Upload folder using huggingface_hub

.gitattributes

@@ -451,3 +451,14 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 2026.04.11/08.20.17_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260411_082153-ecuxlp0j/run-ecuxlp0j.wandb filter=lfs diff=lfs merge=lfs -text
 2026.04.11/08.20.51_train_llmbc_lowdim_maze2d-medium-v0/logs.json.txt filter=lfs diff=lfs merge=lfs -text
 2026.04.11/08.20.51_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260411_082155-p2755cs0/run-p2755cs0.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/logs.json.txt filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/run-6okd5etw.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/run-981b5iwy.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062758-yrc5uqpn/run-yrc5uqpn.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.26.55_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062800-40wpvpiq/run-40wpvpiq.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.09_train_llm_diffusion_unet_lowdim_sweep-v2/logs.json.txt filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.09_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_063146-ar8xzudz/run-ar8xzudz.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.11_train_llm_diffusion_unet_lowdim_sweep-v2/logs.json.txt filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.11_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_063229-tqhbov1v/run-tqhbov1v.wandb filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.13_train_llm_diffusion_unet_lowdim_sweep-v2/logs.json.txt filter=lfs diff=lfs merge=lfs -text
+2026.04.14/06.30.13_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_063231-pm30rgpl/run-pm30rgpl.wandb filter=lfs diff=lfs merge=lfs -text

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/.hydra/config.yaml

@@ -0,0 +1,165 @@
+name: train_llmbc_lowdim
+_target_: llmbc.workspace.train_llmbc_lowdim_workspace.TrainLLMBCLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llmbc_lowdim_policy.LLMBCLowdimPolicy
+  model:
+    _target_: llmbc.model.policy.policy_mlp.PolicyMLP
+    input_size: ${eval:'${n_obs_steps}*${obs_dim}'}
+    hidden_size:
+    - 256
+    - 256
+    - 256
+    - 256
+    output_size: ${eval:'${n_action_steps}*${action_dim}'}
+    activation: tanh
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${task.obs_dim}
+      action_dim: ${task.action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${n_action_steps}
+  loss_bc_weight: 1.0
+  loss_llm_weight: 0.01
+  horizon: ${horizon}
+  n_obs_steps: ${n_obs_steps}
+  n_action_steps: ${n_action_steps}
+  normalize_llm_loss: true
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 44
+  debug: false
+  resume: false
+  lr_scheduler: cosine
+  lr_warmup_steps: 10
+  num_epochs: 801
+  gradient_accumulate_every: 8
+  grad_norm_clip: 0.5
+  rollout_every: 4
+  checkpoint_every: 4
+  val_every: 1
+  sample_every: 4
+  sample_max_batch: 128
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training-mix
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: maze2d-medium-v0
+  obs_dim: 6
+  action_dim: 2
+  env_runner:
+    _target_: llmbc.env_runner.pointmaze_lowdim_runner.PointmazeLowdimRunner
+    env_name: llf-pointmaze-maze2d-medium-v0
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 250
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.99
+  dataset:
+    _target_: llmbc.dataset.pointmaze_lowdim_dataset.PointmazeLowdimDataset
+    data_path: datasets/maze2d-medium-v0-mix(csp5).pt
+    data_path2: datasets/maze2d-medium-v0.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.02
+    dummy_normalizer: true
+  instructor:
+    _target_: llmbc.translator.instructor.pointmaze_instructor.maze2d_medium_v0_instructor.Maze2dMediumV0Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.10.16/11.53.22_train_llm_lowdim_maze2d-medium-v0/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-maze2d-medium-v0/checkpoint-11850
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - training.seed=44
+  job:
+    name: train
+    chdir: null
+    override_dirname: training.seed=44
+    id: ???
+    num: ???
+    config_name: llmbc_maze2d-medium-v0_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- training.seed=44

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/epoch=0264-test_success_rate=0.600.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9569fc1d21c7f33064b324744e05625ad831ed80b5ec0bacbb398c7ce6756a3c
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/epoch=0360-test_success_rate=0.580.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f659d8b720b1bd263c79e37a73077fbadb02d4d487ab5faf6eb8e9322e6ff02
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/epoch=0420-test_success_rate=0.580.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a91bc240b8870982b0650f238bbe7bed8dfd40ff54a4795e493c5c3a28bb2233
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/epoch=0440-test_success_rate=0.600.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d24f2c8cca5303c5d03d7b6cfca5341b36985253ff3f705e952c518c3c798d18
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/epoch=0580-test_success_rate=0.600.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b1c41022079150e20786f934ec85d6bcfcebae53c0a1f00bfac0283ff3cf78c
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/checkpoints/latest.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfdcc92a992e8155fc1a8d2218c91b31bf8643fd8aa00149b22e73eb994d463b
+size 2432984

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/logs.json.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57b8e33c8176eceb59054374e02d710213e4b99dc285b86b421aa7cd4f2b924a
+size 354344598

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/train.log

@@ -0,0 +1,13 @@
+[2026-04-14 06:11:27,692][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:11:27,692][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:11:27,692][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:11:31,100][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:11:31,100][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:11:31,101][datasets][INFO] - JAX version 0.4.30 available.
+[2026-04-14 06:11:35,022][root][INFO] - running build_ext
+[2026-04-14 06:11:35,024][root][INFO] - building 'mujoco_py.cymj' extension
+[2026-04-14 06:11:35,024][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -fPIC -O2 -isystem /venv/llm-bc/include -I/venv/llm-bc/include -fPIC -O2 -isystem /venv/llm-bc/include -fPIC -I/venv/llm-bc/lib/python3.9/site-packages/mujoco_py -I/root/.mujoco/mujoco210/include -I/venv/llm-bc/lib/python3.9/site-packages/numpy/core/include -I/venv/llm-bc/include/python3.9 -c /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.c -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.o -fopenmp -w
+[2026-04-14 06:12:16,476][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -fPIC -O2 -isystem /venv/llm-bc/include -I/venv/llm-bc/include -fPIC -O2 -isystem /venv/llm-bc/include -fPIC -I/venv/llm-bc/lib/python3.9/site-packages/mujoco_py -I/root/.mujoco/mujoco210/include -I/venv/llm-bc/lib/python3.9/site-packages/numpy/core/include -I/venv/llm-bc/include/python3.9 -c /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.c -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.o -fopenmp -w
+[2026-04-14 06:12:16,608][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -shared -Wl,-rpath,/venv/llm-bc/lib -Wl,-rpath-link,/venv/llm-bc/lib -L/venv/llm-bc/lib -L/venv/llm-bc/lib -Wl,-rpath,/venv/llm-bc/lib -Wl,-rpath-link,/venv/llm-bc/lib -L/venv/llm-bc/lib /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.o -L/root/.mujoco/mujoco210/bin -Wl,--enable-new-dtags,-rpath,/root/.mujoco/mujoco210/bin -lmujoco210 -lglewosmesa -lOSMesa -lGL -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/lib.linux-x86_64-cpython-39/mujoco_py/cymj.cpython-39-x86_64-linux-gnu.so -fopenmp
+[2026-04-14 06:12:19,274][absl][INFO] - MUJOCO_GL=osmesa, attempting to import specified OpenGL backend.
+[2026-04-14 06:12:19,279][absl][INFO] - MuJoCo library version is: 2.3.7

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/files/config.yaml

@@ -0,0 +1,273 @@
+_target_:
+    value: llmbc.workspace.train_llmbc_lowdim_workspace.TrainLLMBCLowdimWorkspace
+_wandb:
+    value:
+        cli_version: 0.18.6
+        m: []
+        python_version: 3.9.25
+        t:
+            "1":
+                - 1
+                - 2
+                - 3
+                - 5
+                - 11
+                - 12
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 55
+                - 71
+                - 83
+                - 95
+                - 98
+                - 100
+                - 105
+            "2":
+                - 1
+                - 2
+                - 3
+                - 5
+                - 11
+                - 12
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 55
+                - 71
+                - 83
+                - 95
+                - 98
+                - 100
+                - 105
+            "3":
+                - 13
+                - 15
+                - 16
+                - 23
+                - 55
+                - 61
+            "4": 3.9.25
+            "5": 0.18.6
+            "6": 4.47.1
+            "8":
+                - 5
+            "12": 0.18.6
+            "13": linux-x86_64
+action_dim:
+    value: 2
+checkpoint:
+    value:
+        save_last_ckpt: true
+        save_last_snapshot: false
+        topk:
+            format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+            k: 5
+            mode: max
+            monitor_key: test_success_rate
+dataloader:
+    value:
+        batch_size: 16
+        num_workers: 0
+        persistent_workers: false
+        pin_memory: false
+        shuffle: true
+exp_name:
+    value: default
+horizon:
+    value: 1
+llm:
+    value:
+        causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+        checkpoint: data/outputs/2025.10.16/11.53.22_train_llm_lowdim_maze2d-medium-v0/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-maze2d-medium-v0/checkpoint-11850
+        config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+        finetune_mode: orig
+        hydra:
+            job:
+                override_dirname: HuggingFaceTB/SmolLM2-135M-Instruct
+            run:
+                dir: data/outputs/2026.04.14/06.11.26_HuggingFaceTB/SmolLM2-135M-Instruct
+        llm_mode: ete-finetuned
+        lora_config:
+            bias: none
+            lora_alpha: 64
+            lora_dropout: 0.05
+            r: 32
+            task_type: CAUSAL_LM
+        max_length: 100
+        model_name: SmolLM2-135M-Instruct
+        name: HuggingFaceTB/SmolLM2-135M-Instruct
+        prompter:
+            _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+            use_joint_mlp_projector: true
+        use_joint_mlp_projector: true
+        use_quantization: false
+llm_do_sample:
+    value: false
+llm_orig_expert_feedback:
+    value: true
+logging:
+    value:
+        group: null
+        id: null
+        mode: online
+        name: 2026.04.14-06.11.26_train_llmbc_lowdim_maze2d-medium-v0
+        project: maze2d-medium-v0-training-mix
+        resume: true
+        tags:
+            - train_llmbc_lowdim
+            - maze2d-medium-v0
+            - default
+model_name:
+    value: HuggingFaceTB/SmolLM2-135M-Instruct
+multi_run:
+    value:
+        run_dir: data/outputs/2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0
+        wandb_name_base: 2026.04.14-06.11.26_train_llmbc_lowdim_maze2d-medium-v0
+n_action_steps:
+    value: 1
+n_latency_steps:
+    value: 0
+n_obs_steps:
+    value: 1
+name:
+    value: train_llmbc_lowdim
+obs_dim:
+    value: 6
+optimizer:
+    value:
+        _target_: torch.optim.AdamW
+        betas:
+            - 0.95
+            - 0.999
+        eps: 1e-08
+        lr: 0.0001
+        weight_decay: 1e-06
+output_dir:
+    value: /workspace/LLM-BC/data/outputs/2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0
+past_action_visible:
+    value: false
+policy:
+    value:
+        _target_: llmbc.policy.llmbc_lowdim_policy.LLMBCLowdimPolicy
+        action_dim: 2
+        horizon: 1
+        llm_discriminator:
+            _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+            llm_translator:
+                _target_: llmbc.translator.llm_translator.LLMTranslator
+                action_dim: 2
+                cfg:
+                    causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+                    checkpoint: data/outputs/2025.10.16/11.53.22_train_llm_lowdim_maze2d-medium-v0/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-maze2d-medium-v0/checkpoint-11850
+                    config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+                    finetune_mode: orig
+                    hydra:
+                        job:
+                            override_dirname: HuggingFaceTB/SmolLM2-135M-Instruct
+                        run:
+                            dir: data/outputs/2026.04.14/06.11.26_HuggingFaceTB/SmolLM2-135M-Instruct
+                    llm_mode: ete-finetuned
+                    lora_config:
+                        bias: none
+                        lora_alpha: 64
+                        lora_dropout: 0.05
+                        r: 32
+                        task_type: CAUSAL_LM
+                    max_length: 100
+                    model_name: SmolLM2-135M-Instruct
+                    name: HuggingFaceTB/SmolLM2-135M-Instruct
+                    prompter:
+                        _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+                        use_joint_mlp_projector: true
+                    use_joint_mlp_projector: true
+                    use_quantization: false
+                horizon: 1
+                n_action_steps: 1
+                n_obs_steps: 1
+                obs_dim: 6
+            task_id: maze2d-medium-v0
+        loss_bc_weight: 1
+        loss_llm_weight: 0.01
+        model:
+            _target_: llmbc.model.policy.policy_mlp.PolicyMLP
+            activation: tanh
+            hidden_size:
+                - 256
+                - 256
+                - 256
+                - 256
+            input_size: 6
+            n_action_steps: 1
+            n_obs_steps: 1
+            output_size: 2
+        n_action_steps: 1
+        n_obs_steps: 1
+        normalize_llm_loss: true
+        obs_dim: 6
+task:
+    value:
+        action_dim: 2
+        dataset:
+            _target_: llmbc.dataset.pointmaze_lowdim_dataset.PointmazeLowdimDataset
+            data_path: datasets/maze2d-medium-v0-mix(csp5).pt
+            data_path2: datasets/maze2d-medium-v0.pt
+            dummy_normalizer: true
+            horizon: 1
+            obs_eef_target: true
+            pad_after: 0
+            pad_before: 0
+            use_manual_normalizer: false
+            val_ratio: 0.02
+        env_runner:
+            _target_: llmbc.env_runner.pointmaze_lowdim_runner.PointmazeLowdimRunner
+            discount: 0.99
+            env_name: llf-pointmaze-maze2d-medium-v0
+            feedback_type:
+                - hp
+                - hn
+                - fp
+            instruction_type: b
+            max_steps: 250
+            n_action_steps: 1
+            n_envs: 10
+            n_obs_steps: 1
+            n_test: 50
+            n_train: 10
+            visual: false
+        instructor:
+            _target_: llmbc.translator.instructor.pointmaze_instructor.maze2d_medium_v0_instructor.Maze2dMediumV0Instructor
+        name: maze2d-medium-v0
+        obs_dim: 6
+task_name:
+    value: maze2d-medium-v0
+training:
+    value:
+        checkpoint_every: 4
+        debug: false
+        device: cuda:0
+        grad_norm_clip: 0.5
+        gradient_accumulate_every: 8
+        lr_scheduler: cosine
+        lr_warmup_steps: 10
+        max_train_steps: null
+        max_val_steps: null
+        num_epochs: 801
+        resume: false
+        rollout_every: 4
+        sample_every: 4
+        sample_max_batch: 128
+        seed: 44
+        tqdm_interval_sec: 1
+        val_every: 1
+val_dataloader:
+    value:
+        batch_size: 16
+        num_workers: 0
+        persistent_workers: false
+        pin_memory: false
+        shuffle: true

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/files/output.log

@@ -0,0 +1,5 @@
+/venv/llm-bc/lib/python3.9/site-packages/gym/envs/registration.py:505: UserWarning: WARN: The environment maze2d-medium-v0 is out of date. You should consider upgrading to version `v1` with the environment ID `maze2d-medium-v1`.
+  logger.warn(
+Eval PointmazeLowdimRunner 1/6:   0%|                                                                                                                                             | 0/250 [00:00<?, ?it/s]/workspace/LLM-BC/llmbc/common/llfbench_util.py:39: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:275.)
+  obs = torch.tensor(obs, dtype=torch.float32).unsqueeze(dim=0).to(device)
+                                                                                                                                                                                                          

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/files/wandb-metadata.json

@@ -0,0 +1,56 @@
+{
+  "os": "Linux-5.15.0-168-generic-x86_64-with-glibc2.39",
+  "python": "3.9.25",
+  "startedAt": "2026-04-14T06:12:20.200003Z",
+  "args": [
+    "--config-path",
+    "config/mix_data",
+    "--config-name",
+    "llmbc_maze2d-medium-v0_mix(csp5).yaml",
+    "training.seed=44"
+  ],
+  "program": "/workspace/LLM-BC/./train.py",
+  "codePath": "train.py",
+  "git": {
+    "remote": "https://github.com/CHYang25/LLM-BC.git",
+    "commit": "2cd7c18ceb76b043d0dd650c60c283d7b7721bb3"
+  },
+  "email": "chris920325@gmail.com",
+  "root": "/workspace/LLM-BC/data/outputs/2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0",
+  "host": "c8ef52649900",
+  "username": "root",
+  "executable": "/venv/llm-bc/bin/python3",
+  "codePathLocal": "train.py",
+  "cpu_count": 72,
+  "cpu_count_logical": 144,
+  "gpu": "NVIDIA GeForce RTX 4090",
+  "gpu_count": 2,
+  "disk": {
+    "/": {
+      "total": "545460846592",
+      "used": "352964378624"
+    }
+  },
+  "memory": {
+    "total": "810942898176"
+  },
+  "cpu": {
+    "count": 72,
+    "countLogical": 144
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA GeForce RTX 4090",
+      "memoryTotal": "25757220864",
+      "cudaCores": 16384,
+      "architecture": "Ada"
+    },
+    {
+      "name": "NVIDIA GeForce RTX 4090",
+      "memoryTotal": "25757220864",
+      "cudaCores": 16384,
+      "architecture": "Ada"
+    }
+  ],
+  "cudaVersion": "12.8"
+}

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/files/wandb-summary.json

@@ -0,0 +1 @@
+{"lr":5.629757771075106e-15,"train_loss_llm":0.0035964250564575195,"train/success_rate":0.8,"global_step":1674890,"_step":1674890,"train_loss":0.02662586953478947,"train/cumulative_reward":6.172425918440405,"grad_norm":0.23806339502334595,"test/cumulative_reward":6.285320306195163,"train/mean_score":0.9,"test/mean_score":0.68,"train_loss_bc":0.023785388097167015,"_timestamp":1.7764978261614358e+09,"val_loss":0.18932147324085236,"train_action_mse_error":0.0071886624209582806,"_runtime":351485.9283665,"epoch":800,"_wandb":{"runtime":351485},"test/success_rate":0.52}

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/run-20260414_061220-6okd5etw/run-6okd5etw.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1210535b6b37a158f80541679c961c79af116acda9f6b05f7433bdfcaaa431ea
+size 1496331597

2026.04.14/06.11.26_train_llmbc_lowdim_maze2d-medium-v0/wandb/wandb-resume.json

@@ -0,0 +1 @@
+{"run_id": "6okd5etw"}

2026.04.14/06.20.13_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/config.yaml

@@ -0,0 +1,193 @@
+name: train_llm_diffusion_unet_lowdim
+_target_: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+keypoint_visible_rate: 1.0
+obs_as_local_cond: false
+obs_as_global_cond: true
+pred_action_steps_only: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+  model:
+    _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+    input_dim: '${eval: ${task.action_dim} if ${obs_as_local_cond} or ${obs_as_global_cond}
+      else ${task.obs_dim} + ${task.action_dim}}'
+    local_cond_dim: '${eval: ${task.obs_dim} if ${obs_as_local_cond} else None}'
+    global_cond_dim: '${eval: ${task.obs_dim}*${n_obs_steps} if ${obs_as_global_cond}
+      else None}'
+    diffusion_step_embed_dim: 256
+    down_dims:
+    - 256
+    - 512
+    - 1024
+    kernel_size: 5
+    n_groups: 8
+    cond_predict_scale: true
+  noise_scheduler:
+    _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+    num_train_timesteps: 100
+    beta_start: 0.0001
+    beta_end: 0.02
+    beta_schedule: squaredcos_cap_v2
+    variance_type: fixed_small
+    clip_sample: true
+    prediction_type: epsilon
+  horizon: ${horizon}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  n_obs_steps: ${n_obs_steps}
+  num_inference_steps: 100
+  obs_as_local_cond: ${obs_as_local_cond}
+  obs_as_global_cond: ${obs_as_global_cond}
+  pred_action_steps_only: ${pred_action_steps_only}
+  oa_step_convention: true
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${obs_dim}
+      action_dim: ${action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  loss_dp_weight: 1.0
+  loss_llm_weight: 0.01
+  normalize_llm_loss: true
+  reweight_llm_loss: true
+ema:
+  _target_: llmbc.model.diffusion.ema_model.EMAModel
+  update_after_step: 0
+  inv_gamma: 1.0
+  power: 0.75
+  min_value: 0.0
+  max_value: 0.9999
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: false
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 42
+  debug: false
+  resume: true
+  lr_scheduler: cosine
+  lr_warmup_steps: 500
+  num_epochs: 1001
+  gradient_accumulate_every: 16
+  use_ema: true
+  rollout_every: 5
+  checkpoint_every: 5
+  val_every: 1
+  sample_every: 5
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: sweep-v2
+  obs_dim: 9
+  action_dim: 4
+  env_runner:
+    _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+    env_name: llf-metaworld-sweep-v2
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 30
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.9
+  dataset:
+    _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+    data_path: datasets/sweep-v2-mix(csp5).pt
+    data_path2: datasets/sweep-v2.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.1
+    dummy_normalizer: false
+  instructor:
+    _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.20.13_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - policy.loss_llm_weight=1.0e-2
+  job:
+    name: train
+    chdir: null
+    override_dirname: policy.loss_llm_weight=1.0e-2
+    id: ???
+    num: ???
+    config_name: llmdp_sweep-v2_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.20.13_train_llm_diffusion_unet_lowdim_sweep-v2
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.20.13_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- policy.loss_llm_weight=1.0e-2

2026.04.14/06.20.13_train_llm_diffusion_unet_lowdim_sweep-v2/train.log

@@ -0,0 +1,7 @@
+[2026-04-14 06:20:14,990][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:20:14,990][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:20:14,990][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:20:17,187][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:20:17,187][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:20:17,188][datasets][INFO] - JAX version 0.4.30 available.
+[2026-04-14 06:20:18,923][llmbc.model.diffusion.conditional_unet1d][INFO] - number of parameters: 6.501479e+07

2026.04.14/06.20.23_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/config.yaml

@@ -0,0 +1,193 @@
+name: train_llm_diffusion_unet_lowdim
+_target_: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+keypoint_visible_rate: 1.0
+obs_as_local_cond: false
+obs_as_global_cond: true
+pred_action_steps_only: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+  model:
+    _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+    input_dim: '${eval: ${task.action_dim} if ${obs_as_local_cond} or ${obs_as_global_cond}
+      else ${task.obs_dim} + ${task.action_dim}}'
+    local_cond_dim: '${eval: ${task.obs_dim} if ${obs_as_local_cond} else None}'
+    global_cond_dim: '${eval: ${task.obs_dim}*${n_obs_steps} if ${obs_as_global_cond}
+      else None}'
+    diffusion_step_embed_dim: 256
+    down_dims:
+    - 256
+    - 512
+    - 1024
+    kernel_size: 5
+    n_groups: 8
+    cond_predict_scale: true
+  noise_scheduler:
+    _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+    num_train_timesteps: 100
+    beta_start: 0.0001
+    beta_end: 0.02
+    beta_schedule: squaredcos_cap_v2
+    variance_type: fixed_small
+    clip_sample: true
+    prediction_type: epsilon
+  horizon: ${horizon}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  n_obs_steps: ${n_obs_steps}
+  num_inference_steps: 100
+  obs_as_local_cond: ${obs_as_local_cond}
+  obs_as_global_cond: ${obs_as_global_cond}
+  pred_action_steps_only: ${pred_action_steps_only}
+  oa_step_convention: true
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${obs_dim}
+      action_dim: ${action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  loss_dp_weight: 1.0
+  loss_llm_weight: 0.001
+  normalize_llm_loss: true
+  reweight_llm_loss: true
+ema:
+  _target_: llmbc.model.diffusion.ema_model.EMAModel
+  update_after_step: 0
+  inv_gamma: 1.0
+  power: 0.75
+  min_value: 0.0
+  max_value: 0.9999
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: false
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 42
+  debug: false
+  resume: true
+  lr_scheduler: cosine
+  lr_warmup_steps: 500
+  num_epochs: 1001
+  gradient_accumulate_every: 16
+  use_ema: true
+  rollout_every: 5
+  checkpoint_every: 5
+  val_every: 1
+  sample_every: 5
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: sweep-v2
+  obs_dim: 9
+  action_dim: 4
+  env_runner:
+    _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+    env_name: llf-metaworld-sweep-v2
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 30
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.9
+  dataset:
+    _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+    data_path: datasets/sweep-v2-mix(csp5).pt
+    data_path2: datasets/sweep-v2.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.1
+    dummy_normalizer: false
+  instructor:
+    _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.20.23_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - policy.loss_llm_weight=1.0e-3
+  job:
+    name: train
+    chdir: null
+    override_dirname: policy.loss_llm_weight=1.0e-3
+    id: ???
+    num: ???
+    config_name: llmdp_sweep-v2_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.20.23_train_llm_diffusion_unet_lowdim_sweep-v2
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.20.23_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- policy.loss_llm_weight=1.0e-3

