Add files using upload-large-folder tool

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+.metadata filter=lfs diff=lfs merge=lfs -text
+tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+__0_0.distcp filter=lfs diff=lfs merge=lfs -text

.metadata

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d92db4d020440aa043f898c4acc40c46e34d7565015f4103c531595faa7b930
+size 257851

__0_0.distcp

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5116d9a4aec7407bce8f102ddd0aa152c045179e498f3e20f4bbad53f2b90d41
+size 9321437622

common.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3203c48c508d56071331991ba182a42afd9d7c88152a2c2b4bdc62213519008f
+size 1383

metadata.json

@@ -0,0 +1 @@
+{"sharded_backend": "torch_dist", "sharded_backend_version": 1, "common_backend": "torch", "common_backend_version": 1}

run_config.yaml

@@ -0,0 +1,666 @@
+_target_: megatron.bridge.training.config.ConfigContainer
+checkpoint:
+  _target_: megatron.bridge.training.config.CheckpointConfig
+  async_ckpt_cpu_priority: 10
+  async_ckpt_io_priority: 3
+  async_ckpt_use_cpu_shm: false
+  async_save: false
+  async_strategy: mcore
+  async_write_results_mp_mode: fork
+  auto_detect_ckpt_format: false
+  ckpt_assume_constant_structure: false
+  ckpt_convert_format: null
+  ckpt_convert_save: null
+  ckpt_convert_update_legacy_dist_opt_format: false
+  ckpt_format: torch_dist
+  ckpt_fully_parallel_load_exchange_algo: broadcast
+  ckpt_fully_parallel_load_process_group: dp
+  ckpt_fully_parallel_save_process_group: dp
+  ckpt_load_validate_sharding_integrity: true
+  ckpt_step: null
+  custom_manager_class: null
+  dist_ckpt_optim_fully_reshardable: true
+  dist_ckpt_save_pre_mcore_014: false
+  dist_ckpt_strictness: assume_ok_unexpected
+  dist_ckpt_workers: 1
+  distrib_optim_fully_reshardable_mem_efficient: false
+  exit_on_missing_checkpoint: false
+  finetune: false
+  fully_parallel_load: false
+  fully_parallel_save: true
+  load: null
+  load_main_params_from_ckpt: false
+  load_optim: true
+  load_rng: true
+  most_recent_k: -1
+  non_persistent_ckpt_type: null
+  non_persistent_global_ckpt_dir: null
+  non_persistent_local_ckpt_algo: fully_parallel
+  non_persistent_local_ckpt_dir: null
+  non_persistent_save_interval: null
+  pretrained_checkpoint: null
+  replication: false
+  replication_factor: 2
+  replication_jump: null
+  save: /home/junjie.ye/workspace/humanoid_research/cua/models/checkpoints/slides01_aw_init_from_3x422_func_v1
+  save_activations_interval: null
+  save_dgrads_interval: null
+  save_interval: null
+  save_optim: false
+  save_params_interval: null
+  save_retain_interval: null
+  save_rng: false
+  save_tokenizer_assets: true
+  save_tokens_per_expert_interval: null
+  save_wgrads_interval: null
+  storage_writers_per_rank: 1
+  strict_fsdp_dtensor_load: false
+  use_checkpoint_args: false
+  use_mp_args_from_checkpoint_args: false
+  use_persistent_ckpt_worker: true
+  use_tokenizer_model_from_checkpoint_args: true
+  verify_integrity: false
+comm_overlap: null
+dataset: null
+ddp: null
+dist: null
+ft: null
+inprocess_restart: null
+logger:
+  _target_: megatron.bridge.training.config.LoggerConfig
+  barrier_with_L1_time: true
+  comet_api_key: null
+  comet_experiment_name: null
+  comet_project: null
+  comet_tags: null
+  comet_workspace: null
+  filter_warnings: true
+  log_device_memory_used: false
+  log_energy: false
+  log_interval: 100
+  log_l2_norm_grad_to_tensorboard: false
+  log_loss_scale_to_tensorboard: true
+  log_max_attention_logit: false
