checkpoint:
  broadcast_via_filesystem: 'False'
  dcp_async_mode_enabled: 'False'
  enable_gcs_patch_in_boto3: 'False'
  jit:
    device: cuda
    dtype: bfloat16
    enabled: 'False'
    input_shape: null
    strict: 'True'
  keys_not_to_resume: []
  keys_to_skip_loading: []
  load_ema_to_reg: 'False'
  load_from_object_store:
    bucket: bucket
    credentials: credentials/s3_checkpoint.secret
    enabled: 'True'
  load_path: ''
  load_training_state: 'False'
  only_load_scheduler_state: 'False'
  save_iter: '500'
  save_to_object_store:
    bucket: bucket
    credentials: credentials/s3_checkpoint.secret
    enabled: 'True'
  strict_resume: 'True'
  type:
    _target_: <class 'cosmos_transfer2._src.interactive.checkpointer.dcp_distill.DistributedCheckpointer'>
    callbacks: null
    disable_async: 'False'
  verbose: 'True'
dataloader_train:
  _target_: <class 'cosmos_transfer2._src.predict2.datasets.joint_dataloader.IterativeJointDataLoader'>
  dataloaders:
    image_data:
      dataloader:
        _target_: <function get_cached_replay_dataloader at 0x7f5320132ef0>
        batch_size: '2'
        cache_augment_fn: null
        cache_replay_name: image_dataloader
        cache_size: '32'
        concat_size: '1'
        dataset:
          _target_: <function get_image_dataset at 0x7f53201330a0>
          len_t5: '512'
          resolution: '512'
          t5_dim: '1024'
        num_workers: '8'
        pin_memory: 'True'
        shuffle: 'False'
        use_cache: 'False'
        webdataset: 'False'
      ratio: '1'
    video_data:
      dataloader:
        _target_: <function get_cached_replay_dataloader at 0x7f5320132ef0>
        batch_size: '1'
        cache_augment_fn: null
        cache_replay_name: video_dataloader
        cache_size: '32'
        concat_size: '1'
        dataset:
          _target_: <function get_video_dataset at 0x7f5320133400>
          len_t5: '512'
          num_video_frames: '136'
          resolution: '512'
          t5_dim: '1024'
        num_workers: '8'
        pin_memory: 'True'
        shuffle: 'False'
        use_cache: 'False'
        webdataset: 'False'
      ratio: '1'
  dataset:
    augmentor_name: video_basic_augmentor_v2_with_control
    caption_type: t2w_qwen2p5_7b
    control_input_type: edge
    dataset_resolution_type: gt720p
    embedding_type: null
    max_fps_thres: '60'
    min_fps_thres: '10'
    num_video_frames: '93'
    resolution: '720'
    use_native_fps: 'True'
    video_decoder_name: video_naive_bytes
  num_workers: '4'
dataloader_val: null
defaults:
- _self_
- data_train: null
- data_val: null
- optimizer: fusedadamw
- scheduler: lambdalinear
- callbacks: basic
- model: ddp
- net: null
- net_fake_score: null
- net_teacher: null
- net_discriminator_head: null
- conditioner: video_prediction_conditioner
- ema: power
- tokenizer: wan2pt1_tokenizer
- checkpoint: s3
- ckpt_type: dcp_distill
- condition_postprocessor: null
- experiment: null
job:
  cluster: null
  group: cosmos3_interactive
  name: dmd2_trigflow_distill_cosmos_transfer2p5_2B_bidirectional_edge
  project: cosmos_interactive
  wandb_mode: online
model:
  _recursive_: 'False'
  _target_: <class 'cosmos_transfer2._src.interactive.methods.distribution_matching.dmd2.DMD2Model'>
  config:
    base_load_from: null
    condition_postprocessor:
      _target_: <class 'cosmos_transfer2._src.interactive.conditioner.condition_postprocessors.ControlConditionPostprocessor'>
      hint_keys:
      - edge
    conditional_frames_probs:
      0: 0.6
      1: 0.2
      2: 0.2
    conditioner:
      _target_: <class 'cosmos_transfer2._src.transfer2.configs.vid2vid_transfer.defaults.conditioner.ControlVideo2WorldConditioner'>