2026.04.14/06.20.23_train_llm_diffusion_unet_lowdim_sweep-v2/train.log

@@ -0,0 +1,7 @@
+[2026-04-14 06:20:25,234][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:20:25,235][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:20:25,235][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:20:27,490][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:20:27,491][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:20:27,491][datasets][INFO] - JAX version 0.4.30 available.
+[2026-04-14 06:20:29,155][llmbc.model.diffusion.conditional_unet1d][INFO] - number of parameters: 6.501479e+07

2026.04.14/06.20.32_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/config.yaml

@@ -0,0 +1,193 @@
+name: train_llm_diffusion_unet_lowdim
+_target_: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+keypoint_visible_rate: 1.0
+obs_as_local_cond: false
+obs_as_global_cond: true
+pred_action_steps_only: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+  model:
+    _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+    input_dim: '${eval: ${task.action_dim} if ${obs_as_local_cond} or ${obs_as_global_cond}
+      else ${task.obs_dim} + ${task.action_dim}}'
+    local_cond_dim: '${eval: ${task.obs_dim} if ${obs_as_local_cond} else None}'
+    global_cond_dim: '${eval: ${task.obs_dim}*${n_obs_steps} if ${obs_as_global_cond}
+      else None}'
+    diffusion_step_embed_dim: 256
+    down_dims:
+    - 256
+    - 512
+    - 1024
+    kernel_size: 5
+    n_groups: 8
+    cond_predict_scale: true
+  noise_scheduler:
+    _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+    num_train_timesteps: 100
+    beta_start: 0.0001
+    beta_end: 0.02
+    beta_schedule: squaredcos_cap_v2
+    variance_type: fixed_small
+    clip_sample: true
+    prediction_type: epsilon
+  horizon: ${horizon}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  n_obs_steps: ${n_obs_steps}
+  num_inference_steps: 100
+  obs_as_local_cond: ${obs_as_local_cond}
+  obs_as_global_cond: ${obs_as_global_cond}
+  pred_action_steps_only: ${pred_action_steps_only}
+  oa_step_convention: true
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${obs_dim}
+      action_dim: ${action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  loss_dp_weight: 1.0
+  loss_llm_weight: 0.1
+  normalize_llm_loss: true
+  reweight_llm_loss: true
+ema:
+  _target_: llmbc.model.diffusion.ema_model.EMAModel
+  update_after_step: 0
+  inv_gamma: 1.0
+  power: 0.75
+  min_value: 0.0
+  max_value: 0.9999
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: false
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 42
+  debug: false
+  resume: true
+  lr_scheduler: cosine
+  lr_warmup_steps: 500
+  num_epochs: 1001
+  gradient_accumulate_every: 16
+  use_ema: true
+  rollout_every: 5
+  checkpoint_every: 5
+  val_every: 1
+  sample_every: 5
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: sweep-v2
+  obs_dim: 9
+  action_dim: 4
+  env_runner:
+    _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+    env_name: llf-metaworld-sweep-v2
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 30
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.9
+  dataset:
+    _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+    data_path: datasets/sweep-v2-mix(csp5).pt
+    data_path2: datasets/sweep-v2.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.1
+    dummy_normalizer: false
+  instructor:
+    _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.20.32_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - policy.loss_llm_weight=1.0e-1
+  job:
+    name: train
+    chdir: null
+    override_dirname: policy.loss_llm_weight=1.0e-1
+    id: ???
+    num: ???
+    config_name: llmdp_sweep-v2_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.20.32_train_llm_diffusion_unet_lowdim_sweep-v2
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.20.32_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- policy.loss_llm_weight=1.0e-1

2026.04.14/06.20.32_train_llm_diffusion_unet_lowdim_sweep-v2/train.log

@@ -0,0 +1,7 @@
+[2026-04-14 06:20:34,211][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:20:34,211][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:20:34,211][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:20:36,462][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:20:36,463][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:20:36,463][datasets][INFO] - JAX version 0.4.30 available.
+[2026-04-14 06:20:38,106][llmbc.model.diffusion.conditional_unet1d][INFO] - number of parameters: 6.501479e+07

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/config.yaml

@@ -0,0 +1,193 @@
+name: train_llm_diffusion_unet_lowdim
+_target_: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+keypoint_visible_rate: 1.0
+obs_as_local_cond: false
+obs_as_global_cond: true
+pred_action_steps_only: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+  model:
+    _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+    input_dim: '${eval: ${task.action_dim} if ${obs_as_local_cond} or ${obs_as_global_cond}
+      else ${task.obs_dim} + ${task.action_dim}}'
+    local_cond_dim: '${eval: ${task.obs_dim} if ${obs_as_local_cond} else None}'
+    global_cond_dim: '${eval: ${task.obs_dim}*${n_obs_steps} if ${obs_as_global_cond}
+      else None}'
+    diffusion_step_embed_dim: 256
+    down_dims:
+    - 256
+    - 512
+    - 1024
+    kernel_size: 5
+    n_groups: 8
+    cond_predict_scale: true
+  noise_scheduler:
+    _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+    num_train_timesteps: 100
+    beta_start: 0.0001
+    beta_end: 0.02
+    beta_schedule: squaredcos_cap_v2
+    variance_type: fixed_small
+    clip_sample: true
+    prediction_type: epsilon
+  horizon: ${horizon}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  n_obs_steps: ${n_obs_steps}
+  num_inference_steps: 100
+  obs_as_local_cond: ${obs_as_local_cond}
+  obs_as_global_cond: ${obs_as_global_cond}
+  pred_action_steps_only: ${pred_action_steps_only}
+  oa_step_convention: true
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${obs_dim}
+      action_dim: ${action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  loss_dp_weight: 1.0
+  loss_llm_weight: 0.1
+  normalize_llm_loss: true
+  reweight_llm_loss: true
+ema:
+  _target_: llmbc.model.diffusion.ema_model.EMAModel
+  update_after_step: 0
+  inv_gamma: 1.0
+  power: 0.75
+  min_value: 0.0
+  max_value: 0.9999
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: false
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 42
+  debug: false
+  resume: true
+  lr_scheduler: cosine
+  lr_warmup_steps: 500
+  num_epochs: 1001
+  gradient_accumulate_every: 16
+  use_ema: true
+  rollout_every: 5
+  checkpoint_every: 5
+  val_every: 1
+  sample_every: 5
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: sweep-v2
+  obs_dim: 9
+  action_dim: 4
+  env_runner:
+    _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+    env_name: llf-metaworld-sweep-v2
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 30
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.9
+  dataset:
+    _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+    data_path: datasets/sweep-v2-mix(csp5).pt
+    data_path2: datasets/sweep-v2.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.1
+    dummy_normalizer: false
+  instructor:
+    _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - policy.loss_llm_weight=1.0e-1
+  job:
+    name: train
+    chdir: null
+    override_dirname: policy.loss_llm_weight=1.0e-1
+    id: ???
+    num: ???
+    config_name: llmdp_sweep-v2_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- policy.loss_llm_weight=1.0e-1