+  log_memory_interval: null
+  log_memory_to_tensorboard: false
+  log_num_zeros_in_grad: false
+  log_params_norm: false
+  log_progress: false
+  log_runtime_to_tensorboard: false
+  log_throughput: false
+  log_throughput_to_tensorboard: false
+  log_timers_to_tensorboard: false
+  log_validation_ppl_to_tensorboard: false
+  log_world_size_to_tensorboard: false
+  logging_level: 20
+  memory_keys: null
+  mlflow_description: null
+  mlflow_experiment: null
+  mlflow_log_artifacts: true
+  mlflow_run_name: null
+  mlflow_tags: null
+  mlflow_tracking_uri: null
+  modules_to_filter: null
+  runtime_time_unit: hours
+  save_config_filepath: null
+  set_level_for_all_loggers: false
+  skip_train_metrics_log: false
+  tensorboard_dir: null
+  tensorboard_log_interval: 1
+  tensorboard_queue_size: 1000
+  throughput_window_size: 100
+  timing_log_level: 0
+  timing_log_option: minmax
+  wandb_entity: null
+  wandb_exp_name: null
+  wandb_project: null
+  wandb_save_dir: null
+mixed_precision: null
+model:
+  _target_: megatron.bridge.models.qwen_vl.qwen35_vl_provider.Qwen35VLModelProvider
+  account_for_embedding_in_pipeline_split: false
+  account_for_loss_in_pipeline_split: false
+  activation_func:
+    _call_: false
+    _target_: torch.nn.functional.silu
+  activation_func_clamp_value: null
+  activation_func_fp8_input_store: false
+  add_bias_linear: false
+  add_qkv_bias: false
+  apply_query_key_layer_scaling: false
+  apply_residual_connection_post_layernorm: false
+  apply_rope_fusion: false
+  apply_rotary_pos_emb_in_fp32: false
+  attention_backend:
+    _args_:
+    - 5
+    _call_: true
+    _name_: auto
+    _target_: megatron.core.transformer.enums.AttnBackend
+  attention_dropout: 0.0
+  attention_output_gate: true
+  attention_softmax_in_fp32: true
+  autocast_dtype:
+    _call_: false
+    _target_: torch.bfloat16
+  barrier_with_L1_time: true
+  batch_invariant_mode: false
+  batch_p2p_comm: true
+  batch_p2p_sync: true
+  bf16: true
+  bias_activation_fusion: true
+  bias_dropout_fusion: false
+  bos_token_id: null
+  calculate_per_token_loss: false
+  clone_scatter_output_in_embedding: true
+  config_logger_dir: ''
+  context_parallel_size: 1
+  cp_comm_type: null
+  cpu_offloading: false
+  cpu_offloading_activations: true
+  cpu_offloading_double_buffering: false
+  cpu_offloading_num_layers: 0
+  cpu_offloading_retain_pinned_cpu_buffers: false
+  cpu_offloading_weights: false
+  cross_entropy_fusion_impl: native
+  cross_entropy_loss_fusion: true
+  cuda_graph_impl: none
+  cuda_graph_modules: []
+  cuda_graph_retain_backward_graph: false
+  cuda_graph_scope: null
+  cuda_graph_use_single_mempool: true
+  cuda_graph_warmup_steps: 3
+  deallocate_pipeline_outputs: true
+  defer_embedding_wgrad_compute: false
+  delay_wgrad_compute: false
+  dense_grouped_gemm: false
+  deterministic_mode: false
+  disable_bf16_reduced_precision_matmul: false
+  disable_parameter_transpose_cache: false
+  distribute_saved_activations: false
+  dsa_indexer_head_dim: null
+  dsa_indexer_loss_coeff: null
+  dsa_indexer_n_heads: null
+  dsa_indexer_topk: null
+  dsa_indexer_use_sparse_loss: false
+  embedding_init_method:
+    _args_: []
+    _partial_: true
+    _target_: torch.nn.init.normal_
+    mean: 0.0
+    std: 0.02
+  embedding_init_method_std: 0.02
+  enable_autocast: false
+  enable_cuda_graph: false
+  eos_token_id: 248044
+  ep_overlap_early_attn_memory_release: false
+  experimental_attention_variant: gated_delta_net
+  expert_model_parallel_size: 1