      control_input_depth:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_depth
        output_key: control_input_depth
      control_input_depth_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_depth_mask
        output_key: control_input_depth_mask
      control_input_edge:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_edge
        output_key: control_input_edge
      control_input_edge_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_edge_mask
        output_key: control_input_edge_mask
      control_input_inpaint:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_inpaint
        output_key: control_input_inpaint
      control_input_inpaint_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_inpaint_mask
        output_key: control_input_inpaint_mask
      control_input_seg:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_seg
        output_key: control_input_seg
      control_input_seg_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_seg_mask
        output_key: control_input_seg_mask
      control_input_vis:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_vis
        output_key: control_input_vis
      control_input_vis_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: control_input_vis_mask
        output_key: control_input_vis_mask
      fps:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: fps
        output_key: fps
      padding_mask:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.ReMapkey'>
        dropout_rate: '0.0'
        dtype: null
        input_key: padding_mask
        output_key: padding_mask
      text:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.TextAttr'>
        credential_path: credentials/s3_training.secret
        dropout_rate: '0.2'
        empty_string_embeddings_path: s3://bucket/predict2_assets/reason1_empty_string_embeddings.pt
        input_key:
        - t5_text_embeddings
        use_empty_string: 'False'
      use_video_condition:
        _target_: <class 'cosmos_transfer2._src.predict2.conditioner.BooleanFlag'>
        dropout_rate: '0.0'
        input_key: fps
        output_key: use_video_condition
    ema:
      enabled: false
      iteration_shift: 0
      rate: 0.1
    fsdp_shard_size: 1
    grad_clip: true
    input_caption_key: ai_caption
    input_data_key: video
    input_image_key: images
    intermediate_feature_ids: null
    load_teacher_weights: false
    loss_reduce: mean
    loss_scale_GAN_discriminator: 1.0
    loss_scale_GAN_generator: 1.0
    loss_scale_fake_score: 1.0
    loss_scale_sid: 1.0
    max_num_conditional_frames: 2
    min_num_conditional_frames: 0
    multiply_noise_by_video_len: true
    neg_embed_path: ''
    neg_prompt_str: The video captures a game playing, with bad crappy graphics and
      cartoonish frames. It represents a recording of old outdated games. The lighting
      looks very fake. The textures are very raw and basic. The geometries are very
      primitive. The images are very pixelated and of poor CG quality. There are many
      subtitles in the footage. Overall, the video is unrealistic at all.
    net:
      _target_: <class 'cosmos_transfer2._src.transfer2.networks.minimal_v4_lvg_dit_control_vace.MinimalV4LVGControlVaceDiT'>
      adaln_lora_dim: '256'
      atten_backend: minimal_a2a
      concat_padding_mask: 'True'
      condition_strategy: spaced
      crossattn_emb_channels: '1024'
      crossattn_proj_in_channels: '100352'
      extra_per_block_abs_pos_emb: 'False'
      in_channels: '16'
      max_frames: '128'
      max_img_h: '240'
      max_img_w: '240'
      mlp_ratio: '4.0'
      model_channels: '2048'
      num_blocks: '28'
      num_control_branches: '1'
      num_heads: '16'
      num_max_modalities: '8'
