MaxView

XPK Start: Thu Apr 23 14:57:12 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 14:57:37.570365: 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 14:57:37.779944 136267875350336 max_utils.py:273] Attempting to initialize the jax distributed system...
I0423 14:57:46.822247 136267875350336 distributed.py:149] Starting JAX distributed service on [::]:8482
I0423 14:57:46.824565 136267875350336 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-gknnq-slice-job-0-0.mt-11-optimizer-offload-gknnq:8482
I0423 14:57:47.885259 136267875350336 max_utils.py:284] Jax distributed system initialized!
I0423 14:57:54.094861 136267875350336 max_utils.py:800] System Information: Jax Version: 0.9.2
I0423 14:57:54.094982 136267875350336 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0423 14:57:54.095026 136267875350336 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 14:57:54.095062 136267875350336 train_utils.py:391] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 14:57:54.798308 136267875350336 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 14:57:54.916632 136267875350336 checkpointing.py:688] Setting up checkpoint logger...
I0423 14:57:54.916753 136267875350336 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 14:57:54.916797 136267875350336 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0423 14:57:54.917015 136267875350336 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 0x7bee9e950170>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 14:57:58.196192 136267875350336 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0423 14:57:58.196434 136267875350336 checkpoint_manager.py:708] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bda901f1070>}, handler_registry=None
I0423 14:57:58.196704 136267875350336 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bda901f1070>` 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 14:57:58.196760 136267875350336 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bed92691400>` 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 14:57:58.196803 136267875350336 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 0x7bda901f1070>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bda901f1070>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bed92691400>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bed92691400>}).
I0423 14:57:58.197176 136267875350336 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0423 14:57:58.197263 136267875350336 async_checkpointer.py:192] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bda584276a0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0423 14:57:58.973272 136267875350336 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints
I0423 14:57:58.980893 136267875350336 checkpoint_manager.py:929] [process=3][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/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bda781f3320>
I0423 14:57:58.981031 136267875350336 checkpointing.py:302] Checkpoint manager created!
I0423 14:57:59.957755 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to host
I0423 14:57:59.957872 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.957915 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.957978 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958023 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958071 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.958119 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.958152 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958204 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958244 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958284 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958327 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958368 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0423 14:57:59.958405 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.958436 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958462 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958487 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958511 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.958537 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 14:57:59.958563 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958587 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958611 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958635 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958659 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 14:57:59.958681 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0423 14:57:59.958704 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to host
I0423 14:57:59.958726 136267875350336 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to host
I0423 14:57:59.959839 136267875350336 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 14:57:59.959897 136267875350336 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 14:58:01.428051 136267875350336 maxtext_utils.py:1889]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0423 14:58:01.428171 136267875350336 maxtext_utils.py:1889]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 14:58:01.428223 136267875350336 maxtext_utils.py:1889]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 14:58:01.428265 136267875350336 maxtext_utils.py:1889]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0423 14:58:01.428301 136267875350336 maxtext_utils.py:1889]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 14:58:01.428336 136267875350336 maxtext_utils.py:1889]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 14:58:01.428374 136267875350336 maxtext_utils.py:1889]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 14:58:01.428409 136267875350336 maxtext_utils.py:1889]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0423 14:58:01.428443 136267875350336 maxtext_utils.py:1889]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 14:58:01.428480 136267875350336 maxtext_utils.py:1889]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 14:58:01.428517 136267875350336 maxtext_utils.py:1889]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)

I0423 14:58:01.428550 136267875350336 maxtext_utils.py:1889]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0423 14:58:01.831326 136267875350336 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 14:58:01.831419 136267875350336 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 14:58:01.837113 136267875350336 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 14:58:01.837169 136267875350336 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 14:58:01.853814 136267875350336 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 14:58:01.853873 136267875350336 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 14:58:01.869898 136267875350336 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 14:58:01.869966 136267875350336 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 14:58:01.901254 136267875350336 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 14:58:01.901359 136267875350336 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 14:58:01.917574 136267875350336 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 14:58:01.917634 136267875350336 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 14:58:01.933602 136267875350336 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 14:58:01.933661 136267875350336 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 14:58:01.966135 136267875350336 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 14:58:01.966210 136267875350336 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 14:58:01.986373 136267875350336 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 14:58:01.986438 136267875350336 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 14:58:02.156985 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 14:58:02.157109 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157151 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157189 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157219 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157248 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157276 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157301 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157328 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157354 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157379 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157403 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157428 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 14:58:02.157454 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157480 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157508 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157532 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157556 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157580 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:02.157603 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157627 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157651 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157679 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157709 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:02.157734 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 14:58:02.157757 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 14:58:02.157779 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0423 14:58:13.213230 136267875350336 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0423 14:58:13.215465 136267875350336 metric_logger.py:301] number parameters: 1.104 billion
I0423 14:58:13.770720 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 14:58:13.770844 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.770886 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.770921 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.770968 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771000 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.771028 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.771054 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771079 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771105 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771130 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771155 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771179 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 14:58:13.771204 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.771233 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771258 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771280 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771304 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.771329 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 14:58:13.771353 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771378 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771403 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771430 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771454 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 14:58:13.771481 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 14:58:13.771504 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 14:58:13.771526 136267875350336 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0423 14:58:24.450767 136267875350336 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0423 14:58:24.619252 136267875350336 checkpointing.py:798] Waited 0.1684706211090088 seconds for step 0 to finish before starting checkpointing.