+  expert_tensor_parallel_size: 1
+  external_cuda_graph: false
+  ffn_hidden_size: 9216
+  finalize_model_grads_func: null
+  fine_grained_activation_offloading: false
+  fine_grained_offloading_max_inflight_offloads: null
+  first_last_layers_bf16: false
+  flash_decode: false
+  fp16: false
+  fp16_lm_cross_entropy: false
+  fp32_residual_connection: false
+  fp4: null
+  fp4_param: false
+  fp4_quantizer_factory: null
+  fp4_recipe: nvfp4
+  fp8: null
+  fp8_amax_compute_algo: most_recent
+  fp8_amax_history_len: 1
+  fp8_dot_product_attention: false
+  fp8_interval: 1
+  fp8_margin: 0
+  fp8_multi_head_attention: false
+  fp8_output_proj: false
+  fp8_param: false
+  fp8_quantizer_factory: null
+  fp8_recipe: delayed
+  fp8_wgrad: true
+  freeze_language_model: false
+  freeze_vision_model: false
+  freeze_vision_projection: false
+  fused_residual_rmsnorm: false
+  fused_single_qkv_rope: false
+  gated_linear_unit: true
+  glu_linear_offset: 0.0
+  grad_scale_func: null
+  grad_sync_func: null
+  gradient_accumulation_fusion: false
+  hetereogenous_dist_checkpoint: true
+  heterogeneous_block_specs: false
+  hf_model_id: /home/junjie.ye/workspace/humanoid_research/cua/models/hf_exports/qwen35-vl-4b-cua-3x-full-func-nothink-1ep-20260907/iter_0000422
+  hidden_dropout: 0.0
+  hidden_size: 2560
+  hierarchical_context_parallel_sizes: null
+  high_priority_a2a_comm_stream: false
+  hybrid_context_parallel: false
+  image_token_id: 248056
+  inference_cuda_graph_scope:
+    _args_:
+    - 1
+    _call_: true
+    _name_: none
+    _target_: megatron.core.transformer.enums.InferenceCudaGraphScope
+  inference_disable_triton_nvls_kernels: false
+  inference_fuse_tp_communication: false
+  inference_grouped_gemm_backend: vllm
+  inference_moe_disable_fused_quant_kernels: false
+  inference_moe_token_dispatcher_type: nvls
+  inference_rng_tracker: false
+  inference_sampling_seed: 42
+  init_method:
+    _args_: []
+    _partial_: true
+    _target_: torch.nn.init.normal_
+    mean: 0.0
+    std: 0.02
+  init_method_std: 0.02
+  init_model_with_meta_device: false
+  is_hybrid_model: false
+  kitchen_attention_backend: sdpa
+  kv_channels: 256
+  language_max_sequence_length: 2048
+  layernorm_epsilon: 1.0e-06
+  layernorm_zero_centered_gamma: true
+  linear_attention_freq: 4
+  linear_conv_kernel_dim: 4
+  linear_key_head_dim: 128
+  linear_num_key_heads: 16
+  linear_num_value_heads: 32
+  linear_value_head_dim: 128
+  log_max_attention_logit: false
+  make_vocab_size_divisible_by: 128
+  mamba_head_dim: 64
+  mamba_num_groups: 8
+  mamba_num_heads: null
+  mamba_state_dim: 128
+  masked_softmax_fusion: true
+  max_seqlen_per_dp_cp_rank: null
+  memory_efficient_layer_norm: false
+  microbatch_group_size_per_vp_stage: 1
+  min_offloaded_tensor_size: 1048576
+  mlp_chunks_for_prefill: 1
+  mlp_chunks_for_training: 1
+  moe_apply_probs_on_input: false
+  moe_aux_loss_coeff: 0.0
+  moe_deepep_num_sms: 20
+  moe_enable_deepep: false
+  moe_enable_routing_replay: false
+  moe_expert_capacity_factor: null
+  moe_expert_rank_capacity_factor: null
+  moe_ffn_hidden_size: null
+  moe_flex_dispatcher_backend: deepep
+  moe_grouped_gemm: false
+  moe_hybridep_num_blocks_permute: null
+  moe_hybridep_num_blocks_unpermute: null
+  moe_hybridep_num_sms: null
+  moe_hybridep_num_sms_preprocessing: 108
+  moe_input_jitter_eps: null
+  moe_latent_size: null
+  moe_layer_freq: 1
+  moe_layer_recompute: false
+  moe_mlp_glu_interleave_size: null
+  moe_pad_expert_input_to_capacity: false