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/logs.json.txt

@@ -0,0 +1,551 @@
+{"train_loss": 1.2434064149856567, "train_loss_dp": 1.2109837532043457, "train_loss_llm": 0.32422637939453125, "global_step": 0, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.8936537504196167, "train_loss_dp": 1.5555988550186157, "train_loss_llm": 3.3805484771728516, "global_step": 1, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 0.9923631548881531, "train_loss_dp": 0.8444761037826538, "train_loss_llm": 1.4788703918457031, "global_step": 2, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.6303431987762451, "train_loss_dp": 1.567467212677002, "train_loss_llm": 0.6287602186203003, "global_step": 3, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2874763011932373, "train_loss_dp": 1.2161624431610107, "train_loss_llm": 0.7131385803222656, "global_step": 4, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.6784204244613647, "train_loss_dp": 1.6041865348815918, "train_loss_llm": 0.7423383593559265, "global_step": 5, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.5025584697723389, "train_loss_dp": 1.3234602212905884, "train_loss_llm": 1.7909822463989258, "global_step": 6, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.1995466947555542, "train_loss_dp": 1.1930253505706787, "train_loss_llm": 0.06521368026733398, "global_step": 7, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2473700046539307, "train_loss_dp": 1.006334900856018, "train_loss_llm": 2.410350799560547, "global_step": 8, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.481979489326477, "train_loss_dp": 1.3523192405700684, "train_loss_llm": 1.2966029644012451, "global_step": 9, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.46666419506073, "train_loss_dp": 1.3722383975982666, "train_loss_llm": 0.9442574977874756, "global_step": 10, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.3386297225952148, "train_loss_dp": 1.2402263879776, "train_loss_llm": 0.9840337038040161, "global_step": 11, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2905042171478271, "train_loss_dp": 1.2112371921539307, "train_loss_llm": 0.792670726776123, "global_step": 12, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.1549105644226074, "train_loss_dp": 1.1232653856277466, "train_loss_llm": 0.3164515495300293, "global_step": 13, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2144033908843994, "train_loss_dp": 1.1038504838943481, "train_loss_llm": 1.1055290699005127, "global_step": 14, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.7257826328277588, "train_loss_dp": 1.5052738189697266, "train_loss_llm": 2.205087661743164, "global_step": 15, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 0.9575361013412476, "train_loss_dp": 0.8685603141784668, "train_loss_llm": 0.8897576332092285, "global_step": 16, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.285231590270996, "train_loss_dp": 1.2376809120178223, "train_loss_llm": 0.4755067825317383, "global_step": 17, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.098201036453247, "train_loss_dp": 0.9839760661125183, "train_loss_llm": 1.1422491073608398, "global_step": 18, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.3243927955627441, "train_loss_dp": 1.252960205078125, "train_loss_llm": 0.7143263816833496, "global_step": 19, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.1879470348358154, "train_loss_dp": 1.1879470348358154, "train_loss_llm": 0.0, "global_step": 20, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.4405707120895386, "train_loss_dp": 1.3631517887115479, "train_loss_llm": 0.7741892337799072, "global_step": 21, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.2117116451263428, "train_loss_dp": 1.1086549758911133, "train_loss_llm": 1.0305662155151367, "global_step": 22, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.237113118171692, "train_loss_dp": 1.1755462884902954, "train_loss_llm": 0.6156682968139648, "global_step": 23, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.5508129596710205, "train_loss_dp": 1.3083237409591675, "train_loss_llm": 2.4248924255371094, "global_step": 24, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.516460657119751, "train_loss_dp": 1.387798547744751, "train_loss_llm": 1.2866206169128418, "global_step": 25, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.358413577079773, "train_loss_dp": 1.2061840295791626, "train_loss_llm": 1.5222949981689453, "global_step": 26, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 0.896275520324707, "train_loss_dp": 0.8227578997612, "train_loss_llm": 0.7351760864257812, "global_step": 27, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 0.9893575310707092, "train_loss_dp": 0.906792402267456, "train_loss_llm": 0.8256511688232422, "global_step": 28, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 2.408330202102661, "train_loss_dp": 1.2448729276657104, "train_loss_llm": 11.634572982788086, "global_step": 29, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.3611738681793213, "train_loss_dp": 1.1810848712921143, "train_loss_llm": 1.8008904457092285, "global_step": 30, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.4871776103973389, "train_loss_dp": 1.383911371231079, "train_loss_llm": 1.0326623916625977, "global_step": 31, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 3.552180051803589, "train_loss_dp": 0.9743914604187012, "train_loss_llm": 25.77788543701172, "global_step": 32, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.1520498991012573, "train_loss_dp": 1.0061284303665161, "train_loss_llm": 1.4592151641845703, "global_step": 33, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 2.3061108589172363, "train_loss_dp": 1.1040608882904053, "train_loss_llm": 12.020498275756836, "global_step": 34, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0228338241577148, "train_loss_dp": 0.751134991645813, "train_loss_llm": 2.7169880867004395, "global_step": 35, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3713548183441162, "train_loss_dp": 1.2942302227020264, "train_loss_llm": 0.7712459564208984, "global_step": 36, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.1763675212860107, "train_loss_dp": 1.037414312362671, "train_loss_llm": 1.3895319700241089, "global_step": 37, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3600103855133057, "train_loss_dp": 1.2952556610107422, "train_loss_llm": 0.6475467681884766, "global_step": 38, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.6156158447265625, "train_loss_dp": 1.4624698162078857, "train_loss_llm": 1.5314598083496094, "global_step": 39, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0735447406768799, "train_loss_dp": 1.0008471012115479, "train_loss_llm": 0.7269763946533203, "global_step": 40, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.2104142904281616, "train_loss_dp": 1.13003408908844, "train_loss_llm": 0.8038020133972168, "global_step": 41, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0883771181106567, "train_loss_dp": 1.028773546218872, "train_loss_llm": 0.5960354804992676, "global_step": 42, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.2805579900741577, "train_loss_dp": 1.1822187900543213, "train_loss_llm": 0.983391523361206, "global_step": 43, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3792062997817993, "train_loss_dp": 1.1784553527832031, "train_loss_llm": 2.007509231567383, "global_step": 44, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.207040548324585, "train_loss_dp": 1.1314449310302734, "train_loss_llm": 0.7559566497802734, "global_step": 45, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.6805046796798706, "train_loss_dp": 1.3093578815460205, "train_loss_llm": 3.711467742919922, "global_step": 46, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3399341106414795, "train_loss_dp": 1.296255350112915, "train_loss_llm": 0.4367879033088684, "global_step": 47, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.9866480827331543, "train_loss_dp": 1.7141953706741333, "train_loss_llm": 2.724527359008789, "global_step": 48, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.995402216911316, "train_loss_dp": 1.5114262104034424, "train_loss_llm": 4.839759826660156, "global_step": 49, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.8541259765625, "train_loss_dp": 1.2823514938354492, "train_loss_llm": 5.717743873596191, "global_step": 50, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.172556757926941, "train_loss_dp": 1.0826826095581055, "train_loss_llm": 0.8987417221069336, "global_step": 51, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.427757740020752, "train_loss_dp": 1.3476608991622925, "train_loss_llm": 0.8009681701660156, "global_step": 52, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.450551986694336, "train_loss_dp": 1.3952624797821045, "train_loss_llm": 0.5528949499130249, "global_step": 53, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.1547715663909912, "train_loss_dp": 1.010388731956482, "train_loss_llm": 1.4438278675079346, "global_step": 54, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.4142717123031616, "train_loss_dp": 1.2390402555465698, "train_loss_llm": 1.7523150444030762, "global_step": 55, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.2300602197647095, "train_loss_dp": 1.2300602197647095, "train_loss_llm": 0.0, "global_step": 56, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.4993572235107422, "train_loss_dp": 1.4281704425811768, "train_loss_llm": 0.7118682861328125, "global_step": 57, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.7570635080337524, "train_loss_dp": 1.7570635080337524, "train_loss_llm": 0.0, "global_step": 58, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 0.9507583379745483, "train_loss_dp": 0.9106940031051636, "train_loss_llm": 0.40064334869384766, "global_step": 59, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.105691909790039, "train_loss_dp": 1.058445692062378, "train_loss_llm": 0.4724617004394531, "global_step": 60, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.3795708417892456, "train_loss_dp": 1.2993495464324951, "train_loss_llm": 0.8022127151489258, "global_step": 61, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.1246166229248047, "train_loss_dp": 1.1098010540008545, "train_loss_llm": 0.14815521240234375, "global_step": 62, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.1495836973190308, "train_loss_dp": 1.0959360599517822, "train_loss_llm": 0.5364761352539062, "global_step": 63, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.144653081893921, "train_loss_dp": 1.0287178754806519, "train_loss_llm": 1.1593523025512695, "global_step": 64, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.437492847442627, "train_loss_dp": 1.2668719291687012, "train_loss_llm": 1.7062091827392578, "global_step": 65, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3588780164718628, "train_loss_dp": 1.2626943588256836, "train_loss_llm": 0.9618363380432129, "global_step": 66, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.2253309488296509, "train_loss_dp": 0.9413154721260071, "train_loss_llm": 2.8401551246643066, "global_step": 67, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.4009106159210205, "train_loss_dp": 1.1807156801223755, "train_loss_llm": 2.2019498348236084, "global_step": 68, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.569897174835205, "train_loss_dp": 1.2876338958740234, "train_loss_llm": 2.8226327896118164, "global_step": 69, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3687400817871094, "train_loss_dp": 1.1849956512451172, "train_loss_llm": 1.83744478225708, "global_step": 70, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3966848850250244, "train_loss_dp": 1.3440794944763184, "train_loss_llm": 0.5260534286499023, "global_step": 71, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.1062660217285156, "train_loss_dp": 0.9184344410896301, "train_loss_llm": 1.8783154487609863, "global_step": 72, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.381195068359375, "train_loss_dp": 1.1352195739746094, "train_loss_llm": 2.4597554206848145, "global_step": 73, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.4001133441925049, "train_loss_dp": 1.3563779592514038, "train_loss_llm": 0.4373539686203003, "global_step": 74, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.9038599729537964, "train_loss_dp": 1.8560649156570435, "train_loss_llm": 0.4779500961303711, "global_step": 75, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.5008933544158936, "train_loss_dp": 1.4142041206359863, "train_loss_llm": 0.8668922185897827, "global_step": 76, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.409606695175171, "train_loss_dp": 1.2525367736816406, "train_loss_llm": 1.5706987380981445, "global_step": 77, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.0766390562057495, "train_loss_dp": 0.9175339937210083, "train_loss_llm": 1.5910508632659912, "global_step": 78, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.0880775451660156, "train_loss_dp": 0.7924573421478271, "train_loss_llm": 2.9562015533447266, "global_step": 79, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.0746593475341797, "train_loss_dp": 1.0096971988677979, "train_loss_llm": 0.6496217250823975, "global_step": 80, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.3561511039733887, "train_loss_dp": 1.2802879810333252, "train_loss_llm": 0.7586314678192139, "global_step": 81, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.7920588254928589, "train_loss_dp": 1.4360041618347168, "train_loss_llm": 3.560546875, "global_step": 82, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 4.151108741760254, "train_loss_dp": 1.5098917484283447, "train_loss_llm": 26.412166595458984, "global_step": 83, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 0.9669823050498962, "train_loss_dp": 0.9343065023422241, "train_loss_llm": 0.3267580270767212, "global_step": 84, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1679143905639648, "train_loss_dp": 1.1137746572494507, "train_loss_llm": 0.5413973331451416, "global_step": 85, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1866012811660767, "train_loss_dp": 1.1073408126831055, "train_loss_llm": 0.7926044464111328, "global_step": 86, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.4346221685409546, "train_loss_dp": 1.3543606996536255, "train_loss_llm": 0.8026142716407776, "global_step": 87, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.3397514820098877, "train_loss_dp": 1.2393453121185303, "train_loss_llm": 1.0040616989135742, "global_step": 88, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1538681983947754, "train_loss_dp": 1.0597505569458008, "train_loss_llm": 0.9411766529083252, "global_step": 89, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.0687814950942993, "train_loss_dp": 0.939609706401825, "train_loss_llm": 1.291717529296875, "global_step": 90, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.090595006942749, "train_loss_dp": 0.9575150609016418, "train_loss_llm": 1.3307991027832031, "global_step": 91, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.3333793878555298, "train_loss_dp": 1.251429796218872, "train_loss_llm": 0.8194961547851562, "global_step": 92, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1501396894454956, "train_loss_dp": 1.0854017734527588, "train_loss_llm": 0.6473789215087891, "global_step": 93, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1017917394638062, "train_loss_dp": 1.0115753412246704, "train_loss_llm": 0.9021644592285156, "global_step": 94, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1655091047286987, "train_loss_dp": 1.0675088167190552, "train_loss_llm": 0.9800033569335938, "global_step": 95, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.5182451009750366, "train_loss_dp": 0.9565083384513855, "train_loss_llm": 5.617367744445801, "global_step": 96, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.1374046802520752, "train_loss_dp": 0.9959837198257446, "train_loss_llm": 1.4142093658447266, "global_step": 97, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.1391533613204956, "train_loss_dp": 1.054567575454712, "train_loss_llm": 0.8458576202392578, "global_step": 98, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.659332275390625, "train_loss_dp": 0.8356262445449829, "train_loss_llm": 8.237060546875, "global_step": 99, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.4114913940429688, "train_loss_dp": 1.2123526334762573, "train_loss_llm": 1.9913873672485352, "global_step": 100, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.6131895780563354, "train_loss_dp": 1.5222314596176147, "train_loss_llm": 0.9095816612243652, "global_step": 101, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.434879183769226, "train_loss_dp": 1.1841204166412354, "train_loss_llm": 2.507587432861328, "global_step": 102, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.3973177671432495, "train_loss_dp": 1.2743017673492432, "train_loss_llm": 1.2301597595214844, "global_step": 103, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.5027098655700684, "train_loss_dp": 1.3608131408691406, "train_loss_llm": 1.4189672470092773, "global_step": 104, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.0405128002166748, "train_loss_dp": 0.9271371364593506, "train_loss_llm": 1.1337562799453735, "global_step": 105, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.363061785697937, "train_loss_dp": 1.3135493993759155, "train_loss_llm": 0.49512362480163574, "global_step": 106, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 0.9404826760292053, "train_loss_dp": 0.9404826760292053, "train_loss_llm": 0.0, "global_step": 107, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.5940266847610474, "train_loss_dp": 1.2755205631256104, "train_loss_llm": 3.185061454772949, "global_step": 108, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.520550012588501, "train_loss_dp": 1.3275827169418335, "train_loss_llm": 1.9296727180480957, "global_step": 109, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.4388744831085205, "train_loss_dp": 1.1415976285934448, "train_loss_llm": 2.9727678298950195, "global_step": 110, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.6015065908432007, "train_loss_dp": 1.5133354663848877, "train_loss_llm": 0.8817110061645508, "global_step": 111, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.473227620124817, "train_loss_dp": 1.3589876890182495, "train_loss_llm": 1.142399549484253, "global_step": 112, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.1671924591064453, "train_loss_dp": 0.9679977297782898, "train_loss_llm": 1.9919469356536865, "global_step": 113, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.4365715980529785, "train_loss_dp": 1.2202973365783691, "train_loss_llm": 2.162743091583252, "global_step": 114, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.0032612085342407, "train_loss_dp": 0.8709124326705933, "train_loss_llm": 1.3234872817993164, "global_step": 115, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.392439365386963, "train_loss_dp": 1.2020258903503418, "train_loss_llm": 1.9041342735290527, "global_step": 116, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.329653024673462, "train_loss_dp": 1.133786916732788, "train_loss_llm": 1.9586610794067383, "global_step": 117, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.1006793975830078, "train_loss_dp": 0.7749126553535461, "train_loss_llm": 3.2576675415039062, "global_step": 118, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.9970632791519165, "train_loss_dp": 0.9138763546943665, "train_loss_llm": 0.8318691253662109, "global_step": 119, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.554226040840149, "train_loss_dp": 1.3846684694290161, "train_loss_llm": 1.6955759525299072, "global_step": 120, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.580580711364746, "train_loss_dp": 1.5355420112609863, "train_loss_llm": 0.4503875970840454, "global_step": 121, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.4976978302001953, "train_loss_dp": 1.3972585201263428, "train_loss_llm": 1.0043926239013672, "global_step": 122, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.3272496461868286, "train_loss_dp": 1.2370648384094238, "train_loss_llm": 0.9018478393554688, "global_step": 123, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.3155931234359741, "train_loss_dp": 1.1875865459442139, "train_loss_llm": 1.2800655364990234, "global_step": 124, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.4723001718521118, "train_loss_dp": 1.1729824542999268, "train_loss_llm": 2.9931771755218506, "global_step": 125, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.452314853668213, "train_loss_dp": 1.3155347108840942, "train_loss_llm": 1.3678016662597656, "global_step": 126, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.9460318088531494, "train_loss_dp": 0.8508630990982056, "train_loss_llm": 0.9516868591308594, "global_step": 127, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.1802140474319458, "train_loss_dp": 0.9800631999969482, "train_loss_llm": 2.0015087127685547, "global_step": 128, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.4006552696228027, "train_loss_dp": 1.0460351705551147, "train_loss_llm": 3.5462002754211426, "global_step": 129, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.3679991960525513, "train_loss_dp": 1.2565126419067383, "train_loss_llm": 1.1148653030395508, "global_step": 130, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2788541316986084, "train_loss_dp": 1.211713194847107, "train_loss_llm": 0.6714087724685669, "global_step": 131, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1293612718582153, "train_loss_dp": 0.9124231338500977, "train_loss_llm": 2.169381856918335, "global_step": 132, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.6334517002105713, "train_loss_dp": 1.480604887008667, "train_loss_llm": 1.5284686088562012, "global_step": 133, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.222730040550232, "train_loss_dp": 1.11095130443573, "train_loss_llm": 1.1177878379821777, "global_step": 134, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.7304307818412781, "train_loss_dp": 0.7023981213569641, "train_loss_llm": 0.28032684326171875, "global_step": 135, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.0083627700805664, "train_loss_dp": 0.7776620388031006, "train_loss_llm": 2.3070068359375, "global_step": 136, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.240080714225769, "train_loss_dp": 1.2011940479278564, "train_loss_llm": 0.3888661861419678, "global_step": 137, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.5953550338745117, "train_loss_dp": 1.5124784708023071, "train_loss_llm": 0.8287653923034668, "global_step": 138, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.4363117218017578, "train_loss_dp": 1.165313482284546, "train_loss_llm": 2.709981918334961, "global_step": 139, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.9821711778640747, "train_loss_dp": 0.938874363899231, "train_loss_llm": 0.4329681396484375, "global_step": 140, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.3438633680343628, "train_loss_dp": 1.1623194217681885, "train_loss_llm": 1.8154397010803223, "global_step": 141, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2354316711425781, "train_loss_dp": 1.1815049648284912, "train_loss_llm": 0.5392674207687378, "global_step": 142, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2348833084106445, "train_loss_dp": 1.1790589094161987, "train_loss_llm": 0.5582441091537476, "global_step": 143, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.9522626399993896, "train_loss_dp": 0.7508912682533264, "train_loss_llm": 2.013713836669922, "global_step": 144, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.2936173677444458, "train_loss_dp": 1.0319089889526367, "train_loss_llm": 2.6170835494995117, "global_step": 145, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0112231969833374, "train_loss_dp": 0.8222692012786865, "train_loss_llm": 1.8895397186279297, "global_step": 146, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.18540620803833, "train_loss_dp": 1.1453237533569336, "train_loss_llm": 0.40082502365112305, "global_step": 147, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0036966800689697, "train_loss_dp": 0.9042816758155823, "train_loss_llm": 0.9941501617431641, "global_step": 148, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.2287815809249878, "train_loss_dp": 1.101346731185913, "train_loss_llm": 1.274348258972168, "global_step": 149, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0815613269805908, "train_loss_dp": 0.9836338758468628, "train_loss_llm": 0.9792749881744385, "global_step": 150, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0274279117584229, "train_loss_dp": 0.775547981262207, "train_loss_llm": 2.518798828125, "global_step": 151, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.18129301071167, "train_loss_dp": 1.0132602453231812, "train_loss_llm": 1.6803274154663086, "global_step": 152, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.5826270580291748, "train_loss_dp": 1.457601547241211, "train_loss_llm": 1.2502553462982178, "global_step": 153, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.2897608280181885, "train_loss_dp": 1.105684757232666, "train_loss_llm": 1.8407602310180664, "global_step": 154, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.5086591243743896, "train_loss_dp": 1.002388834953308, "train_loss_llm": 5.062702178955078, "global_step": 155, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.9649895429611206, "train_loss_dp": 0.7761242389678955, "train_loss_llm": 1.88865327835083, "global_step": 156, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0828512907028198, "train_loss_dp": 1.0353147983551025, "train_loss_llm": 0.47536468505859375, "global_step": 157, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.1509294509887695, "train_loss_dp": 1.0870128870010376, "train_loss_llm": 0.6391658782958984, "global_step": 158, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.4991936683654785, "train_loss_dp": 1.1913025379180908, "train_loss_llm": 3.0789108276367188, "global_step": 159, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.1939505338668823, "train_loss_dp": 1.0984699726104736, "train_loss_llm": 0.9548053741455078, "global_step": 160, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.095271348953247, "train_loss_dp": 1.0234202146530151, "train_loss_llm": 0.7185115814208984, "global_step": 161, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.9826719164848328, "train_loss_dp": 0.8854222297668457, "train_loss_llm": 0.9724969863891602, "global_step": 162, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.182403326034546, "train_loss_dp": 1.0527946949005127, "train_loss_llm": 1.296086311340332, "global_step": 163, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.2708439826965332, "train_loss_dp": 1.1911407709121704, "train_loss_llm": 0.797032356262207, "global_step": 164, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.157487392425537, "train_loss_dp": 1.0302231311798096, "train_loss_llm": 1.2726421356201172, "global_step": 165, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.4423418045043945, "train_loss_dp": 1.2782018184661865, "train_loss_llm": 1.6414000988006592, "global_step": 166, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.5062479972839355, "train_loss_dp": 1.4517157077789307, "train_loss_llm": 0.5453231334686279, "global_step": 167, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.9466218948364258, "train_loss_dp": 0.8241947293281555, "train_loss_llm": 1.2242717742919922, "global_step": 168, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.6213325262069702, "train_loss_dp": 1.247808575630188, "train_loss_llm": 3.735239028930664, "global_step": 169, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.6482139825820923, "train_loss_dp": 0.7956172823905945, "train_loss_llm": 8.52596664428711, "global_step": 170, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.5117058753967285, "train_loss_dp": 1.1216180324554443, "train_loss_llm": 3.9008779525756836, "global_step": 171, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.006504774093628, "train_loss_dp": 0.9671895503997803, "train_loss_llm": 0.39315176010131836, "global_step": 172, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.2210654020309448, "train_loss_dp": 1.14951491355896, "train_loss_llm": 0.7155046463012695, "global_step": 173, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.9445791244506836, "train_loss_dp": 0.9445791244506836, "train_loss_llm": 0.0, "global_step": 174, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.1589996814727783, "train_loss_dp": 1.1120543479919434, "train_loss_llm": 0.46945369243621826, "global_step": 175, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.2688140869140625, "train_loss_dp": 1.0855066776275635, "train_loss_llm": 1.8330740928649902, "global_step": 176, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.1326788663864136, "train_loss_dp": 1.0357706546783447, "train_loss_llm": 0.9690818786621094, "global_step": 177, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.6416430473327637, "train_loss_dp": 1.2003788948059082, "train_loss_llm": 4.412641525268555, "global_step": 178, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.0459011793136597, "train_loss_dp": 0.9415539503097534, "train_loss_llm": 1.0434722900390625, "global_step": 179, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.8615846037864685, "train_loss_dp": 0.7749868631362915, "train_loss_llm": 0.8659772872924805, "global_step": 180, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.037363886833191, "train_loss_dp": 0.9090119004249573, "train_loss_llm": 1.2835197448730469, "global_step": 181, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.326940894126892, "train_loss_dp": 1.120017647743225, "train_loss_llm": 2.0692319869995117, "global_step": 182, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.866308331489563, "train_loss_dp": 0.8216971755027771, "train_loss_llm": 0.4461113214492798, "global_step": 183, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.1386281251907349, "train_loss_dp": 1.031313419342041, "train_loss_llm": 1.0731472969055176, "global_step": 184, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.6899389028549194, "train_loss_dp": 0.9879500865936279, "train_loss_llm": 7.019887924194336, "global_step": 185, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9511170983314514, "train_loss_dp": 0.849124014377594, "train_loss_llm": 1.0199308395385742, "global_step": 186, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.0499370098114014, "train_loss_dp": 0.9790463447570801, "train_loss_llm": 0.7089066505432129, "global_step": 187, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.4412920475006104, "train_loss_dp": 1.3331776857376099, "train_loss_llm": 1.0811434984207153, "global_step": 188, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.000285029411316, "train_loss_dp": 0.9592461585998535, "train_loss_llm": 0.4103889465332031, "global_step": 189, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.321751594543457, "train_loss_dp": 1.0179975032806396, "train_loss_llm": 3.0375404357910156, "global_step": 190, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.3318620920181274, "train_loss_dp": 1.2900850772857666, "train_loss_llm": 0.4177696704864502, "global_step": 191, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9957125782966614, "train_loss_dp": 0.9315172433853149, "train_loss_llm": 0.6419534683227539, "global_step": 192, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.699775218963623, "train_loss_dp": 1.2323768138885498, "train_loss_llm": 4.673984527587891, "global_step": 193, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1915502548217773, "train_loss_dp": 1.1699044704437256, "train_loss_llm": 0.21645808219909668, "global_step": 194, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.5729873180389404, "train_loss_dp": 1.2772154808044434, "train_loss_llm": 2.957718849182129, "global_step": 195, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.298560619354248, "train_loss_dp": 1.1789143085479736, "train_loss_llm": 1.196462631225586, "global_step": 196, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.0745209455490112, "train_loss_dp": 0.952931821346283, "train_loss_llm": 1.2158913612365723, "global_step": 197, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.3893707990646362, "train_loss_dp": 1.153257966041565, "train_loss_llm": 2.3611278533935547, "global_step": 198, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.7951567769050598, "train_loss_dp": 0.7324215173721313, "train_loss_llm": 0.6273527145385742, "global_step": 199, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 3.2968602180480957, "train_loss_dp": 0.9798710346221924, "train_loss_llm": 23.169891357421875, "global_step": 200, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.4091683626174927, "train_loss_dp": 1.0128037929534912, "train_loss_llm": 3.9636459350585938, "global_step": 201, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.6895455121994019, "train_loss_dp": 0.6490859985351562, "train_loss_llm": 0.4045950174331665, "global_step": 202, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 2.3663458824157715, "train_loss_dp": 1.3094894886016846, "train_loss_llm": 10.568565368652344, "global_step": 203, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.8081049919128418, "train_loss_dp": 1.3351609706878662, "train_loss_llm": 4.729440689086914, "global_step": 204, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1236177682876587, "train_loss_dp": 1.1033977270126343, "train_loss_llm": 0.20220088958740234, "global_step": 205, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.032827615737915, "train_loss_dp": 0.9335939884185791, "train_loss_llm": 0.9923365116119385, "global_step": 206, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1719549894332886, "train_loss_dp": 1.1318522691726685, "train_loss_llm": 0.40102696418762207, "global_step": 207, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.127489447593689, "train_loss_dp": 1.0305129289627075, "train_loss_llm": 0.9697656631469727, "global_step": 208, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8412598371505737, "train_loss_dp": 0.8412598371505737, "train_loss_llm": 0.0, "global_step": 209, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.9060013890266418, "train_loss_dp": 0.8674270510673523, "train_loss_llm": 0.3857431411743164, "global_step": 210, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.362257480621338, "train_loss_dp": 1.298191785812378, "train_loss_llm": 0.6406564712524414, "global_step": 211, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.2894227504730225, "train_loss_dp": 1.207939624786377, "train_loss_llm": 0.8148307800292969, "global_step": 212, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.036242961883545, "train_loss_dp": 0.995526134967804, "train_loss_llm": 0.407167911529541, "global_step": 213, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.3667150735855103, "train_loss_dp": 1.25553560256958, "train_loss_llm": 1.1117947101593018, "global_step": 214, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.9492236375808716, "train_loss_dp": 0.8719707727432251, "train_loss_llm": 0.7725286483764648, "global_step": 215, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.1446915864944458, "train_loss_dp": 1.1338013410568237, "train_loss_llm": 0.1089019775390625, "global_step": 216, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.959013819694519, "train_loss_dp": 0.8852850794792175, "train_loss_llm": 0.7372875213623047, "global_step": 217, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8863383531570435, "train_loss_dp": 0.7524318695068359, "train_loss_llm": 1.3390650749206543, "global_step": 218, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8720508217811584, "train_loss_dp": 0.8204406499862671, "train_loss_llm": 0.5161018371582031, "global_step": 219, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.0528730154037476, "train_loss_dp": 0.9786262512207031, "train_loss_llm": 0.7424678802490234, "global_step": 220, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.4907286167144775, "train_loss_dp": 1.3778331279754639, "train_loss_llm": 1.128955364227295, "global_step": 221, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.0284879207611084, "train_loss_dp": 0.9562445878982544, "train_loss_llm": 0.7224330902099609, "global_step": 222, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.1713515520095825, "train_loss_dp": 1.1165175437927246, "train_loss_llm": 0.5483397245407104, "global_step": 223, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.6791698932647705, "train_loss_dp": 1.2217121124267578, "train_loss_llm": 4.574577331542969, "global_step": 224, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.1774629354476929, "train_loss_dp": 1.1331934928894043, "train_loss_llm": 0.44269490242004395, "global_step": 225, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.3234145641326904, "train_loss_dp": 1.1474077701568604, "train_loss_llm": 1.760068416595459, "global_step": 226, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.9101505875587463, "train_loss_dp": 0.7784578800201416, "train_loss_llm": 1.3169269561767578, "global_step": 227, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.2944265604019165, "train_loss_dp": 1.2103391885757446, "train_loss_llm": 0.8408737182617188, "global_step": 228, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.3112713098526, "train_loss_dp": 1.0548627376556396, "train_loss_llm": 2.5640859603881836, "global_step": 229, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.0339800119400024, "train_loss_dp": 0.9086443185806274, "train_loss_llm": 1.25335693359375, "global_step": 230, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.9365898370742798, "train_loss_dp": 0.8714820146560669, "train_loss_llm": 0.6510782241821289, "global_step": 231, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.8826104998588562, "train_loss_dp": 0.7803479433059692, "train_loss_llm": 1.02262544631958, "global_step": 232, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.6868939399719238, "train_loss_dp": 0.8306642770767212, "train_loss_llm": 8.562295913696289, "global_step": 233, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.8654317259788513, "train_loss_dp": 0.7952151298522949, "train_loss_llm": 0.7021660804748535, "global_step": 234, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.929224967956543, "train_loss_dp": 1.495288610458374, "train_loss_llm": 4.339364051818848, "global_step": 235, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.5194882154464722, "train_loss_dp": 1.1792947053909302, "train_loss_llm": 3.4019346237182617, "global_step": 236, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.798079490661621, "train_loss_dp": 1.4390504360198975, "train_loss_llm": 3.5902910232543945, "global_step": 237, "epoch": 0, "lr": 3e-06}
+{"train_loss": 8.501764297485352, "train_loss_dp": 1.2990312576293945, "train_loss_llm": 72.02732849121094, "global_step": 238, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.091771125793457, "train_loss_dp": 0.965569257736206, "train_loss_llm": 1.2620182037353516, "global_step": 239, "epoch": 0, "lr": 3e-06}