      out_channels: '16'
      patch_spatial: '2'
      patch_temporal: '1'
      pos_emb_cls: rope3d
      pos_emb_interpolation: crop
      pos_emb_learnable: 'True'
      rope_enable_fps_modulation: 'False'
      rope_h_extrapolation_ratio: '3.0'
      rope_t_extrapolation_ratio: '1.0'
      rope_w_extrapolation_ratio: '3.0'
      sac_config:
        every_n_blocks: 1
        mode: predict2_2b_720_aggressive
      separate_embedders: 'False'
      spatial_compression_factor: '8'
      timestep_scale: '1.0'
      use_adaln_lora: 'True'
      use_after_proj_for_multi_branch: 'True'
      use_crossattn_projection: 'True'
      use_cuda_graphs: 'False'
      use_input_hint_block: 'False'
      use_wan_fp32_strategy: 'True'
      vace_block_every_n: '7'
      vace_has_mask: 'False'
    net_discriminator_head: null
    net_fake_score:
      _target_: <class 'cosmos_transfer2._src.transfer2.networks.minimal_v4_lvg_dit_control_vace.MinimalV4LVGControlVaceDiT'>
      adaln_lora_dim: '256'
      atten_backend: minimal_a2a
      concat_padding_mask: 'True'
      condition_strategy: spaced
      crossattn_emb_channels: '1024'
      crossattn_proj_in_channels: '100352'
      extra_per_block_abs_pos_emb: 'False'
      in_channels: '16'
      max_frames: '128'
      max_img_h: '240'
      max_img_w: '240'
      mlp_ratio: '4.0'
      model_channels: '2048'
      num_blocks: '28'
      num_control_branches: '1'
      num_heads: '16'
      num_max_modalities: '8'
      out_channels: '16'
      patch_spatial: '2'
      patch_temporal: '1'
      pos_emb_cls: rope3d
      pos_emb_interpolation: crop
      pos_emb_learnable: 'True'
      rope_enable_fps_modulation: 'False'
      rope_h_extrapolation_ratio: '3.0'
      rope_t_extrapolation_ratio: '1.0'
      rope_w_extrapolation_ratio: '3.0'
      sac_config:
        every_n_blocks: 1
        mode: predict2_2b_720_aggressive
      separate_embedders: 'False'
      spatial_compression_factor: '8'
      timestep_scale: '1.0'
      use_adaln_lora: 'True'
      use_after_proj_for_multi_branch: 'True'
      use_crossattn_projection: 'True'
      use_cuda_graphs: 'False'
      use_input_hint_block: 'False'
      use_wan_fp32_strategy: 'True'
      vace_block_every_n: '7'
      vace_has_mask: 'False'
    net_teacher:
      _target_: <class 'cosmos_transfer2._src.transfer2.networks.minimal_v4_lvg_dit_control_vace.MinimalV4LVGControlVaceDiT'>
      adaln_lora_dim: '256'
      atten_backend: minimal_a2a
      concat_padding_mask: 'True'
      condition_strategy: spaced
      crossattn_emb_channels: '1024'
      crossattn_proj_in_channels: '100352'
      extra_per_block_abs_pos_emb: 'False'
      in_channels: '16'
      max_frames: '128'
      max_img_h: '240'
      max_img_w: '240'
      mlp_ratio: '4.0'
      model_channels: '2048'
      num_blocks: '28'
      num_control_branches: '1'
      num_heads: '16'
      num_max_modalities: '8'
      out_channels: '16'
      patch_spatial: '2'
      patch_temporal: '1'
      pos_emb_cls: rope3d
      pos_emb_interpolation: crop
      pos_emb_learnable: 'True'
      rope_enable_fps_modulation: 'False'
      rope_h_extrapolation_ratio: '3.0'
      rope_t_extrapolation_ratio: '1.0'
      rope_w_extrapolation_ratio: '3.0'
      sac_config:
        every_n_blocks: 1
        mode: predict2_2b_720_aggressive
      separate_embedders: 'False'
      spatial_compression_factor: '8'
      timestep_scale: '1.0'
      use_adaln_lora: 'True'
      use_after_proj_for_multi_branch: 'True'
      use_crossattn_projection: 'True'
      use_cuda_graphs: 'False'
      use_input_hint_block: 'False'
      use_wan_fp32_strategy: 'True'
      vace_block_every_n: '7'
      vace_has_mask: 'False'
