 138273875728192 metric_logger.py:196] completed step: 14, seconds: 0.563, TFLOP/s/device: 24.147, Tokens/s/device: 3639.724, total_weights: 65536, loss: 6.635, lm_loss: 6.635, perplexity: 761.080
I0418 13:14:57.762586 138273875728192 metric_logger.py:196] completed step: 15, seconds: 0.163, TFLOP/s/device: 83.235, Tokens/s/device: 12546.098, total_weights: 65536, loss: 6.492, lm_loss: 6.492, perplexity: 659.915
I0418 13:14:57.921101 138273875728192 metric_logger.py:196] completed step: 16, seconds: 0.158, TFLOP/s/device: 85.788, Tokens/s/device: 12930.844, total_weights: 65536, loss: 6.380, lm_loss: 6.380, perplexity: 589.948
I0418 13:14:58.079084 138273875728192 metric_logger.py:196] completed step: 17, seconds: 0.158, TFLOP/s/device: 85.825, Tokens/s/device: 12936.480, total_weights: 65536, loss: 6.293, lm_loss: 6.293, perplexity: 540.941
I0418 13:14:58.237353 138273875728192 metric_logger.py:196] completed step: 18, seconds: 0.159, TFLOP/s/device: 85.703, Tokens/s/device: 12918.039, total_weights: 65536, loss: 6.223, lm_loss: 6.223, perplexity: 504.108
I0418 13:14:58.394950 138273875728192 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 13:14:58.395975 138273875728192 checkpointing.py:776] Waited 0.0010447502136230469 seconds for step 19 to finish before starting checkpointing.
I0418 13:14:58.398098 138273875728192 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:15:05.081932 138140794722048 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/10/items/array_metadatas/process_6
I0418 13:15:41.153529 138142992561920 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 35.368 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.66584610939026 s) (per-host)
I0418 13:15:41.153659 138142992561920 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.602333s.
I0418 13:15:50.735993 138142992561920 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 54.184650s.
I0418 13:15:50.736290 138140299814656 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:15:50.736411 138140299814656 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:15:50.736460 138140299814656 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:15:50.737948 138140299814656 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 13:15:50.738121 138273875728192 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.
W0418 13:15:50.738252 138273875728192 checkpoint_manager.py:1452] Waiting for previous save to complete took 52.340158 seconds. If this number is high, consider checkpointing less frequently.
I0418 13:15:50.739865 138273875728192 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 19
I0418 13:15:50.742234 138273875728192 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/19.
I0418 13:15:51.502743 138273875728192 jax_array_handlers.py:360] Scheduling D2H of 81 prioritized jax.Array.
I0418 13:15:51.502859 138273875728192 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
I0418 13:15:51.558573 138273875728192 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.056768s
I0418 13:15:51.558743 138273875728192 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 24.439 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0631253719329834 s) (per-host)
I0418 13:15:51.558793 138273875728192 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.063184s (batch_requests_ready=0.003299s, total_serialization_initiated=0.059821s, others=0.000064s)
I0418 13:15:51.558909 138273875728192 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.067197s (all_items=0.000011s, per_item={'items': '0.00001144'}, temp_paths=0.067185)
I0418 13:15:51.559574 138273875728192 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.82 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/19.
I0418 13:15:51.559910 138140299814656 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:35:51.559873
I0418 13:15:51.561916 138273875728192 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:15:51.562207 138140794722048 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:15:51.562366 138273875728192 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776518098.3980682, 'wait_for_prev_duration_secs': 52.340158224105835, 'time_between_consecutive_saves_sec': 2.622237205505371, 'checkpointer_blocking_start_time': 1776518150.7399032, 'checkpointer_blocking_duration_secs': 0.8201515674591064, 'get_old_steps_start_time': 1776518151.5600727, 'get_old_steps_duration_secs': 2.3603439331054688e-05, 'checkpoint_manager_blocking_start_time': 1776518098.396248, 'checkpoint_manager_blocking_duration_secs': 53.16608691215515}
I0418 13:15:51.562471 138273875728192 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 13:15:51.562514 138273875728192 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:15:58.644558 138141373556480 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_05_fp8/checkpoints/19/items/array_metadatas/process_6
I0418 13:16:35.229857 138140299814656 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 36.122 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.73419809341431 s) (per-host)
I0418 13:16:35.229969 138140299814656 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.669944s.
I0418 13:16:43.820790 138140299814656 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.260747s.
I0418 13:16:43.821096 138140794722048 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:16:43.821215 138140794722048 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:16:43.821256 138140794722048 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:16:43.823195 138140794722048 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 13:16:43.823361 138273875728192 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.
I0418 13:16:43.823516 138273875728192 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:16:43.824536 138273875728192 metric_logger.py:196] completed step: 19, seconds: 0.158, TFLOP/s/device: 86.021, Tokens/s/device: 12966.046, 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: Sat Apr 18 13:16:55 UTC 2026
EXIT_CODE=0