I0423 14:58:24.621703 136267875350336 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 14:58:24.623541 136267875350336 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 0
I0423 14:58:24.624812 136267875350336 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/0.
I0423 14:58:24.932208 136267875350336 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 14:58:24.933200 136267875350336 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0423 14:58:24.933258 136267875350336 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 14:58:25.205241 136267875350336 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.273813s
I0423 14:58:25.205411 136267875350336 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 5.436 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2837789058685303 s) (per-host)
I0423 14:58:25.205468 136267875350336 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283841s (batch_requests_ready=0.003174s, total_serialization_initiated=0.280597s, others=0.000070s)
I0423 14:58:25.205570 136267875350336 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.288068s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.288050)
I0423 14:58:25.206346 136267875350336 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.58 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/0.
I0423 14:58:25.206693 136132511643392 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 15:18:25.206654
I0423 14:58:25.215011 136267875350336 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 14:58:25.215325 136132238042880 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0423 14:58:25.215482 136267875350336 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776956304.6216853, 'wait_for_prev_duration_secs': 6.103515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776956304.6235805, 'checkpointer_blocking_duration_secs': 0.5832705497741699, 'get_old_steps_start_time': 1776956305.2068748, 'get_old_steps_duration_secs': 3.0279159545898438e-05, 'checkpoint_manager_blocking_start_time': 1776956304.6196525, 'checkpoint_manager_blocking_duration_secs': 0.5957918167114258}
I0423 14:58:25.215589 136267875350336 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0423 14:58:25.215656 136267875350336 max_utils.py:750] 
Memstats: After params initialized:
I0423 14:58:25.215704 136267875350336 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_10(process=3,(2,2,0,0))
I0423 14:58:25.215735 136267875350336 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_11(process=3,(3,2,0,0))
I0423 14:58:25.215760 136267875350336 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_14(process=3,(2,3,0,0))
I0423 14:58:25.215785 136267875350336 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_15(process=3,(3,3,0,0))
I0423 14:58:25.536334 136267875350336 metric_logger.py:196] completed step: 0, seconds: 11.234, TFLOP/s/device: 1.209, Tokens/s/device: 182.299, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0423 14:58:25.729810 136267875350336 metric_logger.py:196] completed step: 1, seconds: 1.084, TFLOP/s/device: 12.540, Tokens/s/device: 1890.155, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0423 14:58:26.170029 136267875350336 metric_logger.py:196] completed step: 2, seconds: 0.033, TFLOP/s/device: 415.267, Tokens/s/device: 62593.600, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28632.133
I0423 14:58:26.332976 136267875350336 metric_logger.py:196] completed step: 3, seconds: 0.429, TFLOP/s/device: 31.669, Tokens/s/device: 4773.526, total_weights: 65536, loss: 9.732, lm_loss: 9.732, perplexity: 16852.779
I0423 14:58:26.659265 136267875350336 metric_logger.py:196] completed step: 4, seconds: 0.180, TFLOP/s/device: 75.422, Tokens/s/device: 11368.430, total_weights: 65536, loss: 9.274, lm_loss: 9.274, perplexity: 10659.797
I0423 14:58:26.666722 136267875350336 metric_logger.py:196] completed step: 5, seconds: 0.162, TFLOP/s/device: 83.690, Tokens/s/device: 12614.721, total_weights: 65536, loss: 8.888, lm_loss: 8.888, perplexity: 7244.862
I0423 14:58:28.277046    2786 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 14:58:30.302602 136132494857984 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_3
I0423 14:58:48.969581 136267875350336 metric_logger.py:196] completed step: 6, seconds: 0.327, TFLOP/s/device: 41.551, Tokens/s/device: 6262.978, total_weights: 65536, loss: 8.589, lm_loss: 8.589, perplexity: 5371.227
I0423 14:58:49.132519 136267875350336 metric_logger.py:196] completed step: 7, seconds: 22.141, TFLOP/s/device: 0.614, Tokens/s/device: 92.497, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4360.796
I0423 14:58:49.295363 136267875350336 metric_logger.py:196] completed step: 8, seconds: 0.169, TFLOP/s/device: 80.529, Tokens/s/device: 12138.166, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.198