    noise_level_parameterization: trigflow
    optimizer_discriminator_config:
      _target_: <function get_base_optimizer at 0x7f537145de10>
      betas:
      - '0.9'
      - '0.999'
      capturable: 'True'
      eps: 1e-08
      lr: 2e-07
      master_weights: 'True'
      model: null
      optim_type: fusedadam
      weight_decay: '0.01'
    optimizer_fake_score_config:
      _target_: <function get_base_optimizer at 0x7f537145de10>
      betas:
      - '0.9'
      - '0.999'
      capturable: 'True'
      eps: 1e-08
      lr: 2e-07
      master_weights: 'True'
      model: null
      optim_type: fusedadam
      weight_decay: '0.01'
    precision: bfloat16
    rectified_flow_loss_weight_uniform: false
    rectified_flow_t_scaling_factor: 1.0
    replace_cond_output_with_gt: true
    resolution: '720'
    scaling: rectified_flow
    sde:
      _target_: <class 'cosmos_transfer2._src.imaginaire.modules.edm_sde.EDMSDE'>
      p_mean: '-0.8'
      p_std: '1.6'
      sigma_max: '80'
      sigma_min: '0.0002'
    sde_D:
      _target_: <class 'cosmos_transfer2._src.imaginaire.modules.edm_sde.EDMSDE'>
      p_mean: '0.0'
      p_std: '1.6'
      sigma_max: '80'
      sigma_min: '0.0002'
    selected_sampling_time:
    - 1.5707963267948966
    - 1.5042281630190728
    - 1.373400766945016
    - 1.0303768265243125
    sigma_conditional: 0.0001
    sigma_data: 1.0
    state_ch: 16
    state_t: 24
    student_load_from: null
    student_update_freq: 5
    teacher_guidance: 3.0
    teacher_load_from:
      credentials: credentials/s3_checkpoint.secret
      load_path: s3://bucket/cosmos_transfer2/vid2vid_2B_control/edge_720p_t24_spaced_layer4_cr1pt1_sdev2_lowsigma0.05_nonuniform_hqv3p1_20250714_64N_rectified_flow/checkpoints/iter_000029000/model
    text_encoder_class: reason1p1_7B
    text_encoder_config:
      ckpt_path: s3://bucket/cosmos_reasoning1/sft_exp700/sft_exp721-1_qwen7b_tl_721_5vs5_s3_balanced_n32_resume_16k/checkpoints/iter_000016000/model/
      compute_online: true
      embedding_concat_strategy: full_concat
      model_config:
        _target_: <class 'cosmos_transfer2._src.predict2.text_encoders.reason1.QwenVLBaseModel'>
        model_config:
          _target_: cosmos_transfer2._src.reason1.configs.default.model_config_qwen.QwenModelConfig
          activation_checkpoint:
            mode: selective
            models: vlm
            selective_ac_option: op
          add_answer_tag: 'True'
          add_cross_attention: 'False'
          add_image_start_end_tag: 'False'
          add_tile_tag: 'False'
          architectures:
          - Qwen2_5_VLForConditionalGeneration
          attention_dropout: '0.0'
          attn_implementation: flash_attention_2
          attn_implementation_autoset: 'True'
          aux_loss_coeff: '0.0'
          bad_words_ids: null
          begin_suppress_tokens: null
          bos_token_id: '151643'
          cache_dir: null
          checkpoint:
            async_mode: disabled
            create_seed_checkpoint: false
            enable_checkpoint: false
            export_dtype: float32
            folder: checkpoint
            interval: 500
            interval_type: steps
            model_weights_only: false
          chunk_size_feed_forward: '0'
          ckpt_dir: null
          ckpt_path: null
          comm:
            init_timeout_seconds: 300
            trace_buf_size: 20000
            train_timeout_seconds: 100
          cp_size: null
          cross_attention_hidden_size: null
          decoder_start_token_id: null
          deterministic: 'False'
          diversity_penalty: '0.0'
          do_sample: 'False'
          early_stopping: 'False'
          encoder_no_repeat_ngram_size: '0'