+  moe_pad_experts_for_cuda_graph_inference: false
+  moe_paged_stash: false
+  moe_paged_stash_buffer_size_factor_cpu: 0.0
+  moe_paged_stash_buffer_size_factor_cuda: 1.1
+  moe_paged_stash_page_size: 64
+  moe_per_layer_logging: false
+  moe_permute_fusion: false
+  moe_permute_fusion_into_hybridep: false
+  moe_router_bias_update_rate: 0.001
+  moe_router_dtype: null
+  moe_router_enable_expert_bias: false
+  moe_router_force_biased: null
+  moe_router_force_load_balancing: false
+  moe_router_fusion: false
+  moe_router_group_topk: null
+  moe_router_load_balancing_type: aux_loss
+  moe_router_num_groups: null
+  moe_router_padding_for_fp8: false
+  moe_router_padding_for_quantization: false
+  moe_router_pre_softmax: false
+  moe_router_score_function: softmax
+  moe_router_topk: 2
+  moe_router_topk_limited_devices: null
+  moe_router_topk_scaling_factor: null
+  moe_shared_expert_gate: false
+  moe_shared_expert_intermediate_size: null
+  moe_shared_expert_overlap: false
+  moe_single_grouped_bias: false
+  moe_single_grouped_weight: false
+  moe_token_dispatcher_type: allgather
+  moe_token_drop_policy: probs
+  moe_token_dropping: false
+  moe_z_loss_coeff: null
+  mrope_section:
+  - 11
+  - 11
+  - 10
+  mtp_enabled: false
+  mtp_hybrid_override_pattern: null
+  mtp_loss_scaling_factor: 0.1
+  mtp_num_layers: 1
+  mtp_standalone: false
+  mtp_use_repeated_layer: false
+  multi_latent_attention: false
+  mup_attn_scale_power: 1.0
+  mup_base_head_dim: null
+  mup_base_hidden_size: null
+  mup_embedding_mult: 1.0
+  mup_output_mult: 1.0
+  mup_width_mult: 1.0
+  nccl_all_reduce_for_prefill: false
+  no_rope_freq: null
+  no_sync_func: null
+  normalization: RMSNorm
+  num_attention_heads: 16
+  num_layers: 32
+  num_layers_at_end_in_bf16: 1
+  num_layers_at_start_in_bf16: 1
+  num_layers_in_first_pipeline_stage: null
+  num_layers_in_last_pipeline_stage: null
+  num_microbatches_with_partial_activation_checkpoints: null
+  num_moe_experts: null
+  num_query_groups: 4
+  offload_modules: []
+  output_layer_init_method:
+    _args_: []
+    _partial_: true
+    _target_: torch.nn.init.normal_
+    mean: 0.0
+    std: 0.0025
+  overlap_dispatch_backward_with_experts_wgrad: false
+  overlap_moe_expert_parallel_comm: false
+  overlap_p2p_comm: false
+  overlap_p2p_comm_warmup_flush: false
+  parallel_output: true
+  param_sync_func: null
+  params_dtype:
+    _call_: false
+    _target_: torch.bfloat16
+  patch_size: 16
+  perform_initialization: false
+  persist_layer_norm: false
+  pipeline_dtype:
+    _call_: false
+    _target_: torch.bfloat16
+  pipeline_model_parallel_comm_backend: null
+  pipeline_model_parallel_layout: null
+  pipeline_model_parallel_size: 1
+  position_embedding_type: mrope
+  qk_clip: false
+  qk_clip_alpha: 0.5
+  qk_clip_threshold: 100
+  qk_l2_norm: false
+  qk_layernorm: true
+  quant_recipe: null
+  recompute_granularity: null
+  recompute_method: null
+  recompute_modules:
+  - core_attn
+  recompute_num_layers: null
+  restore_modelopt_state: false
+  rope_scaling: false
+  rope_scaling_factor: 1.0
+  rotary_base: 10000000
+  rotary_interleaved: false
+  rotary_percent: 0.25
+  rotary_scaling_factor: null
+  scatter_embedding_sequence_parallel: false
+  seq_len_interpolation_factor: null
+  seq_length: 262144
+  sequence_parallel: false
+  share_embeddings_and_output_weights: true
+  should_pad_vocab: false
+  softmax_scale: null
+  softmax_type: vanilla
+  spatial_merge_size: 2
+  symmetric_ar_type: null
+  temporal_patch_size: 2
+  tensor_model_parallel_size: 1