I0423 14:58:49.459069 136267875350336 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0423 14:58:49.459743 136267875350336 checkpointing.py:798] Waited 0.0006854534149169922 seconds for step 9 to finish before starting checkpointing.
I0423 14:58:49.462097 136267875350336 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 14:59:02.891248 136132511643392 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 41.600 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.96958565711975 s) (per-host)
I0423 14:59:02.891354 136132511643392 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.684547s.
I0423 14:59:10.192741 136132511643392 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 44.985918s.
I0423 14:59:10.193126 136132238042880 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 14:59:10.193249 136132238042880 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0423 14:59:10.193312 136132238042880 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 14:59:10.195031 136132238042880 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 14:59:10.195214 136267875350336 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0423 14:59:10.195370 136267875350336 checkpoint_manager.py:1452] Waiting for previous save to complete took 20.733265 seconds. If this number is high, consider checkpointing less frequently.
I0423 14:59:10.197302 136267875350336 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 9
I0423 14:59:10.199450 136267875350336 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/9.
I0423 14:59:10.492227 136267875350336 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0423 14:59:10.492318 136267875350336 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 14:59:10.527983 136267875350336 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037294s
I0423 14:59:10.528153 136267875350336 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 34.689 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04446744918823242 s) (per-host)
I0423 14:59:10.528205 136267875350336 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044529s (batch_requests_ready=0.002870s, total_serialization_initiated=0.041590s, others=0.000069s)
I0423 14:59:10.528301 136267875350336 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048681s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.048665)
I0423 14:59:10.529031 136267875350336 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/9.
I0423 14:59:10.529390 136132238042880 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 15:19:10.529352
I0423 14:59:10.531387 136267875350336 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 14:59:10.531640 136113465313024 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0423 14:59:10.531803 136267875350336 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776956329.4620671, 'wait_for_prev_duration_secs': 20.733264923095703, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776956350.197342, 'checkpointer_blocking_duration_secs': 0.33219361305236816, 'get_old_steps_start_time': 1776956350.529559, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776956329.4599898, 'checkpoint_manager_blocking_duration_secs': 21.071778535842896}
I0423 14:59:10.531912 136267875350336 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0423 14:59:10.531971 136267875350336 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 14:59:15.610722 136132494857984 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260423_124541/nnx_xpk_feat_nnx_post_train_fixes_20260423_124541_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_3
I0423 14:59:52.514365 136132238042880 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 37.580 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.03064179420471 s) (per-host)
I0423 14:59:52.514489 136132238042880 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.984989s.
I0423 15:00:01.116494 136132238042880 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 50.586977s.
I0423 15:00:01.116792 136113465313024 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 15:00:01.116916 136113465313024 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0423 15:00:01.116985 136113465313024 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0423 15:00:01.119183 136113465313024 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0423 15:00:01.119350 136267875350336 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0423 15:00:01.119518 136267875350336 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 15:00:01.120518 136267875350336 metric_logger.py:196] completed step: 9, seconds: 0.164, TFLOP/s/device: 83.020, Tokens/s/device: 12513.672, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3550.804
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Thu Apr 23 15:00:13 UTC 2026
EXIT_CODE=0