          eos_token_id: '151645'
          ep_size: null
          experimental:
            enable_async_tensor_parallel: false
            enable_compiled_autograd: false
            pipeline_parallel_degree: 1
          exponential_decay_length_penalty: null
          finetuning_task: null
          float8:
            enable_float8_linear: false
          forced_bos_token_id: null
          forced_eos_token_id: null
          freeze_llm: 'False'
          freeze_mm_projector: 'False'
          freeze_vision_encoder: 'False'
          fsdp_enabled: 'False'
          hidden_act: silu
          hidden_size: '3584'
          id2label:
            0: LABEL_0
            1: LABEL_1
          image_token_id: '151655'
          initializer_range: '0.02'
          intermediate_size: '18944'
          is_decoder: 'False'
          is_encoder_decoder: 'False'
          label2id:
            LABEL_0: '0'
            LABEL_1: '1'
          length_penalty: '1.0'
          loss_per_token: 'True'
          max_batch_size: '1'
          max_length: '20'
          max_position_embeddings: '128000'
          max_seq_len: '128000'
          max_window_layers: '28'
          min_length: '0'
          mm_projector: null
          model_type: qwen2_5_vl
          name_or_path: Qwen/Qwen2.5-VL-7B-Instruct
          no_repeat_ngram_size: '0'
          num_attention_heads: '28'
          num_beam_groups: '1'
          num_beams: '1'
          num_hidden_layers: '28'
          num_key_value_heads: '4'
          num_return_sequences: '1'
          num_tiles: '1'
          optimizer:
            early_step_in_backward: false
            end_lr: 2.5e-05
            fused: false
            init_lr: 1.0e-05
            lr: 0.0003
            lr_multiplier_llm: 1.0
            lr_multiplier_mm_projector: 1.0
            lr_multiplier_vision_encoder: 0.1
            name: AdamW
          output_attentions: 'False'
          output_hidden_states: 'True'
          output_scores: 'False'
          pad_token_id: null
          precision: bfloat16
          prefix: null
          prepend_padding: 'False'
          problem_type: null
          pruned_heads: _Nothing.NOTHING
          remove_invalid_values: 'False'
          repetition_penalty: '1.0'
          return_dict: 'True'
          return_dict_in_generate: 'False'
          rms_norm_eps: 1e-06
          rope_scaling:
            mrope_section:
            - '16'
            - '24'
            - '24'
            rope_type: default
            type: default
          rope_theta: '1000000.0'
          s3_credential_path: credentials/pbss_dir.secret
          seed: '0'
          sep_token_id: null
          sliding_window: '32768'
          suppress_tokens: null
          task_specific_params: null
          temperature: '1.0'
          tf_legacy_loss: 'False'
          tie_encoder_decoder: 'False'
          tie_word_embeddings: 'False'
          tile_tag_type: space_separated
          tokenizer_class: null
          tokenizer_type: Qwen/Qwen2.5-VL-7B-Instruct
          top_k: '50'
          top_p: '1.0'
          torch_dtype: bfloat16
          torchscript: 'False'
          training:
            compile: false
            context_parallel_degree: 1
            data_parallel_replicate_degree: 1
            data_parallel_shard_degree: -1
            disable_loss_parallel: false
            enable_cpu_offload: false
            fsdp_reshard_after_forward: default
            mixed_precision_param: bfloat16
            mixed_precision_reduce: float32
            steps: 400000
            tensor_parallel_degree: 1
            use_cosine_decay: false
            use_linear_decay: true
            warmup_steps: 1000