+{"train_loss": 3.0677552223205566, "train_loss_dp": 1.128534197807312, "train_loss_llm": 19.392210006713867, "global_step": 240, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.2681124210357666, "train_loss_dp": 1.2322348356246948, "train_loss_llm": 0.35877561569213867, "global_step": 241, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.6115072965621948, "train_loss_dp": 0.9573585987091064, "train_loss_llm": 6.541486740112305, "global_step": 242, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9030525088310242, "train_loss_dp": 0.8517443537712097, "train_loss_llm": 0.513081431388855, "global_step": 243, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1892764568328857, "train_loss_dp": 1.0488935708999634, "train_loss_llm": 1.4038286209106445, "global_step": 244, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.5372421741485596, "train_loss_dp": 1.2262842655181885, "train_loss_llm": 3.109579563140869, "global_step": 245, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0118387937545776, "train_loss_dp": 0.9564860463142395, "train_loss_llm": 0.5535271167755127, "global_step": 246, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0099376440048218, "train_loss_dp": 0.9623837471008301, "train_loss_llm": 0.4755389094352722, "global_step": 247, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.216155767440796, "train_loss_dp": 1.1197764873504639, "train_loss_llm": 0.9637928009033203, "global_step": 248, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.2884701490402222, "train_loss_dp": 1.1492207050323486, "train_loss_llm": 1.3924944400787354, "global_step": 249, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1661309003829956, "train_loss_dp": 0.9364465475082397, "train_loss_llm": 2.2968435287475586, "global_step": 250, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1430352926254272, "train_loss_dp": 1.044571876525879, "train_loss_llm": 0.9846341609954834, "global_step": 251, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0817594528198242, "train_loss_dp": 0.9098814725875854, "train_loss_llm": 1.7187795639038086, "global_step": 252, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1721349954605103, "train_loss_dp": 0.8939852714538574, "train_loss_llm": 2.7814974784851074, "global_step": 253, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.7420083284378052, "train_loss_dp": 0.6618357300758362, "train_loss_llm": 0.8017261028289795, "global_step": 254, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0607516765594482, "train_loss_dp": 0.8920992612838745, "train_loss_llm": 1.6865243911743164, "global_step": 255, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.2599736452102661, "train_loss_dp": 1.2013423442840576, "train_loss_llm": 0.5863132476806641, "global_step": 256, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2969942092895508, "train_loss_dp": 1.2171763181686401, "train_loss_llm": 0.7981783151626587, "global_step": 257, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.1864547729492188, "train_loss_dp": 1.0947213172912598, "train_loss_llm": 0.9173346757888794, "global_step": 258, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.383265495300293, "train_loss_dp": 1.2995078563690186, "train_loss_llm": 0.8375759124755859, "global_step": 259, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.033945083618164, "train_loss_dp": 0.7973261475563049, "train_loss_llm": 2.3661887645721436, "global_step": 260, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.8252993226051331, "train_loss_dp": 0.7147562503814697, "train_loss_llm": 1.1054306030273438, "global_step": 261, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.7940130233764648, "train_loss_dp": 0.6915264129638672, "train_loss_llm": 1.0248661041259766, "global_step": 262, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.0615030527114868, "train_loss_dp": 0.9766282439231873, "train_loss_llm": 0.8487482070922852, "global_step": 263, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2189888954162598, "train_loss_dp": 1.0953198671340942, "train_loss_llm": 1.2366905212402344, "global_step": 264, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.3591597080230713, "train_loss_dp": 1.262351393699646, "train_loss_llm": 0.9680832028388977, "global_step": 265, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.1947976350784302, "train_loss_dp": 1.0695462226867676, "train_loss_llm": 1.2525138854980469, "global_step": 266, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2400920391082764, "train_loss_dp": 1.041294813156128, "train_loss_llm": 1.9879717826843262, "global_step": 267, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2374120950698853, "train_loss_dp": 0.9469956159591675, "train_loss_llm": 2.9041647911071777, "global_step": 268, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.9150847792625427, "train_loss_dp": 0.8269492387771606, "train_loss_llm": 0.8813552856445312, "global_step": 269, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.0384037494659424, "train_loss_dp": 1.0384037494659424, "train_loss_llm": 0.0, "global_step": 270, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.4571900367736816, "train_loss_dp": 1.2486485242843628, "train_loss_llm": 2.0854153633117676, "global_step": 271, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.8692492246627808, "train_loss_dp": 0.8016523122787476, "train_loss_llm": 0.675969123840332, "global_step": 272, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.4429532289505005, "train_loss_dp": 1.2417089939117432, "train_loss_llm": 2.0124425888061523, "global_step": 273, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.8958902955055237, "train_loss_dp": 0.84727543592453, "train_loss_llm": 0.4861488342285156, "global_step": 274, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.2953773736953735, "train_loss_dp": 1.2032415866851807, "train_loss_llm": 0.9213581085205078, "global_step": 275, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 2.3143906593322754, "train_loss_dp": 1.1615830659866333, "train_loss_llm": 11.528074264526367, "global_step": 276, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.052037239074707, "train_loss_dp": 0.9317079186439514, "train_loss_llm": 1.2032928466796875, "global_step": 277, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.0975985527038574, "train_loss_dp": 1.050278902053833, "train_loss_llm": 0.47319698333740234, "global_step": 278, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.2353668212890625, "train_loss_dp": 1.1178455352783203, "train_loss_llm": 1.1752128601074219, "global_step": 279, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.237027883529663, "train_loss_dp": 1.1642677783966064, "train_loss_llm": 0.727601170539856, "global_step": 280, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9696657657623291, "train_loss_dp": 0.9659299254417419, "train_loss_llm": 0.03735828399658203, "global_step": 281, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9303364753723145, "train_loss_dp": 0.8971430063247681, "train_loss_llm": 0.33193498849868774, "global_step": 282, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.110910177230835, "train_loss_dp": 1.0967724323272705, "train_loss_llm": 0.14137744903564453, "global_step": 283, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.303040623664856, "train_loss_dp": 0.9593580961227417, "train_loss_llm": 3.4368247985839844, "global_step": 284, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9239597916603088, "train_loss_dp": 0.8029711246490479, "train_loss_llm": 1.2098865509033203, "global_step": 285, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.3574621677398682, "train_loss_dp": 1.1281466484069824, "train_loss_llm": 2.293154716491699, "global_step": 286, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.333818793296814, "train_loss_dp": 0.9414465427398682, "train_loss_llm": 3.923722743988037, "global_step": 287, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.8355175852775574, "train_loss_dp": 0.7921376824378967, "train_loss_llm": 0.43379878997802734, "global_step": 288, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.4687321186065674, "train_loss_dp": 1.1531985998153687, "train_loss_llm": 3.15533447265625, "global_step": 289, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.2272368669509888, "train_loss_dp": 1.0763548612594604, "train_loss_llm": 1.508819818496704, "global_step": 290, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0415282249450684, "train_loss_dp": 0.69901043176651, "train_loss_llm": 3.4251785278320312, "global_step": 291, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.911309003829956, "train_loss_dp": 0.8489445447921753, "train_loss_llm": 0.6236443519592285, "global_step": 292, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.218719720840454, "train_loss_dp": 1.0774226188659668, "train_loss_llm": 1.4129714965820312, "global_step": 293, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9867818355560303, "train_loss_dp": 0.8808246850967407, "train_loss_llm": 1.059571623802185, "global_step": 294, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.2329882383346558, "train_loss_dp": 1.1722691059112549, "train_loss_llm": 0.6071912050247192, "global_step": 295, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.2350484132766724, "train_loss_dp": 1.1274049282073975, "train_loss_llm": 1.0764347314834595, "global_step": 296, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0290387868881226, "train_loss_dp": 1.0290387868881226, "train_loss_llm": 0.0, "global_step": 297, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9906841516494751, "train_loss_dp": 0.9376249313354492, "train_loss_llm": 0.5305919647216797, "global_step": 298, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.7411508560180664, "train_loss_dp": 0.9818929433822632, "train_loss_llm": 7.5925798416137695, "global_step": 299, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9530736207962036, "train_loss_dp": 0.9208890199661255, "train_loss_llm": 0.32184600830078125, "global_step": 300, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.938224732875824, "train_loss_dp": 0.8615167140960693, "train_loss_llm": 0.7670803070068359, "global_step": 301, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.8433783054351807, "train_loss_dp": 1.1833397150039673, "train_loss_llm": 6.600386619567871, "global_step": 302, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0193783044815063, "train_loss_dp": 0.8860270380973816, "train_loss_llm": 1.3335130214691162, "global_step": 303, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.1019195318222046, "train_loss_dp": 1.0630607604980469, "train_loss_llm": 0.3885877728462219, "global_step": 304, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.1835412979125977, "train_loss_dp": 0.889773964881897, "train_loss_llm": 2.9376730918884277, "global_step": 305, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.109869122505188, "train_loss_dp": 1.008467197418213, "train_loss_llm": 1.0140190124511719, "global_step": 306, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.12992525100708, "train_loss_dp": 1.0554054975509644, "train_loss_llm": 0.7451980113983154, "global_step": 307, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.0191370248794556, "train_loss_dp": 0.8569968938827515, "train_loss_llm": 1.6214008331298828, "global_step": 308, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.380019187927246, "train_loss_dp": 1.2170393466949463, "train_loss_llm": 1.6297988891601562, "global_step": 309, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.5667555332183838, "train_loss_dp": 1.0221761465072632, "train_loss_llm": 5.445793151855469, "global_step": 310, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.2844069004058838, "train_loss_dp": 1.2643412351608276, "train_loss_llm": 0.20065689086914062, "global_step": 311, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.1314382553100586, "train_loss_dp": 1.1034005880355835, "train_loss_llm": 0.2803769111633301, "global_step": 312, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.975796639919281, "train_loss_dp": 0.9192400574684143, "train_loss_llm": 0.5655655860900879, "global_step": 313, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.4264872074127197, "train_loss_dp": 1.2695221900939941, "train_loss_llm": 1.5696501731872559, "global_step": 314, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.7536196112632751, "train_loss_dp": 0.6899515390396118, "train_loss_llm": 0.6366806030273438, "global_step": 315, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.5900564193725586, "train_loss_dp": 1.4025930166244507, "train_loss_llm": 1.8746337890625, "global_step": 316, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.167419195175171, "train_loss_dp": 1.0927602052688599, "train_loss_llm": 0.7465896606445312, "global_step": 317, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.6982849836349487, "train_loss_dp": 1.4461236000061035, "train_loss_llm": 2.521613836288452, "global_step": 318, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.8802670240402222, "train_loss_dp": 0.7859880924224854, "train_loss_llm": 0.9427895545959473, "global_step": 319, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.054127812385559, "train_loss_dp": 0.9827301502227783, "train_loss_llm": 0.7139768600463867, "global_step": 320, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1822024583816528, "train_loss_dp": 1.155442714691162, "train_loss_llm": 0.2675971984863281, "global_step": 321, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 0.9470940232276917, "train_loss_dp": 0.8268672823905945, "train_loss_llm": 1.2022676467895508, "global_step": 322, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.3777821063995361, "train_loss_dp": 1.1707093715667725, "train_loss_llm": 2.0707273483276367, "global_step": 323, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1479547023773193, "train_loss_dp": 1.0440549850463867, "train_loss_llm": 1.0389968156814575, "global_step": 324, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1653469800949097, "train_loss_dp": 1.1307355165481567, "train_loss_llm": 0.34611445665359497, "global_step": 325, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.4541990756988525, "train_loss_dp": 0.9085248708724976, "train_loss_llm": 5.4567413330078125, "global_step": 326, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 8.601066589355469, "train_loss_dp": 1.099638819694519, "train_loss_llm": 75.0142822265625, "global_step": 327, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.5911455154418945, "train_loss_dp": 1.1706275939941406, "train_loss_llm": 4.205179214477539, "global_step": 328, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.8013291358947754, "train_loss_dp": 1.501204252243042, "train_loss_llm": 3.001248359680176, "global_step": 329, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.3674952983856201, "train_loss_dp": 1.1691954135894775, "train_loss_llm": 1.9829988479614258, "global_step": 330, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.0689141750335693, "train_loss_dp": 0.9682847261428833, "train_loss_llm": 1.0062944889068604, "global_step": 331, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.712508201599121, "train_loss_dp": 1.0791127681732178, "train_loss_llm": 6.333953857421875, "global_step": 332, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.419217824935913, "train_loss_dp": 1.2654318809509277, "train_loss_llm": 1.5378599166870117, "global_step": 333, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.5188069343566895, "train_loss_dp": 1.1983133554458618, "train_loss_llm": 3.204935073852539, "global_step": 334, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.354443073272705, "train_loss_dp": 1.2561345100402832, "train_loss_llm": 0.9830858707427979, "global_step": 335, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.0936375856399536, "train_loss_dp": 1.0166246891021729, "train_loss_llm": 0.7701292037963867, "global_step": 336, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.5558805465698242, "train_loss_dp": 1.2430956363677979, "train_loss_llm": 3.127849578857422, "global_step": 337, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2406445741653442, "train_loss_dp": 0.9595973491668701, "train_loss_llm": 2.8104724884033203, "global_step": 338, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.4121851921081543, "train_loss_dp": 1.1830607652664185, "train_loss_llm": 2.2912445068359375, "global_step": 339, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.211958408355713, "train_loss_dp": 1.0891669988632202, "train_loss_llm": 1.2279138565063477, "global_step": 340, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2356290817260742, "train_loss_dp": 0.8481951951980591, "train_loss_llm": 3.874338150024414, "global_step": 341, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2205995321273804, "train_loss_dp": 1.128135323524475, "train_loss_llm": 0.9246425628662109, "global_step": 342, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.7872775793075562, "train_loss_dp": 1.3296529054641724, "train_loss_llm": 4.57624626159668, "global_step": 343, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.798534631729126, "train_loss_dp": 1.4708378314971924, "train_loss_llm": 3.276968002319336, "global_step": 344, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.125102162361145, "train_loss_dp": 1.0238349437713623, "train_loss_llm": 1.012671947479248, "global_step": 345, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2339941263198853, "train_loss_dp": 1.1156691312789917, "train_loss_llm": 1.1832504272460938, "global_step": 346, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.023496389389038, "train_loss_dp": 0.9493876695632935, "train_loss_llm": 0.7410869598388672, "global_step": 347, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9859465956687927, "train_loss_dp": 0.9120997786521912, "train_loss_llm": 0.7384681701660156, "global_step": 348, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.0182318687438965, "train_loss_dp": 0.9648668766021729, "train_loss_llm": 0.5336494445800781, "global_step": 349, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.0910234451293945, "train_loss_dp": 1.0466371774673462, "train_loss_llm": 0.44386279582977295, "global_step": 350, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.9113807678222656, "train_loss_dp": 1.3387916088104248, "train_loss_llm": 5.72589111328125, "global_step": 351, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9273109436035156, "train_loss_dp": 0.8427250981330872, "train_loss_llm": 0.8458586931228638, "global_step": 352, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.8119145631790161, "train_loss_dp": 1.340092420578003, "train_loss_llm": 4.718221664428711, "global_step": 353, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.0551027059555054, "train_loss_dp": 0.9461878538131714, "train_loss_llm": 1.089148759841919, "global_step": 354, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.0749891996383667, "train_loss_dp": 0.9747481346130371, "train_loss_llm": 1.002410650253296, "global_step": 355, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.7518798112869263, "train_loss_dp": 1.2779579162597656, "train_loss_llm": 4.739218711853027, "global_step": 356, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.2956786155700684, "train_loss_dp": 1.0528228282928467, "train_loss_llm": 2.428558349609375, "global_step": 357, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.143129587173462, "train_loss_dp": 1.1054202318191528, "train_loss_llm": 0.37709343433380127, "global_step": 358, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.4521291255950928, "train_loss_dp": 1.078309178352356, "train_loss_llm": 3.738199234008789, "global_step": 359, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.212602138519287, "train_loss_dp": 1.1120707988739014, "train_loss_llm": 1.0053133964538574, "global_step": 360, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 0.863084077835083, "train_loss_dp": 0.8114221096038818, "train_loss_llm": 0.5166196227073669, "global_step": 361, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 0.9048126935958862, "train_loss_dp": 0.8028910160064697, "train_loss_llm": 1.019216537475586, "global_step": 362, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.0901223421096802, "train_loss_dp": 1.0465679168701172, "train_loss_llm": 0.43554437160491943, "global_step": 363, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.304418683052063, "train_loss_dp": 1.2207465171813965, "train_loss_llm": 0.8367214202880859, "global_step": 364, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.4729630947113037, "train_loss_dp": 1.0532861948013306, "train_loss_llm": 4.196768760681152, "global_step": 365, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 6.764286994934082, "train_loss_dp": 0.7992550134658813, "train_loss_llm": 59.65032196044922, "global_step": 366, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 0.9501202702522278, "train_loss_dp": 0.9044771194458008, "train_loss_llm": 0.45643162727355957, "global_step": 367, "epoch": 0, "lr": 4.6e-06}
+{"train_loss": 1.2327042818069458, "train_loss_dp": 0.9989014863967896, "train_loss_llm": 2.3380284309387207, "global_step": 368, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.3335950374603271, "train_loss_dp": 0.8981943130493164, "train_loss_llm": 4.354007720947266, "global_step": 369, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.281046986579895, "train_loss_dp": 1.1535141468048096, "train_loss_llm": 1.2753286361694336, "global_step": 370, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.155122995376587, "train_loss_dp": 1.1267238855361938, "train_loss_llm": 0.28399085998535156, "global_step": 371, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.0182808637619019, "train_loss_dp": 0.864569902420044, "train_loss_llm": 1.537109375, "global_step": 372, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.3717584609985352, "train_loss_dp": 1.2809311151504517, "train_loss_llm": 0.9082736968994141, "global_step": 373, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.3684637546539307, "train_loss_dp": 1.1126835346221924, "train_loss_llm": 2.557802200317383, "global_step": 374, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.2146410942077637, "train_loss_dp": 1.1072661876678467, "train_loss_llm": 1.0737495422363281, "global_step": 375, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 0.8189621567726135, "train_loss_dp": 0.7191091775894165, "train_loss_llm": 0.9985296726226807, "global_step": 376, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.1575751304626465, "train_loss_dp": 1.1141235828399658, "train_loss_llm": 0.43451499938964844, "global_step": 377, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.4656541347503662, "train_loss_dp": 0.8398296236991882, "train_loss_llm": 6.258244514465332, "global_step": 378, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 0.9062263369560242, "train_loss_dp": 0.8345751762390137, "train_loss_llm": 0.7165117263793945, "global_step": 379, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 0.785290002822876, "train_loss_dp": 0.7081749439239502, "train_loss_llm": 0.7711505889892578, "global_step": 380, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.5284104347229004, "train_loss_dp": 0.9806740283966064, "train_loss_llm": 5.4773640632629395, "global_step": 381, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 0.9272936582565308, "train_loss_dp": 0.8962379097938538, "train_loss_llm": 0.31055742502212524, "global_step": 382, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 0.9200815558433533, "train_loss_dp": 0.8480089902877808, "train_loss_llm": 0.7207257747650146, "global_step": 383, "epoch": 0, "lr": 4.800000000000001e-06}
+{"train_loss": 1.5686269998550415, "train_loss_dp": 1.0998754501342773, "train_loss_llm": 4.6875152587890625, "global_step": 384, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.3136670589447021, "train_loss_dp": 0.9578797817230225, "train_loss_llm": 3.557872772216797, "global_step": 385, "epoch": 0, "lr": 5e-06}
+{"train_loss": 0.8924936056137085, "train_loss_dp": 0.8924936056137085, "train_loss_llm": 0.0, "global_step": 386, "epoch": 0, "lr": 5e-06}
+{"train_loss": 0.9080877304077148, "train_loss_dp": 0.8301392793655396, "train_loss_llm": 0.7794842720031738, "global_step": 387, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.0032354593276978, "train_loss_dp": 0.8643873929977417, "train_loss_llm": 1.3884801864624023, "global_step": 388, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.1405493021011353, "train_loss_dp": 1.0649242401123047, "train_loss_llm": 0.7562503814697266, "global_step": 389, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.2532830238342285, "train_loss_dp": 1.229729175567627, "train_loss_llm": 0.23553848266601562, "global_step": 390, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.3544565439224243, "train_loss_dp": 1.1068384647369385, "train_loss_llm": 2.4761810302734375, "global_step": 391, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.1640405654907227, "train_loss_dp": 0.87496018409729, "train_loss_llm": 2.8908042907714844, "global_step": 392, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.2347140312194824, "train_loss_dp": 1.0783827304840088, "train_loss_llm": 1.5633125305175781, "global_step": 393, "epoch": 0, "lr": 5e-06}
+{"train_loss": 0.9459535479545593, "train_loss_dp": 0.8556427359580994, "train_loss_llm": 0.9031081199645996, "global_step": 394, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.110896348953247, "train_loss_dp": 0.9188845157623291, "train_loss_llm": 1.9201178550720215, "global_step": 395, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.4927194118499756, "train_loss_dp": 1.3659543991088867, "train_loss_llm": 1.2676506042480469, "global_step": 396, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.116539716720581, "train_loss_dp": 1.0871920585632324, "train_loss_llm": 0.29347705841064453, "global_step": 397, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.317513346672058, "train_loss_dp": 1.1529531478881836, "train_loss_llm": 1.645601511001587, "global_step": 398, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.6743745803833008, "train_loss_dp": 1.047851800918579, "train_loss_llm": 6.265228271484375, "global_step": 399, "epoch": 0, "lr": 5e-06}
+{"train_loss": 1.2950626611709595, "train_loss_dp": 1.2107579708099365, "train_loss_llm": 0.8430471420288086, "global_step": 400, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.2460932731628418, "train_loss_dp": 1.006636619567871, "train_loss_llm": 2.394566535949707, "global_step": 401, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.0128347873687744, "train_loss_dp": 0.9341101050376892, "train_loss_llm": 0.7872467041015625, "global_step": 402, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.7355167865753174, "train_loss_dp": 1.1870709657669067, "train_loss_llm": 5.484457492828369, "global_step": 403, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.7789915800094604, "train_loss_dp": 1.1288546323776245, "train_loss_llm": 6.501369476318359, "global_step": 404, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.7179841995239258, "train_loss_dp": 1.3579055070877075, "train_loss_llm": 3.6007862091064453, "global_step": 405, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 0.8909493088722229, "train_loss_dp": 0.7146732807159424, "train_loss_llm": 1.7627601623535156, "global_step": 406, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.1077065467834473, "train_loss_dp": 0.9934607148170471, "train_loss_llm": 1.142458438873291, "global_step": 407, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.1168633699417114, "train_loss_dp": 1.0349177122116089, "train_loss_llm": 0.8194565773010254, "global_step": 408, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.1338621377944946, "train_loss_dp": 0.8841963410377502, "train_loss_llm": 2.4966583251953125, "global_step": 409, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.6283756494522095, "train_loss_dp": 1.2870562076568604, "train_loss_llm": 3.4131946563720703, "global_step": 410, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.8402247428894043, "train_loss_dp": 0.9473253488540649, "train_loss_llm": 8.928993225097656, "global_step": 411, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.2175792455673218, "train_loss_dp": 1.154588222503662, "train_loss_llm": 0.6299102306365967, "global_step": 412, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.392419457435608, "train_loss_dp": 1.3371851444244385, "train_loss_llm": 0.5523428916931152, "global_step": 413, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.4752495288848877, "train_loss_dp": 1.2088572978973389, "train_loss_llm": 2.6639223098754883, "global_step": 414, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 0.8523682355880737, "train_loss_dp": 0.7866177558898926, "train_loss_llm": 0.6575050354003906, "global_step": 415, "epoch": 0, "lr": 5.2e-06}
+{"train_loss": 1.790925145149231, "train_loss_dp": 1.3473050594329834, "train_loss_llm": 4.436201095581055, "global_step": 416, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 0.8571856021881104, "train_loss_dp": 0.693175196647644, "train_loss_llm": 1.6401042938232422, "global_step": 417, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 0.877697765827179, "train_loss_dp": 0.8483004570007324, "train_loss_llm": 0.2939729690551758, "global_step": 418, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 0.872223973274231, "train_loss_dp": 0.872223973274231, "train_loss_llm": 0.0, "global_step": 419, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.0023760795593262, "train_loss_dp": 0.8468683958053589, "train_loss_llm": 1.5550765991210938, "global_step": 420, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.4021772146224976, "train_loss_dp": 1.2720518112182617, "train_loss_llm": 1.3012542724609375, "global_step": 421, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.8215614557266235, "train_loss_dp": 1.0972496271133423, "train_loss_llm": 7.2431182861328125, "global_step": 422, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.6343142986297607, "train_loss_dp": 1.2888023853302002, "train_loss_llm": 3.4551191329956055, "global_step": 423, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 3.6219587326049805, "train_loss_dp": 0.9287045001983643, "train_loss_llm": 26.932540893554688, "global_step": 424, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.274327039718628, "train_loss_dp": 1.0789847373962402, "train_loss_llm": 1.953423261642456, "global_step": 425, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.3354212045669556, "train_loss_dp": 1.2104196548461914, "train_loss_llm": 1.2500152587890625, "global_step": 426, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.149751901626587, "train_loss_dp": 1.108689785003662, "train_loss_llm": 0.4106205701828003, "global_step": 427, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.2743741273880005, "train_loss_dp": 1.1768531799316406, "train_loss_llm": 0.9752097129821777, "global_step": 428, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.4119147062301636, "train_loss_dp": 1.3391032218933105, "train_loss_llm": 0.7281150817871094, "global_step": 429, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.041386365890503, "train_loss_dp": 0.8086467385292053, "train_loss_llm": 2.3273963928222656, "global_step": 430, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 1.364476203918457, "train_loss_dp": 1.1290169954299927, "train_loss_llm": 2.3545925617218018, "global_step": 431, "epoch": 0, "lr": 5.4e-06}
+{"train_loss": 0.8435541391372681, "train_loss_dp": 0.8017194867134094, "train_loss_llm": 0.4183465242385864, "global_step": 432, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.878893494606018, "train_loss_dp": 1.6203334331512451, "train_loss_llm": 2.5856006145477295, "global_step": 433, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.9828730225563049, "train_loss_dp": 0.935837984085083, "train_loss_llm": 0.4703502655029297, "global_step": 434, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.8847653269767761, "train_loss_dp": 0.8137878775596619, "train_loss_llm": 0.7097742557525635, "global_step": 435, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.9631547927856445, "train_loss_dp": 0.9059398174285889, "train_loss_llm": 0.5721497535705566, "global_step": 436, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.1272735595703125, "train_loss_dp": 1.0568976402282715, "train_loss_llm": 0.7037594318389893, "global_step": 437, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.1863162517547607, "train_loss_dp": 1.1475646495819092, "train_loss_llm": 0.38751548528671265, "global_step": 438, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.0133358240127563, "train_loss_dp": 0.88441002368927, "train_loss_llm": 1.289257526397705, "global_step": 439, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.9175423979759216, "train_loss_dp": 0.9012753963470459, "train_loss_llm": 0.16267013549804688, "global_step": 440, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.657976508140564, "train_loss_dp": 0.6232531070709229, "train_loss_llm": 0.3472341299057007, "global_step": 441, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.1023509502410889, "train_loss_dp": 0.838772177696228, "train_loss_llm": 2.635787010192871, "global_step": 442, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.2490204572677612, "train_loss_dp": 1.1512353420257568, "train_loss_llm": 0.977851152420044, "global_step": 443, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.1950886249542236, "train_loss_dp": 1.0829083919525146, "train_loss_llm": 1.1218023300170898, "global_step": 444, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.1272640228271484, "train_loss_dp": 0.9348856210708618, "train_loss_llm": 1.9237842559814453, "global_step": 445, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.0373958349227905, "train_loss_dp": 0.8847924470901489, "train_loss_llm": 1.5260334014892578, "global_step": 446, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 0.7618308663368225, "train_loss_dp": 0.7300698757171631, "train_loss_llm": 0.317609965801239, "global_step": 447, "epoch": 0, "lr": 5.600000000000001e-06}
+{"train_loss": 1.282822847366333, "train_loss_dp": 1.16628098487854, "train_loss_llm": 1.1654183864593506, "global_step": 448, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.8466892838478088, "train_loss_dp": 0.7841517329216003, "train_loss_llm": 0.6253752708435059, "global_step": 449, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.7437058091163635, "train_loss_dp": 0.6678324937820435, "train_loss_llm": 0.7587330341339111, "global_step": 450, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.3584765195846558, "train_loss_dp": 0.9888741374015808, "train_loss_llm": 3.696023941040039, "global_step": 451, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.8006879091262817, "train_loss_dp": 0.7572035789489746, "train_loss_llm": 0.4348430633544922, "global_step": 452, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 2.631497621536255, "train_loss_dp": 1.1849663257598877, "train_loss_llm": 14.465312957763672, "global_step": 453, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.1777117252349854, "train_loss_dp": 0.9539130926132202, "train_loss_llm": 2.2379865646362305, "global_step": 454, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.7895722389221191, "train_loss_dp": 0.7646273374557495, "train_loss_llm": 0.24944919347763062, "global_step": 455, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.1219508647918701, "train_loss_dp": 1.033158779144287, "train_loss_llm": 0.8879203796386719, "global_step": 456, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.9683038592338562, "train_loss_dp": 0.8938180804252625, "train_loss_llm": 0.7448577880859375, "global_step": 457, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.1356648206710815, "train_loss_dp": 1.0374338626861572, "train_loss_llm": 0.9823098182678223, "global_step": 458, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.9663053154945374, "train_loss_dp": 0.9218496084213257, "train_loss_llm": 0.44455692172050476, "global_step": 459, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.1078722476959229, "train_loss_dp": 1.040675401687622, "train_loss_llm": 0.6719689965248108, "global_step": 460, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 0.9565858840942383, "train_loss_dp": 0.8247555494308472, "train_loss_llm": 1.3183033466339111, "global_step": 461, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.006348729133606, "train_loss_dp": 0.8521592617034912, "train_loss_llm": 1.5418946743011475, "global_step": 462, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.1354830265045166, "train_loss_dp": 1.0032970905303955, "train_loss_llm": 1.3218588829040527, "global_step": 463, "epoch": 0, "lr": 5.8e-06}
+{"train_loss": 1.3391337394714355, "train_loss_dp": 1.2277804613113403, "train_loss_llm": 1.1135326623916626, "global_step": 464, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.1951366662979126, "train_loss_dp": 1.0242871046066284, "train_loss_llm": 1.7084956169128418, "global_step": 465, "epoch": 0, "lr": 6e-06}
+{"train_loss": 0.9524789452552795, "train_loss_dp": 0.8449944853782654, "train_loss_llm": 1.0748443603515625, "global_step": 466, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.1748466491699219, "train_loss_dp": 1.1376163959503174, "train_loss_llm": 0.3723025321960449, "global_step": 467, "epoch": 0, "lr": 6e-06}
+{"train_loss": 0.9307315349578857, "train_loss_dp": 0.8593368530273438, "train_loss_llm": 0.7139468193054199, "global_step": 468, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.0048255920410156, "train_loss_dp": 0.9170733094215393, "train_loss_llm": 0.8775224685668945, "global_step": 469, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.0490206480026245, "train_loss_dp": 0.9156352281570435, "train_loss_llm": 1.3338546752929688, "global_step": 470, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.212315320968628, "train_loss_dp": 1.1209874153137207, "train_loss_llm": 0.9132795333862305, "global_step": 471, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.061944603919983, "train_loss_dp": 0.8268389701843262, "train_loss_llm": 2.3510560989379883, "global_step": 472, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.0483485460281372, "train_loss_dp": 0.9487961530685425, "train_loss_llm": 0.9955241680145264, "global_step": 473, "epoch": 0, "lr": 6e-06}
+{"train_loss": 0.7811214327812195, "train_loss_dp": 0.7548961639404297, "train_loss_llm": 0.26225247979164124, "global_step": 474, "epoch": 0, "lr": 6e-06}
+{"train_loss": 5.385737895965576, "train_loss_dp": 1.445860505104065, "train_loss_llm": 39.398773193359375, "global_step": 475, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.2175748348236084, "train_loss_dp": 1.030663013458252, "train_loss_llm": 1.8691182136535645, "global_step": 476, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.0432624816894531, "train_loss_dp": 0.9906595945358276, "train_loss_llm": 0.5260290503501892, "global_step": 477, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.074748158454895, "train_loss_dp": 0.9921778440475464, "train_loss_llm": 0.8257026672363281, "global_step": 478, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.0475541353225708, "train_loss_dp": 1.0099751949310303, "train_loss_llm": 0.3757896423339844, "global_step": 479, "epoch": 0, "lr": 6e-06}
+{"train_loss": 1.120253086090088, "train_loss_dp": 1.0893492698669434, "train_loss_llm": 0.3090381622314453, "global_step": 480, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.1271027326583862, "train_loss_dp": 1.0818147659301758, "train_loss_llm": 0.4528801441192627, "global_step": 481, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 0.9518130421638489, "train_loss_dp": 0.89238041639328, "train_loss_llm": 0.5943260192871094, "global_step": 482, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.0708280801773071, "train_loss_dp": 1.0392272472381592, "train_loss_llm": 0.3160078525543213, "global_step": 483, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.1280028820037842, "train_loss_dp": 0.9571258425712585, "train_loss_llm": 1.7087697982788086, "global_step": 484, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.0590779781341553, "train_loss_dp": 0.9596970677375793, "train_loss_llm": 0.9938087463378906, "global_step": 485, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.2253177165985107, "train_loss_dp": 1.0896902084350586, "train_loss_llm": 1.3562755584716797, "global_step": 486, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.3276679515838623, "train_loss_dp": 1.1569864749908447, "train_loss_llm": 1.7068142890930176, "global_step": 487, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 0.8685849905014038, "train_loss_dp": 0.7824704647064209, "train_loss_llm": 0.8611454963684082, "global_step": 488, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 0.8577615022659302, "train_loss_dp": 0.7906551361083984, "train_loss_llm": 0.6710638999938965, "global_step": 489, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 0.9480370879173279, "train_loss_dp": 0.8710740804672241, "train_loss_llm": 0.769629955291748, "global_step": 490, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.1902002096176147, "train_loss_dp": 1.042900800704956, "train_loss_llm": 1.4729938507080078, "global_step": 491, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.1359567642211914, "train_loss_dp": 0.929570734500885, "train_loss_llm": 2.0638608932495117, "global_step": 492, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.3535982370376587, "train_loss_dp": 1.219968557357788, "train_loss_llm": 1.336296796798706, "global_step": 493, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 0.8513914942741394, "train_loss_dp": 0.7917338013648987, "train_loss_llm": 0.5965766906738281, "global_step": 494, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.2372939586639404, "train_loss_dp": 1.1521356105804443, "train_loss_llm": 0.8515839576721191, "global_step": 495, "epoch": 0, "lr": 6.2e-06}
+{"train_loss": 1.1444624662399292, "train_loss_dp": 0.8614665269851685, "train_loss_llm": 2.829958915710449, "global_step": 496, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.0589098930358887, "train_loss_dp": 0.9888302087783813, "train_loss_llm": 0.7007968425750732, "global_step": 497, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.0467681884765625, "train_loss_dp": 0.9921185970306396, "train_loss_llm": 0.5464954376220703, "global_step": 498, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.0308401584625244, "train_loss_dp": 0.9958069324493408, "train_loss_llm": 0.35033226013183594, "global_step": 499, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.1642792224884033, "train_loss_dp": 1.1562578678131104, "train_loss_llm": 0.08021354675292969, "global_step": 500, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.2078791856765747, "train_loss_dp": 1.0413066148757935, "train_loss_llm": 1.6657257080078125, "global_step": 501, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 2.1394505500793457, "train_loss_dp": 0.911308765411377, "train_loss_llm": 12.281417846679688, "global_step": 502, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.3157776594161987, "train_loss_dp": 0.8173350095748901, "train_loss_llm": 4.984426498413086, "global_step": 503, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.2698397636413574, "train_loss_dp": 1.2032103538513184, "train_loss_llm": 0.6662936210632324, "global_step": 504, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 0.7326494455337524, "train_loss_dp": 0.6529514789581299, "train_loss_llm": 0.7969799041748047, "global_step": 505, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.1986627578735352, "train_loss_dp": 0.9172288179397583, "train_loss_llm": 2.8143386840820312, "global_step": 506, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.265599250793457, "train_loss_dp": 1.132995367050171, "train_loss_llm": 1.3260383605957031, "global_step": 507, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 0.7992089986801147, "train_loss_dp": 0.6785165071487427, "train_loss_llm": 1.2069251537322998, "global_step": 508, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.0773632526397705, "train_loss_dp": 0.7375759482383728, "train_loss_llm": 3.3978729248046875, "global_step": 509, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.2594934701919556, "train_loss_dp": 0.9950122237205505, "train_loss_llm": 2.64481258392334, "global_step": 510, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 1.08734929561615, "train_loss_dp": 0.9889348745346069, "train_loss_llm": 0.9841439723968506, "global_step": 511, "epoch": 0, "lr": 6.4000000000000006e-06}
+{"train_loss": 0.6992119550704956, "train_loss_dp": 0.6436371803283691, "train_loss_llm": 0.5557479858398438, "global_step": 512, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 0.859641969203949, "train_loss_dp": 0.7813206911087036, "train_loss_llm": 0.7832130193710327, "global_step": 513, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.406069040298462, "train_loss_dp": 1.004441261291504, "train_loss_llm": 4.01627779006958, "global_step": 514, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.269261121749878, "train_loss_dp": 1.2026723623275757, "train_loss_llm": 0.6658878326416016, "global_step": 515, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.359386682510376, "train_loss_dp": 1.2447412014007568, "train_loss_llm": 1.1464545726776123, "global_step": 516, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.6646008491516113, "train_loss_dp": 0.8686027526855469, "train_loss_llm": 7.959980010986328, "global_step": 517, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 0.8993417024612427, "train_loss_dp": 0.8265578150749207, "train_loss_llm": 0.7278385758399963, "global_step": 518, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.098867654800415, "train_loss_dp": 1.0459330081939697, "train_loss_llm": 0.5293464660644531, "global_step": 519, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.465453863143921, "train_loss_dp": 1.0739121437072754, "train_loss_llm": 3.915416717529297, "global_step": 520, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.2951494455337524, "train_loss_dp": 1.153433084487915, "train_loss_llm": 1.4171638488769531, "global_step": 521, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.3767952919006348, "train_loss_dp": 1.256885290145874, "train_loss_llm": 1.1991004943847656, "global_step": 522, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 0.9173856377601624, "train_loss_dp": 0.8117272853851318, "train_loss_llm": 1.0565834045410156, "global_step": 523, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.4451069831848145, "train_loss_dp": 1.106528401374817, "train_loss_llm": 3.3857855796813965, "global_step": 524, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 0.9486806392669678, "train_loss_dp": 0.8122401237487793, "train_loss_llm": 1.3644053936004639, "global_step": 525, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 0.8686879277229309, "train_loss_dp": 0.7733035087585449, "train_loss_llm": 0.9538443088531494, "global_step": 526, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.18575119972229, "train_loss_dp": 1.0694048404693604, "train_loss_llm": 1.1634635925292969, "global_step": 527, "epoch": 0, "lr": 6.6e-06}
+{"train_loss": 1.0660433769226074, "train_loss_dp": 0.9424158334732056, "train_loss_llm": 1.2362759113311768, "global_step": 528, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 3.3765199184417725, "train_loss_dp": 1.0830602645874023, "train_loss_llm": 22.934595108032227, "global_step": 529, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.16761314868927, "train_loss_dp": 1.0733002424240112, "train_loss_llm": 0.9431291222572327, "global_step": 530, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.102622628211975, "train_loss_dp": 0.9997084140777588, "train_loss_llm": 1.0291423797607422, "global_step": 531, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 2.3719563484191895, "train_loss_dp": 0.9761481285095215, "train_loss_llm": 13.958083152770996, "global_step": 532, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 0.9829270839691162, "train_loss_dp": 0.9019390344619751, "train_loss_llm": 0.8098807334899902, "global_step": 533, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 0.7307918667793274, "train_loss_dp": 0.7192401885986328, "train_loss_llm": 0.11551666259765625, "global_step": 534, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.1923335790634155, "train_loss_dp": 0.987787127494812, "train_loss_llm": 2.045464038848877, "global_step": 535, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 0.999212384223938, "train_loss_dp": 0.8321586847305298, "train_loss_llm": 1.6705372333526611, "global_step": 536, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.2036082744598389, "train_loss_dp": 1.1704181432724, "train_loss_llm": 0.33190155029296875, "global_step": 537, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.1001055240631104, "train_loss_dp": 0.9854047894477844, "train_loss_llm": 1.1470069885253906, "global_step": 538, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.8226250410079956, "train_loss_dp": 1.288301706314087, "train_loss_llm": 5.343233108520508, "global_step": 539, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.5440666675567627, "train_loss_dp": 0.9839023351669312, "train_loss_llm": 5.601642608642578, "global_step": 540, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 0.7445494532585144, "train_loss_dp": 0.7130340933799744, "train_loss_llm": 0.3151535987854004, "global_step": 541, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.032609224319458, "train_loss_dp": 0.9593443870544434, "train_loss_llm": 0.7326478958129883, "global_step": 542, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 0.8901945352554321, "train_loss_dp": 0.793813943862915, "train_loss_llm": 0.96380615234375, "global_step": 543, "epoch": 0, "lr": 6.800000000000001e-06}
+{"train_loss": 1.47462797164917, "train_loss_dp": 1.234516978263855, "train_loss_llm": 2.4011101722717285, "global_step": 544, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 1.09365975856781, "train_loss_dp": 1.0563771724700928, "train_loss_llm": 0.37282562255859375, "global_step": 545, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 1.0600663423538208, "train_loss_dp": 0.9242138266563416, "train_loss_llm": 1.358525276184082, "global_step": 546, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 2.1589694023132324, "train_loss_dp": 1.0199064016342163, "train_loss_llm": 11.390628814697266, "global_step": 547, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 1.2346982955932617, "train_loss_dp": 1.1422219276428223, "train_loss_llm": 0.9247641563415527, "global_step": 548, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 1.4717503786087036, "train_loss_dp": 1.0479812622070312, "train_loss_llm": 4.237691402435303, "global_step": 549, "epoch": 0, "lr": 7.000000000000001e-06}
+{"train_loss": 0.8508371114730835, "train_loss_dp": 0.7344393730163574, "train_loss_llm": 1.1639776229858398, "global_step": 550, "epoch": 0, "lr": 7.000000000000001e-06}