+  test_mode: false
+  timers: null
+  tp_comm_atomic_ag: false
+  tp_comm_atomic_rs: false
+  tp_comm_bootstrap_backend: nccl
+  tp_comm_bulk_dgrad: true
+  tp_comm_bulk_wgrad: true
+  tp_comm_overlap: false
+  tp_comm_overlap_ag: true
+  tp_comm_overlap_cfg: null
+  tp_comm_overlap_disable_fc1: false
+  tp_comm_overlap_disable_qkv: false
+  tp_comm_overlap_rs: true
+  tp_comm_overlap_rs_dgrad: false
+  tp_comm_split_ag: true
+  tp_comm_split_rs: true
+  tp_only_amax_red: false
+  transformer_impl: transformer_engine
+  transformer_layer_spec:
+    _call_: false
+    _target_: megatron.core.models.gpt.experimental_attention_variant_module_specs.get_transformer_block_with_experimental_attention_variant_spec
+  use_arbitrary_attention_mask: null
+  use_cpu_initialization: false
+  use_fused_weighted_squared_relu: false
+  use_grouped_gemm_for_dense_mlp: false
+  use_hf_vision_model: false
+  use_inference_optimized_layers: false
+  use_kitchen: false
+  use_kitchen_attention: false
+  use_mamba_mem_eff_path: true
+  use_mup: false
+  use_ring_exchange_p2p: false
+  use_te_activation_func: false
+  use_te_rng_tracker: false
+  use_transformer_engine_full_layer_spec: false
+  use_transformer_engine_op_fuser: false
+  variable_seq_lengths: false
+  video_token_id: 248057
+  virtual_pipeline_model_parallel_size: null
+  vision_config:
+    _target_: transformers.models.qwen3_5.configuration_qwen3_5.Qwen3_5VisionConfig
+    architectures: null
+    chunk_size_feed_forward: 0
+    depth: 24
+    dtype:
+      _call_: false
+      _target_: torch.bfloat16
+    hidden_act: gelu_pytorch_tanh
+    hidden_size: 1024
+    id2label:
+      0: LABEL_0
+      1: LABEL_1
+    in_channels: 3
+    initializer_range: 0.02
+    intermediate_size: 4096
+    is_encoder_decoder: false
+    label2id:
+      LABEL_0: 0
+      LABEL_1: 1
+    num_heads: 16
+    num_position_embeddings: 2304
+    out_hidden_size: 2560
+    output_hidden_states: false
+    patch_size: 16
+    problem_type: null
+    return_dict: true
+    spatial_merge_size: 2
+    temporal_patch_size: 2
+    transformers_version: null
+  vision_dp_when_cp: false
+  vision_end_token_id: 248054
+  vision_start_token_id: 248053
+  vocab_size: 248320
+  wgrad_deferral_limit: 0
+  window_attn_skip_freq: null
+  window_size: null
+  yarn_beta_fast: null
+  yarn_beta_slow: null
+  yarn_correction_range_round_to_int: null
+  yarn_mscale: null
+  yarn_mscale_all_dim: null
+  yarn_original_max_position_embeddings: null
+  yarn_rotary_scaling_factor: null
+nvrx_straggler: null
+optimizer:
+  _target_: megatron.core.optimizer.optimizer_config.OptimizerConfig
+  adam_beta1: 0.9
+  adam_beta2: 0.999
+  adam_eps: 1.0e-08
+  adaptive_muon_beta2: 0.95
+  adaptive_muon_eps: 1.0e-08
+  adaptive_muon_moment2_method: adamuon
+  apply_wd_to_qk_layernorm: false
+  barrier_with_L1_time: false
+  bf16: false
+  clip_grad: 1.0
+  config_logger_dir: ''
+  decoupled_lr: null
+  decoupled_min_lr: null
+  decoupled_weight_decay: true
+  exp_avg_dtype:
+    _call_: false
+    _target_: torch.float32
+  exp_avg_sq_dtype:
+    _call_: false
+    _target_: torch.float32
+  fp16: false
+  fp8_recipe: null
+  grad_norm_skip_threshold: .inf
+  hysteresis: 2
+  initial_loss_scale: 4294967296
+  lion_beta1: 0.95
+  lion_beta2: 0.98
+  log_num_zeros_in_grad: false
+  loss_scale: null
+  loss_scale_window: 1000
+  lr: null
+  main_grads_dtype:
+    _call_: false
+    _target_: torch.float32
+  main_params_dtype:
+    _call_: false
+    _target_: torch.float32
+  min_loss_scale: 1.0