          training_seq_len: '4096'
          transformers_version: 4.51.0.dev0
          typical_p: '1.0'
          use_bfloat16: 'False'
          use_cache: 'False'
          use_fsdp2: 'True'
          use_return_dict: 'True'
          use_rope_from_torchtitan: 'False'
          use_sliding_window: 'False'
          video_token_id: '151656'
          vision_config:
            _target_: cosmos_transfer2._src.reason1.configs.default.model_config_qwen.QwenVisionConfig
            add_cross_attention: 'False'
            architectures: null
            attn_implementation: flash_attention_2
            attn_implementation_autoset: 'True'
            bad_words_ids: null
            begin_suppress_tokens: null
            bos_token_id: null
            chunk_size_feed_forward: '0'
            cross_attention_hidden_size: null
            decoder_start_token_id: null
            depth: '32'
            diversity_penalty: '0.0'
            do_sample: 'False'
            early_stopping: 'False'
            embed_dim: null
            encoder_no_repeat_ngram_size: '0'
            eos_token_id: null
            exponential_decay_length_penalty: null
            finetuning_task: null
            forced_bos_token_id: null
            forced_eos_token_id: null
            fullatt_block_indexes:
            - '7'
            - '15'
            - '23'
            - '31'
            hidden_act: silu
            hidden_size: '1280'
            id2label:
              0: LABEL_0
              1: LABEL_1
            in_channels: '3'
            in_chans: '3'
            intermediate_size: '3420'
            is_decoder: 'False'
            is_encoder_decoder: 'False'
            label2id:
              LABEL_0: '0'
              LABEL_1: '1'
            length_penalty: '1.0'
            max_length: '20'
            min_length: '0'
            mlp_ratio: null
            model_type: qwen2_5_vl
            name_or_path: ''
            no_repeat_ngram_size: '0'
            num_beam_groups: '1'
            num_beams: '1'
            num_heads: '16'
            num_return_sequences: '1'
            out_hidden_size: '3584'
            output_attentions: 'False'
            output_hidden_states: 'False'
            output_scores: 'False'
            pad_token_id: null
            patch_size: '14'
            prefix: null
            problem_type: null
            pruned_heads: _Nothing.NOTHING
            remove_invalid_values: 'False'
            repetition_penalty: '1.0'
            return_dict: 'True'
            return_dict_in_generate: 'False'
            sep_token_id: null
            spatial_merge_size: '2'
            spatial_patch_size: '14'
            suppress_tokens: null
            task_specific_params: null
            temperature: '1.0'
            temporal_patch_size: '2'
            tf_legacy_loss: 'False'
            tie_encoder_decoder: 'False'
            tie_word_embeddings: 'True'
            tokenizer_class: null
            tokens_per_second: '2'
            top_k: '50'
            top_p: '1.0'
            torch_dtype: bfloat16
            torchscript: 'False'
            typical_p: '1.0'
            use_bfloat16: 'False'
            window_size: '112'
          vision_encoder: openai/clip-vit-base-patch32
          vision_encoder_config:
            depth_init: true
            dim: 1024
            ffn_dim_multiplier: null
            head_dim: null
            hidden_act: null
            hidden_dim: 4096
            image_size: 1024
            image_token_id: null
            multiple_of: null
            n_heads: 16
            n_kv_heads: null
            n_layers: 24
            norm_eps: 1.0e-05
            norm_type: rmsnorm
            num_channels: 3