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/train.log

@@ -0,0 +1,14 @@
+[2026-04-14 06:25:33,511][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:25:33,511][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:25:33,511][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:25:35,970][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:25:35,972][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:25:35,972][datasets][INFO] - JAX version 0.4.30 available.
+[2026-04-14 06:25:37,623][llmbc.model.diffusion.conditional_unet1d][INFO] - number of parameters: 6.501479e+07
+[2026-04-14 06:25:41,925][root][INFO] - running build_ext
+[2026-04-14 06:25:41,927][root][INFO] - building 'mujoco_py.cymj' extension
+[2026-04-14 06:25:41,927][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -fPIC -O2 -isystem /venv/llm-bc/include -I/venv/llm-bc/include -fPIC -O2 -isystem /venv/llm-bc/include -fPIC -I/venv/llm-bc/lib/python3.9/site-packages/mujoco_py -I/root/.mujoco/mujoco210/include -I/venv/llm-bc/lib/python3.9/site-packages/numpy/core/include -I/venv/llm-bc/include/python3.9 -c /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.c -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.o -fopenmp -w
+[2026-04-14 06:26:23,961][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -O2 -Wall -fPIC -O2 -isystem /venv/llm-bc/include -I/venv/llm-bc/include -fPIC -O2 -isystem /venv/llm-bc/include -fPIC -I/venv/llm-bc/lib/python3.9/site-packages/mujoco_py -I/root/.mujoco/mujoco210/include -I/venv/llm-bc/lib/python3.9/site-packages/numpy/core/include -I/venv/llm-bc/include/python3.9 -c /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.c -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.o -fopenmp -w
+[2026-04-14 06:26:24,083][root][INFO] - gcc -pthread -B /venv/llm-bc/compiler_compat -shared -Wl,-rpath,/venv/llm-bc/lib -Wl,-rpath-link,/venv/llm-bc/lib -L/venv/llm-bc/lib -L/venv/llm-bc/lib -Wl,-rpath,/venv/llm-bc/lib -Wl,-rpath-link,/venv/llm-bc/lib -L/venv/llm-bc/lib /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/cymj.o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/temp.linux-x86_64-cpython-39/venv/llm-bc/lib/python3.9/site-packages/mujoco_py/gl/osmesashim.o -L/root/.mujoco/mujoco210/bin -Wl,--enable-new-dtags,-rpath,/root/.mujoco/mujoco210/bin -lmujoco210 -lglewosmesa -lOSMesa -lGL -o /venv/llm-bc/lib/python3.9/site-packages/mujoco_py/generated/_pyxbld_2.1.2.14_39_linuxcpuextensionbuilder/lib.linux-x86_64-cpython-39/mujoco_py/cymj.cpython-39-x86_64-linux-gnu.so -fopenmp
+[2026-04-14 06:26:27,226][absl][INFO] - MUJOCO_GL=osmesa, attempting to import specified OpenGL backend.
+[2026-04-14 06:26:27,293][absl][INFO] - MuJoCo library version is: 2.3.7

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/files/config.yaml

@@ -0,0 +1,304 @@
+_target_:
+    value: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+_wandb:
+    value:
+        cli_version: 0.18.6
+        m: []
+        python_version: 3.9.25
+        t:
+            "1":
+                - 1
+                - 2
+                - 3
+                - 5
+                - 11
+                - 12
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 55
+                - 71
+                - 83
+                - 95
+                - 98
+                - 100
+                - 105
+            "2":
+                - 1
+                - 2
+                - 3
+                - 5
+                - 11
+                - 12
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 55
+                - 71
+                - 83
+                - 95
+                - 98
+                - 100
+                - 105
+            "3":
+                - 13
+                - 15
+                - 16
+                - 23
+                - 55
+                - 61
+            "4": 3.9.25
+            "5": 0.18.6
+            "6": 4.47.1
+            "8":
+                - 5
+            "12": 0.18.6
+            "13": linux-x86_64
+action_dim:
+    value: 4
+checkpoint:
+    value:
+        save_last_ckpt: true
+        save_last_snapshot: false
+        topk:
+            format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+            k: 5
+            mode: max
+            monitor_key: test_success_rate
+dataloader:
+    value:
+        batch_size: 16
+        num_workers: 0
+        persistent_workers: false
+        pin_memory: false
+        shuffle: true
+ema:
+    value:
+        _target_: llmbc.model.diffusion.ema_model.EMAModel
+        inv_gamma: 1
+        max_value: 0.9999
+        min_value: 0
+        power: 0.75
+        update_after_step: 0
+exp_name:
+    value: default
+horizon:
+    value: 1
+keypoint_visible_rate:
+    value: 1
+llm:
+    value:
+        causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+        checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+        config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+        finetune_mode: orig
+        hydra:
+            job:
+                override_dirname: HuggingFaceTB/SmolLM2-135M-Instruct
+            run:
+                dir: data/outputs/2026.04.14/06.25.31_HuggingFaceTB/SmolLM2-135M-Instruct
+        llm_mode: ete-finetuned
+        lora_config:
+            bias: none
+            lora_alpha: 64
+            lora_dropout: 0.05
+            r: 32
+            task_type: CAUSAL_LM
+        max_length: 100
+        model_name: SmolLM2-135M-Instruct
+        name: HuggingFaceTB/SmolLM2-135M-Instruct
+        prompter:
+            _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+            use_joint_mlp_projector: true
+        use_joint_mlp_projector: true
+        use_quantization: false
+llm_do_sample:
+    value: false
+llm_orig_expert_feedback:
+    value: true
+logging:
+    value:
+        group: null
+        id: null
+        mode: online
+        name: 2026.04.14-06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2
+        project: sweep-v2-training
+        resume: true
+        tags:
+            - train_llm_diffusion_unet_lowdim
+            - sweep-v2
+            - default
+model_name:
+    value: HuggingFaceTB/SmolLM2-135M-Instruct
+multi_run:
+    value:
+        run_dir: data/outputs/2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2
+        wandb_name_base: 2026.04.14-06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2
+n_action_steps:
+    value: 1
+n_latency_steps:
+    value: 0
+n_obs_steps:
+    value: 1
+name:
+    value: train_llm_diffusion_unet_lowdim
+obs_as_global_cond:
+    value: true
+obs_as_local_cond:
+    value: false
+obs_dim:
+    value: 9
+optimizer:
+    value:
+        _target_: torch.optim.AdamW
+        betas:
+            - 0.95
+            - 0.999
+        eps: 1e-08
+        lr: 0.0001
+        weight_decay: 1e-06
+output_dir:
+    value: /workspace/LLM-BC/data/outputs/2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2
+past_action_visible:
+    value: false
+policy:
+    value:
+        _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+        action_dim: 4
+        horizon: 1
+        llm_discriminator:
+            _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+            llm_translator:
+                _target_: llmbc.translator.llm_translator.LLMTranslator
+                action_dim: 4
+                cfg:
+                    causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+                    checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+                    config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+                    finetune_mode: orig
+                    hydra:
+                        job:
+                            override_dirname: HuggingFaceTB/SmolLM2-135M-Instruct
+                        run:
+                            dir: data/outputs/2026.04.14/06.25.31_HuggingFaceTB/SmolLM2-135M-Instruct
+                    llm_mode: ete-finetuned
+                    lora_config:
+                        bias: none
+                        lora_alpha: 64
+                        lora_dropout: 0.05
+                        r: 32
+                        task_type: CAUSAL_LM
+                    max_length: 100
+                    model_name: SmolLM2-135M-Instruct
+                    name: HuggingFaceTB/SmolLM2-135M-Instruct
+                    prompter:
+                        _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+                        use_joint_mlp_projector: true
+                    use_joint_mlp_projector: true
+                    use_quantization: false
+                horizon: 1
+                n_action_steps: 1
+                n_obs_steps: 1
+                obs_dim: 9
+            task_id: sweep-v2
+        loss_dp_weight: 1
+        loss_llm_weight: 0.1
+        model:
+            _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+            cond_predict_scale: true
+            diffusion_step_embed_dim: 256
+            down_dims:
+                - 256
+                - 512
+                - 1024
+            global_cond_dim: 9
+            input_dim: 4
+            kernel_size: 5
+            local_cond_dim: null
+            n_groups: 8
+        n_action_steps: 1
+        n_obs_steps: 1
+        noise_scheduler:
+            _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+            beta_end: 0.02
+            beta_schedule: squaredcos_cap_v2
+            beta_start: 0.0001
+            clip_sample: true
+            num_train_timesteps: 100
+            prediction_type: epsilon
+            variance_type: fixed_small
+        normalize_llm_loss: true
+        num_inference_steps: 100
+        oa_step_convention: true
+        obs_as_global_cond: true
+        obs_as_local_cond: false
+        obs_dim: 9
+        pred_action_steps_only: false
+        reweight_llm_loss: true
+pred_action_steps_only:
+    value: false
+task:
+    value:
+        action_dim: 4
+        dataset:
+            _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+            data_path: datasets/sweep-v2-mix(csp5).pt
+            data_path2: datasets/sweep-v2.pt
+            dummy_normalizer: false
+            horizon: 1
+            obs_eef_target: true
+            pad_after: 0
+            pad_before: 0
+            use_manual_normalizer: false
+            val_ratio: 0.1
+        env_runner:
+            _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+            discount: 0.9
+            env_name: llf-metaworld-sweep-v2
+            feedback_type:
+                - hp
+                - hn
+                - fp
+            instruction_type: b
+            max_steps: 30
+            n_action_steps: 1
+            n_envs: 10
+            n_obs_steps: 1
+            n_test: 50
+            n_train: 10
+            visual: false
+        instructor:
+            _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+        name: sweep-v2
+        obs_dim: 9
+task_name:
+    value: sweep-v2
+training:
+    value:
+        checkpoint_every: 5
+        debug: false
+        device: cuda:0
+        gradient_accumulate_every: 16
+        lr_scheduler: cosine
+        lr_warmup_steps: 500
+        max_train_steps: null
+        max_val_steps: null
+        num_epochs: 1001
+        resume: true
+        rollout_every: 5
+        sample_every: 5
+        seed: 42
+        tqdm_interval_sec: 1
+        use_ema: true
+        val_every: 1
+val_dataloader:
+    value:
+        batch_size: 16
+        num_workers: 0
+        persistent_workers: false
+        pin_memory: false
+        shuffle: false

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/files/output.log

@@ -0,0 +1,72 @@
+Traceback (most recent call last):                                                                                                                                                                        
+  File "/workspace/LLM-BC/./train.py", line 35, in <module>
+    main()
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/main.py", line 90, in decorated_main
+    _run_hydra(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 389, in _run_hydra
+    _run_app(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 452, in _run_app
+    run_and_report(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 213, in run_and_report
+    return func()
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 453, in <lambda>
+    lambda: hydra.run(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/hydra.py", line 119, in run
+    ret = run_job(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/core/utils.py", line 186, in run_job
+    ret.return_value = task_function(task_cfg)
+  File "/workspace/LLM-BC/./train.py", line 32, in main
+    workspace.run()
+  File "/workspace/LLM-BC/llmbc/workspace/train_llm_diffusion_unet_lowdim_workspace.py", line 252, in run
+    runner_log = env_runner.run(policy)
+  File "/workspace/LLM-BC/llmbc/env_runner/metaworld_lowdim_runner.py", line 113, in run
+    self.env = AsyncVectorEnv(self.env_fns)
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 168, in __init__
+    self._check_observation_spaces()
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 374, in _check_observation_spaces
+    same_spaces, successes = zip(*[pipe.recv() for pipe in self.parent_pipes])
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 374, in <listcomp>
+    same_spaces, successes = zip(*[pipe.recv() for pipe in self.parent_pipes])
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 250, in recv
+    buf = self._recv_bytes()
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 414, in _recv_bytes
+    buf = self._recv(4)
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 379, in _recv
+    chunk = read(handle, remaining)
+KeyboardInterrupt
+Traceback (most recent call last):
+  File "/workspace/LLM-BC/./train.py", line 35, in <module>
+    main()
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/main.py", line 90, in decorated_main
+    _run_hydra(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 389, in _run_hydra
+    _run_app(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 452, in _run_app
+    run_and_report(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 213, in run_and_report
+    return func()
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/utils.py", line 453, in <lambda>
+    lambda: hydra.run(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/_internal/hydra.py", line 119, in run
+    ret = run_job(
+  File "/venv/llm-bc/lib/python3.9/site-packages/hydra/core/utils.py", line 186, in run_job
+    ret.return_value = task_function(task_cfg)
+  File "/workspace/LLM-BC/./train.py", line 32, in main
+    workspace.run()
+  File "/workspace/LLM-BC/llmbc/workspace/train_llm_diffusion_unet_lowdim_workspace.py", line 252, in run
+    runner_log = env_runner.run(policy)
+  File "/workspace/LLM-BC/llmbc/env_runner/metaworld_lowdim_runner.py", line 113, in run
+    self.env = AsyncVectorEnv(self.env_fns)
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 168, in __init__
+    self._check_observation_spaces()
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 374, in _check_observation_spaces
+    same_spaces, successes = zip(*[pipe.recv() for pipe in self.parent_pipes])
+  File "/workspace/LLM-BC/llmbc/gym_util/async_vector_env.py", line 374, in <listcomp>
+    same_spaces, successes = zip(*[pipe.recv() for pipe in self.parent_pipes])
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 250, in recv
+    buf = self._recv_bytes()
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 414, in _recv_bytes
+    buf = self._recv(4)
+  File "/venv/llm-bc/lib/python3.9/multiprocessing/connection.py", line 379, in _recv
+    chunk = read(handle, remaining)
+KeyboardInterrupt