+  min_lr: null
+  muon_coefficient_type: quintic
+  muon_extra_scale_factor: 1.0
+  muon_fp32_matmul_prec: medium
+  muon_momentum: 0.95
+  muon_nesterov: false
+  muon_num_ns_steps: 5
+  muon_scalar_optimizer: adam
+  muon_scale_mode: spectral
+  muon_split_qkv: true
+  muon_tp_mode: blockwise
+  optimizer: adam
+  optimizer_cpu_offload: false
+  optimizer_cuda_graph: false
+  optimizer_offload_fraction: 0.0
+  overlap_cpu_optimizer_d2h_h2d: false
+  overlap_param_gather: false
+  overlap_param_gather_with_optimizer_step: false
+  params_dtype:
+    _call_: false
+    _target_: torch.float32
+  pin_cpu_grads: true
+  pin_cpu_params: true
+  reuse_grad_buf_for_mxfp8_param_ag: false
+  sgd_momentum: 0.9
+  soap_precondition_frequency: 1
+  soap_shampoo_beta: 0.95
+  soap_use_kl_shampoo: true
+  store_param_remainders: true
+  timers: null
+  use_distributed_optimizer: false
+  use_layer_wise_distributed_optimizer: false
+  use_precision_aware_optimizer: false
+  use_torch_optimizer_for_cpu_offload: false
+  weight_decay: 0.01
+optimizer_config_override_provider:
+  _target_: megatron.bridge.training.config.OptimizerConfigOverrideProvider
+peft: null
+profiling:
+  _target_: megatron.bridge.training.config.ProfilingConfig
+  memory_snapshot_path: /nemo_run/snapshot.pickle
+  nvtx_ranges: false
+  profile_ranks:
+  - 0
+  profile_step_end: 12
+  profile_step_start: 10
+  pytorch_profiler_collect_callstack: false
+  pytorch_profiler_collect_chakra: false
+  pytorch_profiler_collect_shapes: false
+  record_memory_history: false
+  record_shapes: false
+  use_nsys_profiler: false
+  use_pytorch_profiler: false
+rerun_state_machine:
+  _target_: megatron.bridge.training.config.RerunStateMachineConfig
+  check_for_nan_in_loss: true
+  check_for_spiky_loss: false
+  error_injection_rate: 0
+  error_injection_type: transient_error
+  rerun_mode: disabled
+  spiky_loss_factor: 10.0
+rng:
+  _target_: megatron.training.config.common_config.RNGConfig
+  data_parallel_random_init: false
+  inference_rng_tracker: false
+  seed: 1234
+  te_rng_tracker: false
+scheduler: null
+straggler: null
+tensor_inspect: null
+tokenizer:
+  _target_: megatron.bridge.training.tokenizers.config.TokenizerConfig
+  chat_template: null
+  force_system_message: false
+  hf_tokenizer_kwargs: {}
+  image_tag_type: null
+  make_vocab_size_divisible_by: 1
+  merge_file: null
+  metadata_path: null
+  null_tokenizer_eod_id: null
+  null_tokenizer_pad_id: -1
+  pad_vocab_size: false
+  padded_vocab_size: null
+  rank: 0
+  sp_tokenizer_kwargs: {}
+  special_tokens: null
+  tensor_model_parallel_size: 1
+  tiktoken_num_special_tokens: 1000
+  tiktoken_pattern: null
+  tokenizer_hf_no_include_special_tokens: false
+  tokenizer_hf_no_use_fast: false
+  tokenizer_model: /home/junjie.ye/workspace/humanoid_research/cua/models/hf_exports/qwen35-vl-4b-cua-3x-full-func-nothink-1ep-20260907/iter_0000422
+  tokenizer_prompt_format: null
+  tokenizer_sentencepiece_ignore_extra_whitespaces: true
+  tokenizer_sentencepiece_legacy: false
+  tokenizer_type: HuggingFaceTokenizer
+  trust_remote_code: false
+  vocab_extra_ids: 0
+  vocab_file: null
+  vocab_size: null
+train: null
+validation:
+  _target_: megatron.training.config.training_config.ValidationConfig
+  eval_global_batch_size: null
+  eval_interval: null
+  eval_iters: 100
+  eval_micro_batch_size: null
+  full_validation: false
+  multiple_validation_sets: false
+  skip_train: false
+  start_eval_at_iter: null
+  test_mode: false
+  validation_set_names: null