            patch_size: 16
            proj_bias: null
            qkv_bias: null
            rope_theta: 10000.0
            use_cache: false
            use_rope_from_torchtitan: false
          vision_encoder_in_channels: '3'
          vision_end_token_id: '151653'
          vision_start_token_id: '151652'
          vision_token_id: '151654'
          vocab_size: '152064'
          z_loss_coeff: '0.0'
        tokenizer:
          _target_: <function build_tokenizer at 0x7f5375b13ac0>
          cache_dir: null
          tokenizer_type: Qwen/Qwen2.5-VL-7B-Instruct
      n_layers_per_group: 5
      s3_credential_path: credentials/s3_checkpoint.secret
    timestep_shift: 5.0
    tokenizer:
      _target_: <class 'cosmos_transfer2._src.predict2.tokenizers.wan2pt1.Wan2pt1VAEInterface'>
      chunk_duration: '81'
      load_mean_std: 'False'
      name: wan2pt1_tokenizer
    use_clean_cond_timesteps: true
    use_neg_prompt_str: false
    use_torch_compile: false
    vis_debug: false
    warmup_steps: 1
model_parallel:
  _cpu_offloading_context: null
  async_tensor_model_parallel_allreduce: false
  autocast_dtype: torch.float32
  barrier_with_L1_time: true
  batch_p2p_comm: true
  batch_p2p_sync: true
  bf16: false
  context_parallel_size: 4
  cpu_offloading: false
  cpu_offloading_activations: true
  cpu_offloading_double_buffering: false
  cpu_offloading_num_layers: 0
  cpu_offloading_weights: true
  cross_entropy_fusion_impl: native
  cross_entropy_loss_fusion: false
  deallocate_pipeline_outputs: false
  defer_embedding_wgrad_compute: false
  delay_wgrad_compute: false
  deterministic_mode: false
  enable_autocast: false
  expert_model_parallel_size: 1
  expert_tensor_parallel_size: 1
  finalize_model_grads_func: null
  fp16: false
  grad_scale_func: null
  grad_sync_func: null
  gradient_accumulation_fusion: false
  hierarchical_context_parallel_sizes: null
  microbatch_group_size_per_vp_stage: 1
  moe_extended_tp: false
  no_sync_func: null
  num_microbatches_with_partial_activation_checkpoints: null
  overlap_moe_expert_parallel_comm: false
  overlap_p2p_comm: false
  overlap_p2p_comm_warmup_flush: false
  param_sync_func: null
  params_dtype: torch.float32
  perform_initialization: true
  pipeline_dtype: null
  pipeline_model_parallel_comm_backend: null
  pipeline_model_parallel_size: 1
  sequence_parallel: false
  tensor_model_parallel_size: 1
  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_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
  use_cpu_initialization: false
  use_ring_exchange_p2p: false
  use_te_rng_tracker: false
  variable_seq_lengths: false
  virtual_pipeline_model_parallel_size: null
  wgrad_deferral_limit: 0
optimizer:
  _target_: <function get_base_optimizer at 0x7f537145de10>
  betas:
  - '0.9'
  - '0.999'
  capturable: 'True'
  eps: 1e-08
  lr: 1e-06
  master_weights: 'True'
  model: null
  optim_type: fusedadam
  weight_decay: '0.01'
scheduler:
  _target_: <class 'cosmos_transfer2._src.imaginaire.functional.lr_scheduler.LambdaLinearScheduler'>
  cycle_lengths:
  - '400000'
  f_max:
  - '0.99'
  f_min:
  - '0.4'
  f_start:
  - 1e-06
  verbosity_interval: '0'
  warm_up_steps:
  - '100'
trainer:
  callbacks:
    compile_tokenizer:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.compile_tokenizer.CompileTokenizer'>
      compile_after_iterations: '4'
      dynamic: 'False'
      enabled: 'True'
    dataloader_speed:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.dataloading_monitor.DetailedDataLoadingSpeedMonitor'>
      every_n: '10'
      save_s3: 'True'
      step_size: '1'
    device_monitor:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.device_monitor.DeviceMonitor'>
      every_n: '10'