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/files/requirements.txt

@@ -0,0 +1,864 @@
+aiohappyeyeballs==2.6.1
+antlr4-python3-runtime==4.9.3
+asciitree==0.3.3
+async-timeout==5.0.1
+attrs==25.4.0
+certifi==2025.10.5
+charset-normalizer==3.4.4
+colorama==0.4.6
+dill==0.3.8
+docstring-parser==0.16
+eval_type_backport==0.2.2
+fasteners==0.20
+filelock==3.17.0
+frozenlist==1.8.0
+fsspec==2024.3.1
+gmpy2==2.2.1
+hf-xet==1.1.8
+idna==3.11
+llvmlite==0.39.1
+MarkupSafe==3.0.2
+mdurl==0.1.2
+mkl-service==2.4.0
+mpmath==1.3.0
+msgpack==1.1.1
+networkx==3.2.1
+packaging==25.0
+propcache==0.3.1
+psutil==7.0.0
+pycparser==2.23
+Pygments==2.19.1
+PyOpenGL==3.1.10
+PySocks==1.7.1
+tzdata==2025.3
+xxhash==3.5.0
+pytz==2025.2
+PyYAML==6.0.2
+regex==2025.9.1
+safetensors==0.5.3
+setuptools==80.9.0
+shtab==1.7.2
+six==1.17.0
+tqdm==4.67.1
+tensorboard==2.15.2
+uvloop==0.22.1
+wheel==0.45.1
+zipp==3.21.0
+tensorflow-estimator==2.15.0
+tensorboard-data-server==0.7.2
+google-auth-oauthlib==1.3.1
+tensorflow==2.15.1
+Jinja2==3.1.6
+markdown-it-py==2.2.0
+multiprocess==0.70.15
+numpy==1.23.5
+pillow==11.3.0
+pip==23.3.2
+python-dateutil==2.9.0.post0
+aiosignal==1.4.0
+brotlicffi==1.0.9.2
+hydra-core==1.2.0
+multidict==6.7.0
+pycares==4.10.0
+rich==14.2.0
+sympy==1.14.0
+typeguard==4.4.4
+aiodns==3.5.0
+urllib3==2.5.0
+yarl==1.18.0
+aiohttp==3.12.15
+pyarrow==21.0.0
+pyarrow-hotfix==0.7
+tokenizers==0.21.0
+av==10.0.0
+einops==0.4.1
+Bottleneck==1.4.2
+mkl_fft==1.3.11
+mkl_random==1.2.8
+numba==0.56.4
+numcodecs==0.12.1
+numexpr==2.10.1
+pandas==2.3.3
+zarr==2.12.0
+datasets==2.19.0
+transformers==4.47.1
+accelerate==1.0.1
+peft==0.14.0
+trl==0.11.4
+torch==2.2.2
+tabulate==0.9.0
+sentencepiece==0.2.0
+pyperclip==1.11.0
+PyOpenGL-accelerate==3.1.10
+pydub==0.25.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+nvidia-ml-py==13.595.45
+hjson==3.1.0
+gym-notices==0.1.0
+glfw==2.0.0
+Farama-Notifications==0.0.4
+websockets==15.0.1
+wcwidth==0.6.0
+uc-micro-py==1.0.3
+typing-inspection==0.4.2
+triton==2.2.0
+trimesh==4.11.5
+transforms3d==0.4.2
+traitlets==5.14.3
+threadpoolctl==3.6.0
+tenacity==9.1.2
+sniffio==1.3.1
+smmap==5.0.3
+setproctitle==1.3.7
+sentry-sdk==2.57.0
+semantic-version==2.10.0
+scipy==1.13.1
+rpds-py==0.27.1
+python-multipart==0.0.20
+pyparsing==3.3.2
+pynvml==13.0.1
+pydantic_core==2.41.5
+pycryptodomex==3.23.0
+pyasn1==0.6.3
+grpcio==1.80.0
+platformdirs==4.4.0
+pexpect==4.9.0
+parso==0.8.6
+orjson==3.11.5
+opencv-python==4.11.0.86
+stack-data==0.6.3
+omegaconf==2.2.1
+nvidia-nvtx-cu12==12.1.105
+nvidia-nvjitlink-cu12==12.9.86
+nvidia-nccl-cu12==2.19.3
+nvidia-curand-cu12==10.3.2.106
+nvidia-cufft-cu12==11.0.2.54
+nvidia-cuda-runtime-cu12==12.1.105
+nvidia-cuda-nvrtc-cu12==12.1.105
+nvidia-cuda-cupti-cu12==12.1.105
+nvidia-cublas-cu12==12.1.3.1
+ninja==1.13.0
+narwhals==2.19.0
+lxml==6.0.3
+kiwisolver==1.4.7
+joblib==1.5.3
+jiter==0.13.0
+imageio-ffmpeg==0.6.0
+ImageIO==2.37.2
+h5py==3.14.0
+h11==0.16.0
+fonttools==4.60.2
+ffmpy==1.0.0
+fast_kinematics==0.2.2
+executing==2.2.1
+exceptiongroup==1.3.1
+docker-pycreds==0.4.0
+distro==1.9.0
+decorator==5.2.1
+dacite==1.9.2
+cycler==0.12.1
+contourpy==1.3.0
+cloudpickle==3.1.2
+click==8.1.8
+asttokens==3.0.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+aiofiles==25.1.0
+absl-py==2.3.1
+uvicorn==0.39.0
+tiktoken==0.8.0
+scikit-learn==1.6.1
+sapien==3.0.0b1
+referencing==0.36.2
+pydantic==2.12.5
+pyasn1_modules==0.4.2
+prompt_toolkit==3.0.52
+nvidia-cusparse-cu12==12.1.0.106
+nvidia-cudnn-cu12==8.9.2.26
+nltk==3.9.2
+mujoco==2.3.7
+mdit-py-plugins==0.3.3
+matplotlib-inline==0.2.1
+matplotlib==3.7.5
+linkify-it-py==2.0.3
+jedi==0.19.2
+huggingface-hub==0.26.2
+httpcore==1.0.9
+gymnasium==0.29.1
+google-pasta==0.2.0
+gitdb==4.0.12
+cryptography==46.0.7
+blobfile==3.0.0
+anyio==4.12.1
+tyro==0.9.1
+toppra==0.6.6
+starlette==0.49.3
+seaborn==0.13.2
+rouge_score==0.1.2
+pynndescent==0.6.0
+nvidia-cusolver-cu12==11.4.5.107
+jsonschema-specifications==2025.9.1
+ipython==8.18.1
+httpx==0.28.1
+gymnasium-robotics==1.2.0
+google-auth==2.49.2
+GitPython==3.1.46
+diffusers==0.31.0
+wandb==0.18.6
+umap-learn==0.5.9.post2
+requests-oauthlib==2.0.0
+openai==2.8.1
+mplib==0.1.1
+jsonschema==4.25.1
+gradio_client==0.2.9
+google-genai==1.47.0
+fastapi==0.128.8
+torchvision==0.17.2
+torchaudio==2.2.2
+sentence-transformers==3.2.1
+pytorch-seed==0.2.0
+deepspeed==0.16.1
+bitsandbytes==0.45.0
+altair==6.0.0
+timm==1.0.22
+gradio==3.36.1
+evaluate==0.4.3
+arm_pytorch_utilities==0.5.0
+tensorflow-io-gcs-filesystem==0.37.1
+Markdown==3.9
+pybullet==3.2.6
+wrapt==1.14.2
+tensorflow-probability==0.23.0
+parse==1.19.1
+dm-tree==0.1.8
+termcolor==3.1.0
+requests==2.32.0
+cffi==2.0.0
+opt_einsum==3.4.0
+labmaze==1.0.6
+typing_extensions==4.15.0
+importlib-resources==5.13.0
+importlib-metadata==5.2.0
+FLAML==2.3.6
+dm-env==1.6
+diskcache==5.6.3
+Cython==0.29.37
+mujoco-py==2.1.2.14
+jaxlib==0.4.30
+dm-control==1.0.14
+cmudict==1.0.13
+syllables==1.0.9
+metaworld==2.0.0
+jax==0.4.30
+highway-env==1.9.1
+gym_bandits==0.0.2
+d4rl==1.1
+pyautogen==0.1.0
+protobuf==4.25.9
+tf-agents==0.19.0
+oauthlib==3.3.1
+pytorch-kinematics==0.7.5
+pygame==2.6.1
+gin-config==0.5.0
+mani_skill==3.0.0b20
+tifffile==2024.8.30
+Shapely==1.8.4
+PyWavelets==1.6.0
+Werkzeug==3.1.8
+astunparse==1.6.3
+scikit-image==0.19.3
+pymunk==6.2.1
+ml-dtypes==0.3.2
+gast==0.7.0
+stable-baselines3==2.2.1
+keras==2.15.0
+libclang==18.1.1
+flatbuffers==25.12.19
+gym==0.23.1
+llmbc==0.0.0
+mjrl==1.0.0
+llfbench==0.1.0
+aiohappyeyeballs==2.6.1
+antlr4-python3-runtime==4.9.3
+asciitree==0.3.3
+async-timeout==5.0.1
+attrs==25.4.0
+certifi==2025.10.5
+charset-normalizer==3.4.4
+colorama==0.4.6
+dill==0.3.8
+docstring-parser==0.16
+eval_type_backport==0.2.2
+fasteners==0.20
+filelock==3.17.0
+frozenlist==1.8.0
+fsspec==2024.3.1
+gmpy2==2.2.1
+hf-xet==1.1.8
+idna==3.11
+llvmlite==0.39.1
+MarkupSafe==3.0.2
+mdurl==0.1.2
+mkl-service==2.4.0
+mpmath==1.3.0
+msgpack==1.1.1
+networkx==3.2.1
+packaging==25.0
+propcache==0.3.1
+psutil==7.0.0
+pycparser==2.23
+Pygments==2.19.1
+PyOpenGL==3.1.10
+PySocks==1.7.1
+tzdata==2025.3
+xxhash==3.5.0
+pytz==2025.2
+PyYAML==6.0.2
+regex==2025.9.1
+safetensors==0.5.3
+setuptools==80.9.0
+shtab==1.7.2
+six==1.17.0
+tqdm==4.67.1
+tensorboard==2.15.2
+uvloop==0.22.1
+wheel==0.45.1
+zipp==3.21.0
+tensorflow-estimator==2.15.0
+tensorboard-data-server==0.7.2
+google-auth-oauthlib==1.3.1
+tensorflow==2.15.1
+Jinja2==3.1.6
+markdown-it-py==2.2.0
+multiprocess==0.70.15
+numpy==1.23.5
+pillow==11.3.0
+pip==23.3.2
+python-dateutil==2.9.0.post0
+aiosignal==1.4.0
+brotlicffi==1.0.9.2
+hydra-core==1.2.0
+multidict==6.7.0
+pycares==4.10.0
+rich==14.2.0
+sympy==1.14.0
+typeguard==4.4.4
+aiodns==3.5.0
+urllib3==2.5.0
+yarl==1.18.0
+aiohttp==3.12.15
+pyarrow==21.0.0
+pyarrow-hotfix==0.7
+tokenizers==0.21.0
+av==10.0.0
+einops==0.4.1
+Bottleneck==1.4.2
+mkl_fft==1.3.11
+mkl_random==1.2.8
+numba==0.56.4
+numcodecs==0.12.1
+numexpr==2.10.1
+pandas==2.3.3
+zarr==2.12.0
+datasets==2.19.0
+transformers==4.47.1
+accelerate==1.0.1
+peft==0.14.0
+trl==0.11.4
+torch==2.2.2
+tabulate==0.9.0
+sentencepiece==0.2.0
+pyperclip==1.11.0
+PyOpenGL-accelerate==3.1.10
+pydub==0.25.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+nvidia-ml-py==13.595.45
+hjson==3.1.0
+gym-notices==0.1.0
+glfw==2.0.0
+Farama-Notifications==0.0.4
+websockets==15.0.1
+wcwidth==0.6.0
+uc-micro-py==1.0.3
+typing-inspection==0.4.2
+triton==2.2.0
+trimesh==4.11.5
+transforms3d==0.4.2
+traitlets==5.14.3
+threadpoolctl==3.6.0
+tenacity==9.1.2
+sniffio==1.3.1
+smmap==5.0.3
+setproctitle==1.3.7
+sentry-sdk==2.57.0
+semantic-version==2.10.0
+scipy==1.13.1
+rpds-py==0.27.1
+python-multipart==0.0.20
+pyparsing==3.3.2
+pynvml==13.0.1
+pydantic_core==2.41.5
+pycryptodomex==3.23.0
+pyasn1==0.6.3
+grpcio==1.80.0
+platformdirs==4.4.0
+pexpect==4.9.0
+parso==0.8.6
+orjson==3.11.5
+opencv-python==4.11.0.86
+stack-data==0.6.3
+omegaconf==2.2.1
+nvidia-nvtx-cu12==12.1.105
+nvidia-nvjitlink-cu12==12.9.86
+nvidia-nccl-cu12==2.19.3
+nvidia-curand-cu12==10.3.2.106
+nvidia-cufft-cu12==11.0.2.54
+nvidia-cuda-runtime-cu12==12.1.105
+nvidia-cuda-nvrtc-cu12==12.1.105
+nvidia-cuda-cupti-cu12==12.1.105
+nvidia-cublas-cu12==12.1.3.1
+ninja==1.13.0
+narwhals==2.19.0
+lxml==6.0.3
+kiwisolver==1.4.7
+joblib==1.5.3
+jiter==0.13.0
+imageio-ffmpeg==0.6.0
+ImageIO==2.37.2
+h5py==3.14.0
+h11==0.16.0
+fonttools==4.60.2
+ffmpy==1.0.0
+fast_kinematics==0.2.2
+executing==2.2.1
+exceptiongroup==1.3.1
+docker-pycreds==0.4.0
+distro==1.9.0
+decorator==5.2.1
+dacite==1.9.2
+cycler==0.12.1
+contourpy==1.3.0
+cloudpickle==3.1.2
+click==8.1.8
+asttokens==3.0.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+aiofiles==25.1.0
+absl-py==2.3.1
+uvicorn==0.39.0
+tiktoken==0.8.0
+scikit-learn==1.6.1
+sapien==3.0.0b1
+referencing==0.36.2
+pydantic==2.12.5
+pyasn1_modules==0.4.2
+prompt_toolkit==3.0.52
+nvidia-cusparse-cu12==12.1.0.106
+nvidia-cudnn-cu12==8.9.2.26
+nltk==3.9.2
+mujoco==2.3.7
+mdit-py-plugins==0.3.3
+matplotlib-inline==0.2.1
+matplotlib==3.7.5
+linkify-it-py==2.0.3
+jedi==0.19.2
+huggingface-hub==0.26.2
+httpcore==1.0.9
+gymnasium==0.29.1
+google-pasta==0.2.0
+gitdb==4.0.12
+cryptography==46.0.7
+blobfile==3.0.0
+anyio==4.12.1
+tyro==0.9.1
+toppra==0.6.6
+starlette==0.49.3
+seaborn==0.13.2
+rouge_score==0.1.2
+pynndescent==0.6.0
+nvidia-cusolver-cu12==11.4.5.107
+jsonschema-specifications==2025.9.1
+ipython==8.18.1
+httpx==0.28.1
+gymnasium-robotics==1.2.0
+google-auth==2.49.2
+GitPython==3.1.46
+diffusers==0.31.0
+wandb==0.18.6
+umap-learn==0.5.9.post2
+requests-oauthlib==2.0.0
+openai==2.8.1
+mplib==0.1.1
+jsonschema==4.25.1
+gradio_client==0.2.9
+google-genai==1.47.0
+fastapi==0.128.8
+torchvision==0.17.2
+torchaudio==2.2.2
+sentence-transformers==3.2.1
+pytorch-seed==0.2.0
+deepspeed==0.16.1
+bitsandbytes==0.45.0
+altair==6.0.0
+timm==1.0.22
+gradio==3.36.1
+evaluate==0.4.3
+arm_pytorch_utilities==0.5.0
+tensorflow-io-gcs-filesystem==0.37.1
+Markdown==3.9
+pybullet==3.2.6
+wrapt==1.14.2
+tensorflow-probability==0.23.0
+parse==1.19.1
+dm-tree==0.1.8
+termcolor==3.1.0
+requests==2.32.0
+cffi==2.0.0
+opt_einsum==3.4.0
+labmaze==1.0.6
+typing_extensions==4.15.0
+importlib-resources==5.13.0
+importlib-metadata==5.2.0
+FLAML==2.3.6
+dm-env==1.6
+diskcache==5.6.3
+Cython==0.29.37
+mujoco-py==2.1.2.14
+jaxlib==0.4.30
+dm-control==1.0.14
+cmudict==1.0.13
+syllables==1.0.9
+metaworld==2.0.0
+jax==0.4.30
+highway-env==1.9.1
+gym_bandits==0.0.2
+d4rl==1.1
+pyautogen==0.1.0
+protobuf==4.25.9
+tf-agents==0.19.0
+oauthlib==3.3.1
+pytorch-kinematics==0.7.5
+pygame==2.6.1
+gin-config==0.5.0
+mani_skill==3.0.0b20
+tifffile==2024.8.30
+Shapely==1.8.4
+PyWavelets==1.6.0
+Werkzeug==3.1.8
+astunparse==1.6.3
+scikit-image==0.19.3
+pymunk==6.2.1
+ml-dtypes==0.3.2
+gast==0.7.0
+stable-baselines3==2.2.1
+keras==2.15.0
+libclang==18.1.1
+flatbuffers==25.12.19
+gym==0.23.1
+llmbc==0.0.0
+mjrl==1.0.0
+llfbench==0.1.0
+llmbc==0.0.0
+llmbc==0.0.0
+aiohappyeyeballs==2.6.1
+antlr4-python3-runtime==4.9.3
+asciitree==0.3.3
+async-timeout==5.0.1
+attrs==25.4.0
+certifi==2025.10.5
+charset-normalizer==3.4.4
+colorama==0.4.6
+dill==0.3.8
+docstring-parser==0.16
+eval_type_backport==0.2.2
+fasteners==0.20
+filelock==3.17.0
+frozenlist==1.8.0
+fsspec==2024.3.1
+gmpy2==2.2.1
+hf-xet==1.1.8
+idna==3.11
+llvmlite==0.39.1
+MarkupSafe==3.0.2
+mdurl==0.1.2
+mkl-service==2.4.0
+mpmath==1.3.0
+msgpack==1.1.1
+networkx==3.2.1
+packaging==25.0
+propcache==0.3.1
+psutil==7.0.0
+pycparser==2.23
+Pygments==2.19.1
+PyOpenGL==3.1.10
+PySocks==1.7.1
+tzdata==2025.3
+xxhash==3.5.0
+pytz==2025.2
+PyYAML==6.0.2
+regex==2025.9.1
+safetensors==0.5.3
+setuptools==80.9.0
+shtab==1.7.2
+six==1.17.0
+tqdm==4.67.1
+tensorboard==2.15.2
+uvloop==0.22.1
+wheel==0.45.1
+zipp==3.21.0
+tensorflow-estimator==2.15.0
+tensorboard-data-server==0.7.2
+google-auth-oauthlib==1.3.1
+tensorflow==2.15.1
+Jinja2==3.1.6
+markdown-it-py==2.2.0
+multiprocess==0.70.15
+numpy==1.23.5
+pillow==11.3.0
+pip==23.3.2
+python-dateutil==2.9.0.post0
+aiosignal==1.4.0
+brotlicffi==1.0.9.2
+hydra-core==1.2.0
+multidict==6.7.0
+pycares==4.10.0
+rich==14.2.0
+sympy==1.14.0
+typeguard==4.4.4
+aiodns==3.5.0
+urllib3==2.5.0
+yarl==1.18.0
+aiohttp==3.12.15
+pyarrow==21.0.0
+pyarrow-hotfix==0.7
+tokenizers==0.21.0
+av==10.0.0
+einops==0.4.1
+Bottleneck==1.4.2
+mkl_fft==1.3.11
+mkl_random==1.2.8
+numba==0.56.4
+numcodecs==0.12.1
+numexpr==2.10.1
+pandas==2.3.3
+zarr==2.12.0
+datasets==2.19.0
+transformers==4.47.1
+accelerate==1.0.1
+peft==0.14.0
+trl==0.11.4
+torch==2.2.2
+tabulate==0.9.0
+sentencepiece==0.2.0
+pyperclip==1.11.0
+PyOpenGL-accelerate==3.1.10
+pydub==0.25.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+nvidia-ml-py==13.595.45
+hjson==3.1.0
+gym-notices==0.1.0
+glfw==2.0.0
+Farama-Notifications==0.0.4
+websockets==15.0.1
+wcwidth==0.6.0
+uc-micro-py==1.0.3
+typing-inspection==0.4.2
+triton==2.2.0
+trimesh==4.11.5
+transforms3d==0.4.2
+traitlets==5.14.3
+threadpoolctl==3.6.0
+tenacity==9.1.2
+sniffio==1.3.1
+smmap==5.0.3
+setproctitle==1.3.7
+sentry-sdk==2.57.0
+semantic-version==2.10.0
+scipy==1.13.1
+rpds-py==0.27.1
+python-multipart==0.0.20
+pyparsing==3.3.2
+pynvml==13.0.1
+pydantic_core==2.41.5
+pycryptodomex==3.23.0
+pyasn1==0.6.3
+grpcio==1.80.0
+platformdirs==4.4.0
+pexpect==4.9.0
+parso==0.8.6
+orjson==3.11.5
+opencv-python==4.11.0.86
+stack-data==0.6.3
+omegaconf==2.2.1
+nvidia-nvtx-cu12==12.1.105
+nvidia-nvjitlink-cu12==12.9.86
+nvidia-nccl-cu12==2.19.3
+nvidia-curand-cu12==10.3.2.106
+nvidia-cufft-cu12==11.0.2.54
+nvidia-cuda-runtime-cu12==12.1.105
+nvidia-cuda-nvrtc-cu12==12.1.105
+nvidia-cuda-cupti-cu12==12.1.105
+nvidia-cublas-cu12==12.1.3.1
+ninja==1.13.0
+narwhals==2.19.0
+lxml==6.0.3
+kiwisolver==1.4.7
+joblib==1.5.3
+jiter==0.13.0
+imageio-ffmpeg==0.6.0
+ImageIO==2.37.2
+h5py==3.14.0
+h11==0.16.0
+fonttools==4.60.2
+ffmpy==1.0.0
+fast_kinematics==0.2.2
+executing==2.2.1
+exceptiongroup==1.3.1
+docker-pycreds==0.4.0
+distro==1.9.0
+decorator==5.2.1
+dacite==1.9.2
+cycler==0.12.1
+contourpy==1.3.0
+cloudpickle==3.1.2
+click==8.1.8
+asttokens==3.0.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+aiofiles==25.1.0
+absl-py==2.3.1
+uvicorn==0.39.0
+tiktoken==0.8.0
+scikit-learn==1.6.1
+sapien==3.0.0b1
+referencing==0.36.2
+pydantic==2.12.5
+pyasn1_modules==0.4.2
+prompt_toolkit==3.0.52
+nvidia-cusparse-cu12==12.1.0.106
+nvidia-cudnn-cu12==8.9.2.26
+nltk==3.9.2
+mujoco==2.3.7
+mdit-py-plugins==0.3.3
+matplotlib-inline==0.2.1
+matplotlib==3.7.5
+linkify-it-py==2.0.3
+jedi==0.19.2
+huggingface-hub==0.26.2
+httpcore==1.0.9
+gymnasium==0.29.1
+google-pasta==0.2.0
+gitdb==4.0.12
+cryptography==46.0.7
+blobfile==3.0.0
+anyio==4.12.1
+tyro==0.9.1
+toppra==0.6.6
+starlette==0.49.3
+seaborn==0.13.2
+rouge_score==0.1.2
+pynndescent==0.6.0
+nvidia-cusolver-cu12==11.4.5.107
+jsonschema-specifications==2025.9.1
+ipython==8.18.1
+httpx==0.28.1
+gymnasium-robotics==1.2.0
+google-auth==2.49.2
+GitPython==3.1.46
+diffusers==0.31.0
+wandb==0.18.6
+umap-learn==0.5.9.post2
+requests-oauthlib==2.0.0
+openai==2.8.1
+mplib==0.1.1
+jsonschema==4.25.1
+gradio_client==0.2.9
+google-genai==1.47.0
+fastapi==0.128.8
+torchvision==0.17.2
+torchaudio==2.2.2
+sentence-transformers==3.2.1
+pytorch-seed==0.2.0
+deepspeed==0.16.1
+bitsandbytes==0.45.0
+altair==6.0.0
+timm==1.0.22
+gradio==3.36.1
+evaluate==0.4.3
+arm_pytorch_utilities==0.5.0
+tensorflow-io-gcs-filesystem==0.37.1
+Markdown==3.9
+pybullet==3.2.6
+wrapt==1.14.2
+tensorflow-probability==0.23.0
+parse==1.19.1
+dm-tree==0.1.8
+termcolor==3.1.0
+requests==2.32.0
+cffi==2.0.0
+opt_einsum==3.4.0
+labmaze==1.0.6
+typing_extensions==4.15.0
+importlib-resources==5.13.0
+importlib-metadata==5.2.0
+FLAML==2.3.6
+dm-env==1.6
+diskcache==5.6.3
+Cython==0.29.37
+mujoco-py==2.1.2.14
+jaxlib==0.4.30
+dm-control==1.0.14
+cmudict==1.0.13
+syllables==1.0.9
+metaworld==2.0.0
+jax==0.4.30
+highway-env==1.9.1
+gym_bandits==0.0.2
+d4rl==1.1
+pyautogen==0.1.0
+protobuf==4.25.9
+tf-agents==0.19.0
+oauthlib==3.3.1
+pytorch-kinematics==0.7.5
+pygame==2.6.1
+gin-config==0.5.0
+mani_skill==3.0.0b20
+tifffile==2024.8.30
+Shapely==1.8.4
+PyWavelets==1.6.0
+Werkzeug==3.1.8
+astunparse==1.6.3
+scikit-image==0.19.3
+pymunk==6.2.1
+ml-dtypes==0.3.2
+gast==0.7.0
+stable-baselines3==2.2.1
+keras==2.15.0
+libclang==18.1.1
+flatbuffers==25.12.19
+gym==0.23.1
+llmbc==0.0.0
+mjrl==1.0.0
+llfbench==0.1.0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/files/wandb-metadata.json