tokenizer/chat_template.jinja

@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+    {%- if content is string %}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping %}
+        {%- for item in content %}
+            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain images.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+            {%- elif 'video' in item or item.type == 'video' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain videos.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Video ' ~ video_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+            {%- elif 'text' in item %}
+                {{- item.text }}
+            {%- else %}
+                {{- raise_exception('Unexpected item type in content.') }}
+            {%- endif %}
+        {%- endfor %}
+    {%- elif content is none or content is undefined %}
+        {{- '' }}
+    {%- else %}
+        {{- raise_exception('Unexpected content type.') }}
+    {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+    {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>" }}
+    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {%- if content %}
+            {{- '\n\n' + content }}
+        {%- endif %}
+    {%- endif %}
+    {{- '<|im_end|>\n' }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" %}
+        {%- set content = render_content(message.content, false)|trim %}
+        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
+            {%- set ns.multi_step_tool = false %}
+            {%- set ns.last_query_index = index %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+    {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content, true)|trim %}
+    {%- if message.role == "system" %}
+        {%- if not loop.first %}
+            {{- raise_exception('System message must be at the beginning.') }}
+        {%- endif %}
+    {%- elif message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {%- if loop.first %}
+                    {%- if content|trim %}
+                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- else %}
+                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- endif %}
+                {%- else %}
+                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- endif %}
+                {%- if tool_call.arguments is defined %}
+                    {%- for args_name, args_value in tool_call.arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.previtem and loop.previtem.role != "tool" %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if not loop.last and loop.nextitem.role != "tool" %}
+            {{- '<|im_end|>\n' }}
+        {%- elif loop.last %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- else %}
+        {{- raise_exception('Unexpected message role.') }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- else %}
+        {{- '<think>\n' }}
+    {%- endif %}
+{%- endif %}

tokenizer/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325

tokenizer/tokenizer_config.json

@@ -0,0 +1,34 @@
+{
+  "add_prefix_space": false,
+  "audio_bos_token": "<|audio_start|>",
+  "audio_eos_token": "<|audio_end|>",
+  "audio_token": "<|audio_pad|>",
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "image_token": "<|image_pad|>",
+  "is_local": true,
+  "local_files_only": true,
+  "model_max_length": 262144,
+  "model_specific_special_tokens": {
+    "audio_bos_token": "<|audio_start|>",
+    "audio_eos_token": "<|audio_end|>",
+    "audio_token": "<|audio_pad|>",
+    "image_token": "<|image_pad|>",
+    "video_token": "<|video_pad|>",
+    "vision_bos_token": "<|vision_start|>",
+    "vision_eos_token": "<|vision_end|>"
+  },
+  "pad_token": "<|endoftext|>",
+  "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+  "processor_class": "Qwen3VLProcessor",
+  "sep_token": "<|im_end|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}

train_state.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aee94cd44a5abd78b6a7dbf199b514f17967f64c0403da1e17e7891c70f95788
+size 3461