      log_memory_detail: 'True'
      save_s3: 'True'
      step_size: '1'
      upload_every_n_mul: '10'
    every_n_sample_ema:
      _target_: <class 'cosmos_transfer2._src.interactive.callbacks.every_n_draw_sample_distilled.EveryNDrawSample'>
      do_sample_val_prompts: 'True'
      every_n: '250'
      is_ema: 'True'
      is_image: 'False'
      is_sample: 'True'
      n_sample_to_save: '128'
      num_prompts: '4'
      num_samples_per_prompt: '3'
      num_sampling_step: '4'
      num_sampling_step_teacher: '35'
      save_s3: 'True'
      show_all_frames: 'False'
      step_size: '1'
      use_negative_prompt: 'False'
    every_n_sample_reg:
      _target_: <class 'cosmos_transfer2._src.interactive.callbacks.every_n_draw_sample_distilled.EveryNDrawSample'>
      do_sample_val_prompts: 'True'
      every_n: '250'
      is_ema: 'False'
      is_image: 'False'
      is_sample: 'True'
      n_sample_to_save: '128'
      num_prompts: '4'
      num_samples_per_prompt: '3'
      num_sampling_step: '4'
      num_sampling_step_teacher: '35'
      save_s3: 'True'
      show_all_frames: 'False'
      step_size: '1'
      use_negative_prompt: 'False'
    grad_clip:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.grad_clip.GradClip'>
      clip_norm: '1.0'
      force_finite: 'True'
    heart_beat:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.heart_beat.HeartBeat'>
      every_n: '10'
      save_s3: 'True'
      step_size: '1'
      update_interval_in_minute: '20'
    iter_speed:
      _target_: <class 'cosmos_transfer2._src.predict2.callbacks.iter_speed.IterSpeed'>
      every_n: '10'
      hit_thres: '100'
      save_s3: 'True'
      save_s3_every_log_n: '10'
    low_prec:
      _target_: <class 'cosmos_transfer2._src.imaginaire.utils.callback.LowPrecisionCallback'>
      config: null
      trainer: null
      update_iter: '1'
    manual_gc:
      _target_: <class 'cosmos_transfer2._src.imaginaire.callbacks.manual_gc.ManualGarbageCollection'>
      every_n: '5'
      warm_up: '5'
    wandb:
      _target_: <class 'cosmos_transfer2._src.interactive.callbacks.wandb_log_rcm.WandbCallback'>
      logging_iter_multipler: '1'
      save_logging_iter_multipler: '10'
      save_s3: 'True'
    wandb_10x:
      _target_: <class 'cosmos_transfer2._src.interactive.callbacks.wandb_log_rcm.WandbCallback'>
      logging_iter_multipler: '10'
      save_logging_iter_multipler: '1'
      save_s3: 'True'
  compile_config:
    recompile_limit: '8'
    use_duck_shape: 'True'
  cudnn:
    benchmark: 'True'
    deterministic: 'False'
  ddp:
    broadcast_buffers: 'True'
    find_unused_parameters: 'False'
    static_graph: 'True'
  distributed_parallelism: fsdp
  grad_accum_iter: '1'
  grad_scaler_args:
    enabled: 'False'
  logging_iter: '10'
  max_iter: '30000'
  max_val_iter: null
  memory_format: torch.preserve_format
  profiling:
    enable_memory_snapshot: 'False'
    enable_profiling: 'False'
    profile_freq: '1'
    profile_memory: 'False'
    record_shape: 'False'
    save_s3: 'False'
    target_ranks:
    - '0'
    - '1'
    - '2'
    - '3'
    - '4'
    - '5'
    - '6'
    - '7'
    with_modules: 'True'
    with_stack: 'True'
  run_validation: 'False'
  run_validation_on_start: 'False'
  seed: '0'
  straggler_detection:
    analyze_backward: 'True'
    analyze_dataloading: 'True'
    analyze_forward: 'True'
    analyze_optimizer: 'True'
    enabled: 'False'
    max_diff: '2.0'
    profile_freq: '1'
    raise_error: 'True'
    report_freq: '100'
  timeout_period: '999999999'
  type: <class 'cosmos_transfer2._src.interactive.trainer.trainer_distillation.ImaginaireDistillationTrainer'>
  validation_iter: '100'
upload_reproducible_setup: 'True'