@@ -0,0 +1,56 @@
+{
+  "os": "Linux-5.15.0-168-generic-x86_64-with-glibc2.39",
+  "python": "3.9.25",
+  "startedAt": "2026-04-14T06:26:28.532096Z",
+  "args": [
+    "--config-path",
+    "config/mix_data",
+    "--config-name",
+    "llmdp_sweep-v2_mix(csp5).yaml",
+    "policy.loss_llm_weight=1.0e-1"
+  ],
+  "program": "/workspace/LLM-BC/./train.py",
+  "codePath": "train.py",
+  "git": {
+    "remote": "https://github.com/CHYang25/LLM-BC.git",
+    "commit": "16b534f5ac2111b9cbfa638ab62d7944bff711d2"
+  },
+  "email": "chris920325@gmail.com",
+  "root": "/workspace/LLM-BC/data/outputs/2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2",
+  "host": "c8ef52649900",
+  "username": "root",
+  "executable": "/venv/llm-bc/bin/python3",
+  "codePathLocal": "train.py",
+  "cpu_count": 72,
+  "cpu_count_logical": 144,
+  "gpu": "NVIDIA GeForce RTX 4090",
+  "gpu_count": 2,
+  "disk": {
+    "/": {
+      "total": "545460846592",
+      "used": "352992382976"
+    }
+  },
+  "memory": {
+    "total": "810942898176"
+  },
+  "cpu": {
+    "count": 72,
+    "countLogical": 144
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA GeForce RTX 4090",
+      "memoryTotal": "25757220864",
+      "cudaCores": 16384,
+      "architecture": "Ada"
+    },
+    {
+      "name": "NVIDIA GeForce RTX 4090",
+      "memoryTotal": "25757220864",
+      "cudaCores": 16384,
+      "architecture": "Ada"
+    }
+  ],
+  "cudaVersion": "12.8"
+}

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/files/wandb-summary.json

@@ -0,0 +1 @@
+{"_wandb":{"runtime":213},"lr":7.000000000000001e-06,"_runtime":213.817319835,"_step":550,"train_loss_dp":0.7344393730163574,"global_step":550,"epoch":0,"_timestamp":1.7761481617795644e+09,"train_loss":0.8508371114730835,"train_loss_llm":1.1639776229858398}

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/run-20260414_062628-981b5iwy/run-981b5iwy.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0e61bd891da7f59f0cae5d086053a23d4eea0080b8d39b73c79790f25250279
+size 491520

2026.04.14/06.25.31_train_llm_diffusion_unet_lowdim_sweep-v2/wandb/wandb-resume.json

@@ -0,0 +1 @@
+{"run_id": "981b5iwy"}

2026.04.14/06.25.35_train_llmbc_lowdim_maze2d-medium-v0/.hydra/config.yaml

@@ -0,0 +1,165 @@
+name: train_llmbc_lowdim
+_target_: llmbc.workspace.train_llmbc_lowdim_workspace.TrainLLMBCLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llmbc_lowdim_policy.LLMBCLowdimPolicy
+  model:
+    _target_: llmbc.model.policy.policy_mlp.PolicyMLP
+    input_size: ${eval:'${n_obs_steps}*${obs_dim}'}
+    hidden_size:
+    - 256
+    - 256
+    - 256
+    - 256
+    output_size: ${eval:'${n_action_steps}*${action_dim}'}
+    activation: tanh
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${task.obs_dim}
+      action_dim: ${task.action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${n_action_steps}
+  loss_bc_weight: 1.0
+  loss_llm_weight: 0.01
+  horizon: ${horizon}
+  n_obs_steps: ${n_obs_steps}
+  n_action_steps: ${n_action_steps}
+  normalize_llm_loss: true
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 43
+  debug: false
+  resume: false
+  lr_scheduler: cosine
+  lr_warmup_steps: 10
+  num_epochs: 801
+  gradient_accumulate_every: 8
+  grad_norm_clip: 0.5
+  rollout_every: 4
+  checkpoint_every: 4
+  val_every: 1
+  sample_every: 4
+  sample_max_batch: 128
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training-mix
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: maze2d-medium-v0
+  obs_dim: 6
+  action_dim: 2
+  env_runner:
+    _target_: llmbc.env_runner.pointmaze_lowdim_runner.PointmazeLowdimRunner
+    env_name: llf-pointmaze-maze2d-medium-v0
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 250
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.99
+  dataset:
+    _target_: llmbc.dataset.pointmaze_lowdim_dataset.PointmazeLowdimDataset
+    data_path: datasets/maze2d-medium-v0-mix(csp5).pt
+    data_path2: datasets/maze2d-medium-v0.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.02
+    dummy_normalizer: true
+  instructor:
+    _target_: llmbc.translator.instructor.pointmaze_instructor.maze2d_medium_v0_instructor.Maze2dMediumV0Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.10.16/11.53.22_train_llm_lowdim_maze2d-medium-v0/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-maze2d-medium-v0/checkpoint-11850
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.25.35_train_llmbc_lowdim_maze2d-medium-v0/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - training.seed=43
+  job:
+    name: train
+    chdir: null
+    override_dirname: training.seed=43
+    id: ???
+    num: ???
+    config_name: llmbc_maze2d-medium-v0_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.25.35_train_llmbc_lowdim_maze2d-medium-v0
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.25.35_train_llmbc_lowdim_maze2d-medium-v0/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- training.seed=43

2026.04.14/06.25.35_train_llmbc_lowdim_maze2d-medium-v0/train.log

@@ -0,0 +1,6 @@
+[2026-04-14 06:25:36,871][numexpr.utils][INFO] - Note: detected 144 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable.
+[2026-04-14 06:25:36,871][numexpr.utils][INFO] - Note: NumExpr detected 144 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 16.
+[2026-04-14 06:25:36,871][numexpr.utils][INFO] - NumExpr defaulting to 16 threads.
+[2026-04-14 06:25:40,671][datasets][INFO] - PyTorch version 2.2.2 available.
+[2026-04-14 06:25:40,672][datasets][INFO] - TensorFlow version 2.15.1 available.
+[2026-04-14 06:25:40,673][datasets][INFO] - JAX version 0.4.30 available.

2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/config.yaml

@@ -0,0 +1,193 @@
+name: train_llm_diffusion_unet_lowdim
+_target_: llmbc.workspace.train_llm_diffusion_unet_lowdim_workspace.TrainLLMDiffusionUnetLowdimWorkspace
+obs_dim: ${task.obs_dim}
+action_dim: ${task.action_dim}
+task_name: ${task.name}
+exp_name: default
+model_name: ${llm.name}
+horizon: 1
+n_obs_steps: 1
+n_action_steps: 1
+n_latency_steps: 0
+past_action_visible: false
+keypoint_visible_rate: 1.0
+obs_as_local_cond: false
+obs_as_global_cond: true
+pred_action_steps_only: false
+llm_orig_expert_feedback: true
+llm_do_sample: false
+policy:
+  _target_: llmbc.policy.llm_diffusion_unet_lowdim_policy.LLMDiffusionUnetLowdimPolicy
+  model:
+    _target_: llmbc.model.diffusion.conditional_unet1d.ConditionalUnet1D
+    input_dim: '${eval: ${task.action_dim} if ${obs_as_local_cond} or ${obs_as_global_cond}
+      else ${task.obs_dim} + ${task.action_dim}}'
+    local_cond_dim: '${eval: ${task.obs_dim} if ${obs_as_local_cond} else None}'
+    global_cond_dim: '${eval: ${task.obs_dim}*${n_obs_steps} if ${obs_as_global_cond}
+      else None}'
+    diffusion_step_embed_dim: 256
+    down_dims:
+    - 256
+    - 512
+    - 1024
+    kernel_size: 5
+    n_groups: 8
+    cond_predict_scale: true
+  noise_scheduler:
+    _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
+    num_train_timesteps: 100
+    beta_start: 0.0001
+    beta_end: 0.02
+    beta_schedule: squaredcos_cap_v2
+    variance_type: fixed_small
+    clip_sample: true
+    prediction_type: epsilon
+  horizon: ${horizon}
+  obs_dim: ${obs_dim}
+  action_dim: ${action_dim}
+  n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  n_obs_steps: ${n_obs_steps}
+  num_inference_steps: 100
+  obs_as_local_cond: ${obs_as_local_cond}
+  obs_as_global_cond: ${obs_as_global_cond}
+  pred_action_steps_only: ${pred_action_steps_only}
+  oa_step_convention: true
+  llm_discriminator:
+    _target_: llmbc.discriminator.llm_ce_discriminator.LLMCEDiscriminator
+    task_id: ${task_name}
+    llm_translator:
+      _target_: llmbc.translator.llm_translator.LLMTranslator
+      cfg: ${llm}
+      obs_dim: ${obs_dim}
+      action_dim: ${action_dim}
+      horizon: ${horizon}
+      n_obs_steps: ${n_obs_steps}
+      n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
+  loss_dp_weight: 1.0
+  loss_llm_weight: 0.01
+  normalize_llm_loss: true
+  reweight_llm_loss: true
+ema:
+  _target_: llmbc.model.diffusion.ema_model.EMAModel
+  update_after_step: 0
+  inv_gamma: 1.0
+  power: 0.75
+  min_value: 0.0
+  max_value: 0.9999
+dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: true
+  pin_memory: false
+  persistent_workers: false
+val_dataloader:
+  batch_size: 16
+  num_workers: 0
+  shuffle: false
+  pin_memory: false
+  persistent_workers: false
+optimizer:
+  _target_: torch.optim.AdamW
+  lr: 0.0001
+  betas:
+  - 0.95
+  - 0.999
+  eps: 1.0e-08
+  weight_decay: 1.0e-06
+training:
+  device: cuda:0
+  seed: 42
+  debug: false
+  resume: true
+  lr_scheduler: cosine
+  lr_warmup_steps: 500
+  num_epochs: 1001
+  gradient_accumulate_every: 16
+  use_ema: true
+  rollout_every: 5
+  checkpoint_every: 5
+  val_every: 1
+  sample_every: 5
+  max_train_steps: null
+  max_val_steps: null
+  tqdm_interval_sec: 1.0
+logging:
+  project: ${task.name}-training
+  resume: true
+  mode: online
+  name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+  tags:
+  - ${name}
+  - ${task_name}
+  - ${exp_name}
+  id: null
+  group: null
+checkpoint:
+  topk:
+    monitor_key: test_success_rate
+    mode: max
+    k: 5
+    format_str: epoch={epoch:04d}-test_success_rate={test_success_rate:.3f}.ckpt
+  save_last_ckpt: true
+  save_last_snapshot: false
+multi_run:
+  run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
+task:
+  name: sweep-v2
+  obs_dim: 9
+  action_dim: 4
+  env_runner:
+    _target_: llmbc.env_runner.metaworld_lowdim_runner.MetaworldLowdimRunner
+    env_name: llf-metaworld-sweep-v2
+    n_train: 10
+    n_test: 50
+    n_envs: 10
+    max_steps: 30
+    n_obs_steps: ${n_obs_steps}
+    n_action_steps: ${n_action_steps}
+    instruction_type: b
+    feedback_type:
+    - hp
+    - hn
+    - fp
+    visual: false
+    discount: 0.9
+  dataset:
+    _target_: llmbc.dataset.metaworld_lowdim_dataset.MetaworldLowdimDataset
+    data_path: datasets/sweep-v2-mix(csp5).pt
+    data_path2: datasets/sweep-v2.pt
+    horizon: ${horizon}
+    pad_before: ${eval:'${n_obs_steps}-1'}
+    pad_after: ${eval:'${n_action_steps}-1'}
+    obs_eef_target: true
+    use_manual_normalizer: false
+    val_ratio: 0.1
+    dummy_normalizer: false
+  instructor:
+    _target_: llmbc.translator.instructor.metaworld_instructor.sweep_v2_instructor.SweepV2Instructor
+llm:
+  name: HuggingFaceTB/SmolLM2-135M-Instruct
+  model_name: SmolLM2-135M-Instruct
+  config_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaConfig
+  causal_lm_target: llmbc.model.llm.llama_lowdim_model.LowdimLlamaForCausalLM
+  use_quantization: false
+  use_joint_mlp_projector: true
+  llm_mode: ete-finetuned
+  finetune_mode: orig
+  checkpoint: data/outputs/2025.11.01/23.03.01_train_llm_lowdim_sweep-v2/HuggingFaceTB/SmolLM2-135M-Instruct-finetuned-sweep-v2/checkpoint-5690
+  max_length: 100
+  lora_config:
+    r: 32
+    lora_alpha: 64
+    lora_dropout: 0.05
+    bias: none
+    task_type: CAUSAL_LM
+  prompter:
+    _target_: llmbc.translator.prompter.smollm2_prompter.SmolLM2Prompter
+    use_joint_mlp_projector: true
+  hydra:
+    job:
+      override_dirname: ${model_name}
+    run:
+      dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${model_name}

2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/hydra.yaml

@@ -0,0 +1,155 @@
+hydra:
+  run:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+  sweep:
+    dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
+    subdir: ${hydra.job.num}
+  launcher:
+    _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - policy.loss_llm_weight=1.0e-2
+  job:
+    name: train
+    chdir: null
+    override_dirname: policy.loss_llm_weight=1.0e-2
+    id: ???
+    num: ???
+    config_name: llmdp_sweep-v2_mix(csp5).yaml
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.2.0
+    version_base: '1.2'
+    cwd: /workspace/LLM-BC
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /workspace/LLM-BC/config/mix_data
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /workspace/LLM-BC/data/outputs/2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2
+    choices:
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- policy.loss_llm_weight=1.0e-2

2026.04.14/06.26.20_train_llm_diffusion_unet_lowdim_sweep-v2/logs.json.txt

@@ -0,0 +1,350 @@
+{"train_loss": 1.2142260074615479, "train_loss_dp": 1.2109837532043457, "train_loss_llm": 0.32422637939453125, "global_step": 0, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.5894043445587158, "train_loss_dp": 1.5555988550186157, "train_loss_llm": 3.3805484771728516, "global_step": 1, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 0.8592647910118103, "train_loss_dp": 0.8444761037826538, "train_loss_llm": 1.4788703918457031, "global_step": 2, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.5737547874450684, "train_loss_dp": 1.567467212677002, "train_loss_llm": 0.6287602186203003, "global_step": 3, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2232937812805176, "train_loss_dp": 1.2161624431610107, "train_loss_llm": 0.7131385803222656, "global_step": 4, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.611609935760498, "train_loss_dp": 1.6041865348815918, "train_loss_llm": 0.7423383593559265, "global_step": 5, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.3413699865341187, "train_loss_dp": 1.3234602212905884, "train_loss_llm": 1.7909822463989258, "global_step": 6, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.193677544593811, "train_loss_dp": 1.1930253505706787, "train_loss_llm": 0.06521368026733398, "global_step": 7, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.0304384231567383, "train_loss_dp": 1.006334900856018, "train_loss_llm": 2.410350799560547, "global_step": 8, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.3652852773666382, "train_loss_dp": 1.3523192405700684, "train_loss_llm": 1.2966029644012451, "global_step": 9, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.381680965423584, "train_loss_dp": 1.3722383975982666, "train_loss_llm": 0.9442574977874756, "global_step": 10, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2500667572021484, "train_loss_dp": 1.2402263879776, "train_loss_llm": 0.9840337038040161, "global_step": 11, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.2191638946533203, "train_loss_dp": 1.2112371921539307, "train_loss_llm": 0.792670726776123, "global_step": 12, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.1264299154281616, "train_loss_dp": 1.1232653856277466, "train_loss_llm": 0.3164515495300293, "global_step": 13, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.114905834197998, "train_loss_dp": 1.1038504838943481, "train_loss_llm": 1.1055290699005127, "global_step": 14, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 1.5273246765136719, "train_loss_dp": 1.5052738189697266, "train_loss_llm": 2.205087661743164, "global_step": 15, "epoch": 0, "lr": 2.0000000000000002e-07}
+{"train_loss": 0.8774579167366028, "train_loss_dp": 0.8685603141784668, "train_loss_llm": 0.8897576332092285, "global_step": 16, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.242292881011963, "train_loss_dp": 1.2375354766845703, "train_loss_llm": 0.4757366180419922, "global_step": 17, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 0.9952828288078308, "train_loss_dp": 0.9838780164718628, "train_loss_llm": 1.1404805183410645, "global_step": 18, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.2600748538970947, "train_loss_dp": 1.2529277801513672, "train_loss_llm": 0.7147130966186523, "global_step": 19, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.1880123615264893, "train_loss_dp": 1.1880123615264893, "train_loss_llm": 0.0, "global_step": 20, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.3708230257034302, "train_loss_dp": 1.3630808591842651, "train_loss_llm": 0.774214506149292, "global_step": 21, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.1187998056411743, "train_loss_dp": 1.108497977256775, "train_loss_llm": 1.0301799774169922, "global_step": 22, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.1816767454147339, "train_loss_dp": 1.1755201816558838, "train_loss_llm": 0.6156516075134277, "global_step": 23, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.3325159549713135, "train_loss_dp": 1.3082664012908936, "train_loss_llm": 2.424954414367676, "global_step": 24, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.4005565643310547, "train_loss_dp": 1.3876872062683105, "train_loss_llm": 1.2869410514831543, "global_step": 25, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.2212674617767334, "train_loss_dp": 1.2060461044311523, "train_loss_llm": 1.5221405029296875, "global_step": 26, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 0.8301150798797607, "train_loss_dp": 0.8227673768997192, "train_loss_llm": 0.7347679138183594, "global_step": 27, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 0.9149899482727051, "train_loss_dp": 0.9067434072494507, "train_loss_llm": 0.8246545791625977, "global_step": 28, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.3611483573913574, "train_loss_dp": 1.2447832822799683, "train_loss_llm": 11.636505126953125, "global_step": 29, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.1991043090820312, "train_loss_dp": 1.1810945272445679, "train_loss_llm": 1.8009734153747559, "global_step": 30, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.39414644241333, "train_loss_dp": 1.3838207721710205, "train_loss_llm": 1.0325665473937988, "global_step": 31, "epoch": 0, "lr": 4.0000000000000003e-07}
+{"train_loss": 1.232168436050415, "train_loss_dp": 0.9743912220001221, "train_loss_llm": 25.777721405029297, "global_step": 32, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0207141637802124, "train_loss_dp": 1.0061120986938477, "train_loss_llm": 1.4602069854736328, "global_step": 33, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.224158763885498, "train_loss_dp": 1.1038836240768433, "train_loss_llm": 12.02751350402832, "global_step": 34, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 0.7783562541007996, "train_loss_dp": 0.7511874437332153, "train_loss_llm": 2.716879367828369, "global_step": 35, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.30172598361969, "train_loss_dp": 1.2940138578414917, "train_loss_llm": 0.7712092399597168, "global_step": 36, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0512481927871704, "train_loss_dp": 1.0373506546020508, "train_loss_llm": 1.389752984046936, "global_step": 37, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3016198873519897, "train_loss_dp": 1.2951383590698242, "train_loss_llm": 0.648158073425293, "global_step": 38, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.477652907371521, "train_loss_dp": 1.4623231887817383, "train_loss_llm": 1.5329742431640625, "global_step": 39, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0078511238098145, "train_loss_dp": 1.0005815029144287, "train_loss_llm": 0.7269659042358398, "global_step": 40, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.1381250619888306, "train_loss_dp": 1.1300842761993408, "train_loss_llm": 0.8040800094604492, "global_step": 41, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.0345423221588135, "train_loss_dp": 1.02858567237854, "train_loss_llm": 0.5956590175628662, "global_step": 42, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.1918821334838867, "train_loss_dp": 1.182042121887207, "train_loss_llm": 0.9839978218078613, "global_step": 43, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.1983389854431152, "train_loss_dp": 1.1782596111297607, "train_loss_llm": 2.007936477661133, "global_step": 44, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.13905668258667, "train_loss_dp": 1.1314916610717773, "train_loss_llm": 0.7565021514892578, "global_step": 45, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.3465195894241333, "train_loss_dp": 1.3093535900115967, "train_loss_llm": 3.7166032791137695, "global_step": 46, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.300544261932373, "train_loss_dp": 1.2961735725402832, "train_loss_llm": 0.43706512451171875, "global_step": 47, "epoch": 0, "lr": 6.000000000000001e-07}
+{"train_loss": 1.7413785457611084, "train_loss_dp": 1.714151382446289, "train_loss_llm": 2.722715377807617, "global_step": 48, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.559943437576294, "train_loss_dp": 1.5114814043045044, "train_loss_llm": 4.8462066650390625, "global_step": 49, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.3392784595489502, "train_loss_dp": 1.2820918560028076, "train_loss_llm": 5.718661308288574, "global_step": 50, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.091318130493164, "train_loss_dp": 1.0823571681976318, "train_loss_llm": 0.896092414855957, "global_step": 51, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.3553695678710938, "train_loss_dp": 1.3473711013793945, "train_loss_llm": 0.7998428344726562, "global_step": 52, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.400208592414856, "train_loss_dp": 1.3946783542633057, "train_loss_llm": 0.553019642829895, "global_step": 53, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.024678111076355, "train_loss_dp": 1.0102272033691406, "train_loss_llm": 1.4450857639312744, "global_step": 54, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.2564771175384521, "train_loss_dp": 1.2389534711837769, "train_loss_llm": 1.7523598670959473, "global_step": 55, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.229862928390503, "train_loss_dp": 1.229862928390503, "train_loss_llm": 0.0, "global_step": 56, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.4351203441619873, "train_loss_dp": 1.4280118942260742, "train_loss_llm": 0.7108421325683594, "global_step": 57, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.7568683624267578, "train_loss_dp": 1.7568683624267578, "train_loss_llm": 0.0, "global_step": 58, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 0.9146701693534851, "train_loss_dp": 0.9106614589691162, "train_loss_llm": 0.4008716344833374, "global_step": 59, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.0629714727401733, "train_loss_dp": 1.0582225322723389, "train_loss_llm": 0.47489166259765625, "global_step": 60, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.3070189952850342, "train_loss_dp": 1.2989965677261353, "train_loss_llm": 0.8022446632385254, "global_step": 61, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.1110565662384033, "train_loss_dp": 1.1095645427703857, "train_loss_llm": 0.14920806884765625, "global_step": 62, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.1010279655456543, "train_loss_dp": 1.095658302307129, "train_loss_llm": 0.536961555480957, "global_step": 63, "epoch": 0, "lr": 8.000000000000001e-07}
+{"train_loss": 1.040368676185608, "train_loss_dp": 1.0287753343582153, "train_loss_llm": 1.159337043762207, "global_step": 64, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.283882975578308, "train_loss_dp": 1.2667810916900635, "train_loss_llm": 1.7101936340332031, "global_step": 65, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.2713927030563354, "train_loss_dp": 1.2617037296295166, "train_loss_llm": 0.9688944816589355, "global_step": 66, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 0.9697917699813843, "train_loss_dp": 0.9413739442825317, "train_loss_llm": 2.8417840003967285, "global_step": 67, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.2020347118377686, "train_loss_dp": 1.1800479888916016, "train_loss_llm": 2.198676824569702, "global_step": 68, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3157751560211182, "train_loss_dp": 1.2875651121139526, "train_loss_llm": 2.8210039138793945, "global_step": 69, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.202554702758789, "train_loss_dp": 1.1841760873794556, "train_loss_llm": 1.8378667831420898, "global_step": 70, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3491004705429077, "train_loss_dp": 1.3438347578048706, "train_loss_llm": 0.5265769958496094, "global_step": 71, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 0.936764121055603, "train_loss_dp": 0.9179775714874268, "train_loss_llm": 1.8786544799804688, "global_step": 72, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.1598823070526123, "train_loss_dp": 1.135289192199707, "train_loss_llm": 2.4593095779418945, "global_step": 73, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.3598875999450684, "train_loss_dp": 1.3555123805999756, "train_loss_llm": 0.4375239610671997, "global_step": 74, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.860397458076477, "train_loss_dp": 1.8556218147277832, "train_loss_llm": 0.47756195068359375, "global_step": 75, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.4227266311645508, "train_loss_dp": 1.414045810699463, "train_loss_llm": 0.8680804967880249, "global_step": 76, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.267741322517395, "train_loss_dp": 1.25203275680542, "train_loss_llm": 1.5708551406860352, "global_step": 77, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 0.9328470230102539, "train_loss_dp": 0.9169472455978394, "train_loss_llm": 1.5899808406829834, "global_step": 78, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 0.8221402168273926, "train_loss_dp": 0.7923790216445923, "train_loss_llm": 2.9761199951171875, "global_step": 79, "epoch": 0, "lr": 1.0000000000000002e-06}
+{"train_loss": 1.015797734260559, "train_loss_dp": 1.0092968940734863, "train_loss_llm": 0.6500859260559082, "global_step": 80, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.287087082862854, "train_loss_dp": 1.279496669769287, "train_loss_llm": 0.759040355682373, "global_step": 81, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.4711720943450928, "train_loss_dp": 1.4356005191802979, "train_loss_llm": 3.5571517944335938, "global_step": 82, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.7746796607971191, "train_loss_dp": 1.5105911493301392, "train_loss_llm": 26.408855438232422, "global_step": 83, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 0.937280535697937, "train_loss_dp": 0.9340149760246277, "train_loss_llm": 0.32655787467956543, "global_step": 84, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.119026780128479, "train_loss_dp": 1.1136049032211304, "train_loss_llm": 0.5421831607818604, "global_step": 85, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.1145763397216797, "train_loss_dp": 1.1065962314605713, "train_loss_llm": 0.7980070114135742, "global_step": 86, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.3616853952407837, "train_loss_dp": 1.353653907775879, "train_loss_llm": 0.803144633769989, "global_step": 87, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.248418927192688, "train_loss_dp": 1.2383646965026855, "train_loss_llm": 1.005427360534668, "global_step": 88, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.0686962604522705, "train_loss_dp": 1.0592976808547974, "train_loss_llm": 0.9398558139801025, "global_step": 89, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 0.9523566961288452, "train_loss_dp": 0.9393671751022339, "train_loss_llm": 1.2989501953125, "global_step": 90, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 0.9704211354255676, "train_loss_dp": 0.9571309089660645, "train_loss_llm": 1.3290214538574219, "global_step": 91, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.2581830024719238, "train_loss_dp": 1.2500872611999512, "train_loss_llm": 0.8095684051513672, "global_step": 92, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.091536521911621, "train_loss_dp": 1.0850541591644287, "train_loss_llm": 0.6482415199279785, "global_step": 93, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.018674373626709, "train_loss_dp": 1.0096831321716309, "train_loss_llm": 0.8991279602050781, "global_step": 94, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.0770442485809326, "train_loss_dp": 1.067256212234497, "train_loss_llm": 0.9788017272949219, "global_step": 95, "epoch": 0, "lr": 1.2000000000000002e-06}
+{"train_loss": 1.0118895769119263, "train_loss_dp": 0.9556567668914795, "train_loss_llm": 5.6232805252075195, "global_step": 96, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.0088353157043457, "train_loss_dp": 0.9946292638778687, "train_loss_llm": 1.4206085205078125, "global_step": 97, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.0618704557418823, "train_loss_dp": 1.0533971786499023, "train_loss_llm": 0.8473281860351562, "global_step": 98, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 0.9173486232757568, "train_loss_dp": 0.8348127603530884, "train_loss_llm": 8.253589630126953, "global_step": 99, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.2309874296188354, "train_loss_dp": 1.2110412120819092, "train_loss_llm": 1.9946260452270508, "global_step": 100, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.530369520187378, "train_loss_dp": 1.5212764739990234, "train_loss_llm": 0.9093027114868164, "global_step": 101, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.2080981731414795, "train_loss_dp": 1.183016061782837, "train_loss_llm": 2.508209228515625, "global_step": 102, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.28593111038208, "train_loss_dp": 1.2735224962234497, "train_loss_llm": 1.240865707397461, "global_step": 103, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.374464511871338, "train_loss_dp": 1.3602583408355713, "train_loss_llm": 1.4206228256225586, "global_step": 104, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 0.9377275109291077, "train_loss_dp": 0.9263818860054016, "train_loss_llm": 1.1345635652542114, "global_step": 105, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.3163727521896362, "train_loss_dp": 1.311410903930664, "train_loss_llm": 0.4961848258972168, "global_step": 106, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 0.9401397109031677, "train_loss_dp": 0.9401397109031677, "train_loss_llm": 0.0, "global_step": 107, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.3074995279312134, "train_loss_dp": 1.2755658626556396, "train_loss_llm": 3.193366050720215, "global_step": 108, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.3464398384094238, "train_loss_dp": 1.3270540237426758, "train_loss_llm": 1.938586711883545, "global_step": 109, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.1693884134292603, "train_loss_dp": 1.1397781372070312, "train_loss_llm": 2.961027145385742, "global_step": 110, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.522181510925293, "train_loss_dp": 1.5133485794067383, "train_loss_llm": 0.8832952976226807, "global_step": 111, "epoch": 0, "lr": 1.4000000000000001e-06}
+{"train_loss": 1.3709678649902344, "train_loss_dp": 1.3595249652862549, "train_loss_llm": 1.144289255142212, "global_step": 112, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.987101137638092, "train_loss_dp": 0.9671929478645325, "train_loss_llm": 1.9908201694488525, "global_step": 113, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.2418595552444458, "train_loss_dp": 1.2201751470565796, "train_loss_llm": 2.1684465408325195, "global_step": 114, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.8841862082481384, "train_loss_dp": 0.8708856105804443, "train_loss_llm": 1.3300609588623047, "global_step": 115, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.22050142288208, "train_loss_dp": 1.2014448642730713, "train_loss_llm": 1.905655860900879, "global_step": 116, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.1527434587478638, "train_loss_dp": 1.1332242488861084, "train_loss_llm": 1.951920509338379, "global_step": 117, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.8063086867332458, "train_loss_dp": 0.7741774320602417, "train_loss_llm": 3.2131271362304688, "global_step": 118, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.9218640923500061, "train_loss_dp": 0.913447916507721, "train_loss_llm": 0.8416156768798828, "global_step": 119, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.399662733078003, "train_loss_dp": 1.3826619386672974, "train_loss_llm": 1.7000749111175537, "global_step": 120, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.5391050577163696, "train_loss_dp": 1.53459632396698, "train_loss_llm": 0.45086777210235596, "global_step": 121, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.4053776264190674, "train_loss_dp": 1.3952162265777588, "train_loss_llm": 1.016134262084961, "global_step": 122, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.246399998664856, "train_loss_dp": 1.2373460531234741, "train_loss_llm": 0.9053974151611328, "global_step": 123, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.199317455291748, "train_loss_dp": 1.186524510383606, "train_loss_llm": 1.2792892456054688, "global_step": 124, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.202831506729126, "train_loss_dp": 1.1726295948028564, "train_loss_llm": 3.020185708999634, "global_step": 125, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.3294272422790527, "train_loss_dp": 1.3156859874725342, "train_loss_llm": 1.3741207122802734, "global_step": 126, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 0.8595767617225647, "train_loss_dp": 0.8500500917434692, "train_loss_llm": 0.9526662826538086, "global_step": 127, "epoch": 0, "lr": 1.6000000000000001e-06}
+{"train_loss": 1.0012800693511963, "train_loss_dp": 0.9812033176422119, "train_loss_llm": 2.0076799392700195, "global_step": 128, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.0801386833190918, "train_loss_dp": 1.0447351932525635, "train_loss_llm": 3.540346622467041, "global_step": 129, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2674883604049683, "train_loss_dp": 1.2562992572784424, "train_loss_llm": 1.1189069747924805, "global_step": 130, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2178473472595215, "train_loss_dp": 1.21113920211792, "train_loss_llm": 0.6708115339279175, "global_step": 131, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.9338388442993164, "train_loss_dp": 0.9120755195617676, "train_loss_llm": 2.176330804824829, "global_step": 132, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.495168685913086, "train_loss_dp": 1.4798719882965088, "train_loss_llm": 1.5296683311462402, "global_step": 133, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1203300952911377, "train_loss_dp": 1.109201431274414, "train_loss_llm": 1.112870693206787, "global_step": 134, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.7030981183052063, "train_loss_dp": 0.7003169059753418, "train_loss_llm": 0.2781219482421875, "global_step": 135, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.7989674806594849, "train_loss_dp": 0.7763664126396179, "train_loss_llm": 2.2601051330566406, "global_step": 136, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.2034188508987427, "train_loss_dp": 1.1995105743408203, "train_loss_llm": 0.39082586765289307, "global_step": 137, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.5198830366134644, "train_loss_dp": 1.5114814043045044, "train_loss_llm": 0.8401651382446289, "global_step": 138, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1898471117019653, "train_loss_dp": 1.1625717878341675, "train_loss_llm": 2.7275333404541016, "global_step": 139, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.9431044459342957, "train_loss_dp": 0.9387279748916626, "train_loss_llm": 0.4376487731933594, "global_step": 140, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1791592836380005, "train_loss_dp": 1.1608953475952148, "train_loss_llm": 1.8263983726501465, "global_step": 141, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1842983961105347, "train_loss_dp": 1.1788935661315918, "train_loss_llm": 0.5404876470565796, "global_step": 142, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 1.1830167770385742, "train_loss_dp": 1.177420973777771, "train_loss_llm": 0.5595860481262207, "global_step": 143, "epoch": 0, "lr": 1.8e-06}
+{"train_loss": 0.7706676721572876, "train_loss_dp": 0.7500808835029602, "train_loss_llm": 2.0586776733398438, "global_step": 144, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0568273067474365, "train_loss_dp": 1.0306296348571777, "train_loss_llm": 2.6197643280029297, "global_step": 145, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.8390830159187317, "train_loss_dp": 0.8200579881668091, "train_loss_llm": 1.9025001525878906, "global_step": 146, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.1467828750610352, "train_loss_dp": 1.1427737474441528, "train_loss_llm": 0.40091609954833984, "global_step": 147, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.9128871560096741, "train_loss_dp": 0.9029089212417603, "train_loss_llm": 0.9978246688842773, "global_step": 148, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.1126600503921509, "train_loss_dp": 1.0998445749282837, "train_loss_llm": 1.2815475463867188, "global_step": 149, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.9934484362602234, "train_loss_dp": 0.983627200126648, "train_loss_llm": 0.9821264743804932, "global_step": 150, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.8006670475006104, "train_loss_dp": 0.7753530740737915, "train_loss_llm": 2.5313987731933594, "global_step": 151, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0276238918304443, "train_loss_dp": 1.0108177661895752, "train_loss_llm": 1.6806116104125977, "global_step": 152, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.4698054790496826, "train_loss_dp": 1.4573051929473877, "train_loss_llm": 1.2500245571136475, "global_step": 153, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.1208338737487793, "train_loss_dp": 1.1033995151519775, "train_loss_llm": 1.7434301376342773, "global_step": 154, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0510036945343018, "train_loss_dp": 1.000337839126587, "train_loss_llm": 5.066581726074219, "global_step": 155, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 0.7954217195510864, "train_loss_dp": 0.7764338254928589, "train_loss_llm": 1.898789882659912, "global_step": 156, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.0372687578201294, "train_loss_dp": 1.0325254201889038, "train_loss_llm": 0.4743328094482422, "global_step": 157, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.092544436454773, "train_loss_dp": 1.0861330032348633, "train_loss_llm": 0.6411471366882324, "global_step": 158, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.2194020748138428, "train_loss_dp": 1.1884307861328125, "train_loss_llm": 3.0971336364746094, "global_step": 159, "epoch": 0, "lr": 2.0000000000000003e-06}
+{"train_loss": 1.107576608657837, "train_loss_dp": 1.0977963209152222, "train_loss_llm": 0.9780311584472656, "global_step": 160, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.0311685800552368, "train_loss_dp": 1.0239753723144531, "train_loss_llm": 0.7193160057067871, "global_step": 161, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.8929820656776428, "train_loss_dp": 0.8832759857177734, "train_loss_llm": 0.9706106185913086, "global_step": 162, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.0656144618988037, "train_loss_dp": 1.052584171295166, "train_loss_llm": 1.3030333518981934, "global_step": 163, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.196394920349121, "train_loss_dp": 1.188420295715332, "train_loss_llm": 0.7974569797515869, "global_step": 164, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.0412317514419556, "train_loss_dp": 1.028942584991455, "train_loss_llm": 1.228917121887207, "global_step": 165, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.2923147678375244, "train_loss_dp": 1.275970697402954, "train_loss_llm": 1.634401559829712, "global_step": 166, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.457854986190796, "train_loss_dp": 1.4523682594299316, "train_loss_llm": 0.5486674308776855, "global_step": 167, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.8358233571052551, "train_loss_dp": 0.8237012624740601, "train_loss_llm": 1.2122089862823486, "global_step": 168, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.2824134826660156, "train_loss_dp": 1.2447845935821533, "train_loss_llm": 3.762887954711914, "global_step": 169, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.8838927149772644, "train_loss_dp": 0.7977854013442993, "train_loss_llm": 8.610729217529297, "global_step": 170, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.159955620765686, "train_loss_dp": 1.1208444833755493, "train_loss_llm": 3.9111108779907227, "global_step": 171, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.9689618349075317, "train_loss_dp": 0.9650105237960815, "train_loss_llm": 0.39513182640075684, "global_step": 172, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.1551589965820312, "train_loss_dp": 1.1478490829467773, "train_loss_llm": 0.7309961318969727, "global_step": 173, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 0.9433655738830566, "train_loss_dp": 0.9433655738830566, "train_loss_llm": 0.0, "global_step": 174, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.115327000617981, "train_loss_dp": 1.1106475591659546, "train_loss_llm": 0.46794044971466064, "global_step": 175, "epoch": 0, "lr": 2.2e-06}
+{"train_loss": 1.100792646408081, "train_loss_dp": 1.082282543182373, "train_loss_llm": 1.8510119915008545, "global_step": 176, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.0397549867630005, "train_loss_dp": 1.0299686193466187, "train_loss_llm": 0.9786415100097656, "global_step": 177, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.2450857162475586, "train_loss_dp": 1.2004895210266113, "train_loss_llm": 4.459615707397461, "global_step": 178, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9495880007743835, "train_loss_dp": 0.9388677477836609, "train_loss_llm": 1.0720233917236328, "global_step": 179, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.7823573350906372, "train_loss_dp": 0.7736498117446899, "train_loss_llm": 0.8707506656646729, "global_step": 180, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9191964268684387, "train_loss_dp": 0.9063225984573364, "train_loss_llm": 1.2873849868774414, "global_step": 181, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.1402461528778076, "train_loss_dp": 1.1191234588623047, "train_loss_llm": 2.1122732162475586, "global_step": 182, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.8272892832756042, "train_loss_dp": 0.8228100538253784, "train_loss_llm": 0.44792068004608154, "global_step": 183, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.0445338487625122, "train_loss_dp": 1.0336251258850098, "train_loss_llm": 1.0908684730529785, "global_step": 184, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.0622131824493408, "train_loss_dp": 0.991152286529541, "train_loss_llm": 7.106090545654297, "global_step": 185, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.8597482442855835, "train_loss_dp": 0.8495693206787109, "train_loss_llm": 1.0178899765014648, "global_step": 186, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9844797253608704, "train_loss_dp": 0.9774206280708313, "train_loss_llm": 0.7059082984924316, "global_step": 187, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.3411743640899658, "train_loss_dp": 1.33035409450531, "train_loss_llm": 1.0820218324661255, "global_step": 188, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.963067352771759, "train_loss_dp": 0.9588446021080017, "train_loss_llm": 0.4222745895385742, "global_step": 189, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.047143816947937, "train_loss_dp": 1.0167937278747559, "train_loss_llm": 3.0350112915039062, "global_step": 190, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 1.2898896932601929, "train_loss_dp": 1.2857036590576172, "train_loss_llm": 0.418597936630249, "global_step": 191, "epoch": 0, "lr": 2.4000000000000003e-06}
+{"train_loss": 0.9388867020606995, "train_loss_dp": 0.9324394464492798, "train_loss_llm": 0.6447279453277588, "global_step": 192, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.2793828248977661, "train_loss_dp": 1.231186866760254, "train_loss_llm": 4.819598197937012, "global_step": 193, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1666911840438843, "train_loss_dp": 1.1645325422286987, "train_loss_llm": 0.2158680558204651, "global_step": 194, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.303589105606079, "train_loss_dp": 1.2741291522979736, "train_loss_llm": 2.9459991455078125, "global_step": 195, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1922321319580078, "train_loss_dp": 1.1801778078079224, "train_loss_llm": 1.2054328918457031, "global_step": 196, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.9651893973350525, "train_loss_dp": 0.9529061317443848, "train_loss_llm": 1.2283272743225098, "global_step": 197, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1780668497085571, "train_loss_dp": 1.1543831825256348, "train_loss_llm": 2.3683700561523438, "global_step": 198, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.7351898550987244, "train_loss_dp": 0.7292382717132568, "train_loss_llm": 0.5951566696166992, "global_step": 199, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.2100989818572998, "train_loss_dp": 0.9783383011817932, "train_loss_llm": 23.176063537597656, "global_step": 200, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.0502445697784424, "train_loss_dp": 1.0106562376022339, "train_loss_llm": 3.9588356018066406, "global_step": 201, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.6521593332290649, "train_loss_dp": 0.6481107473373413, "train_loss_llm": 0.40485620498657227, "global_step": 202, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.4098745584487915, "train_loss_dp": 1.303734540939331, "train_loss_llm": 10.614002227783203, "global_step": 203, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.3819440603256226, "train_loss_dp": 1.3341147899627686, "train_loss_llm": 4.782923698425293, "global_step": 204, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.103819489479065, "train_loss_dp": 1.1017730236053467, "train_loss_llm": 0.2046513557434082, "global_step": 205, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 0.940501868724823, "train_loss_dp": 0.9306132197380066, "train_loss_llm": 0.9888644218444824, "global_step": 206, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.1322269439697266, "train_loss_dp": 1.1281698942184448, "train_loss_llm": 0.40570664405822754, "global_step": 207, "epoch": 0, "lr": 2.6e-06}
+{"train_loss": 1.0362513065338135, "train_loss_dp": 1.026556134223938, "train_loss_llm": 0.9695119857788086, "global_step": 208, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8448707461357117, "train_loss_dp": 0.8448707461357117, "train_loss_llm": 0.0, "global_step": 209, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8706929087638855, "train_loss_dp": 0.8668233752250671, "train_loss_llm": 0.3869504928588867, "global_step": 210, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.2980793714523315, "train_loss_dp": 1.2917020320892334, "train_loss_llm": 0.6377372741699219, "global_step": 211, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.2147033214569092, "train_loss_dp": 1.20652174949646, "train_loss_llm": 0.8181605339050293, "global_step": 212, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.9953945279121399, "train_loss_dp": 0.9913030862808228, "train_loss_llm": 0.4091458320617676, "global_step": 213, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.2621140480041504, "train_loss_dp": 1.2506753206253052, "train_loss_llm": 1.1438686847686768, "global_step": 214, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8768635988235474, "train_loss_dp": 0.8692286014556885, "train_loss_llm": 0.7634992599487305, "global_step": 215, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.1312974691390991, "train_loss_dp": 1.1301722526550293, "train_loss_llm": 0.11252593994140625, "global_step": 216, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8877076506614685, "train_loss_dp": 0.8807100057601929, "train_loss_llm": 0.6997666358947754, "global_step": 217, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.7655230164527893, "train_loss_dp": 0.752112627029419, "train_loss_llm": 1.3410396575927734, "global_step": 218, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.8268309235572815, "train_loss_dp": 0.8216547966003418, "train_loss_llm": 0.5176153182983398, "global_step": 219, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.9836443066596985, "train_loss_dp": 0.9762389659881592, "train_loss_llm": 0.740534782409668, "global_step": 220, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.3871673345565796, "train_loss_dp": 1.3758190870285034, "train_loss_llm": 1.1348257064819336, "global_step": 221, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 0.9618462920188904, "train_loss_dp": 0.9548556804656982, "train_loss_llm": 0.6990604400634766, "global_step": 222, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.121734380722046, "train_loss_dp": 1.1161866188049316, "train_loss_llm": 0.5547780990600586, "global_step": 223, "epoch": 0, "lr": 2.8000000000000003e-06}
+{"train_loss": 1.2717845439910889, "train_loss_dp": 1.2252352237701416, "train_loss_llm": 4.654937744140625, "global_step": 224, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.133670449256897, "train_loss_dp": 1.1292202472686768, "train_loss_llm": 0.445019006729126, "global_step": 225, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.1635395288467407, "train_loss_dp": 1.1458921432495117, "train_loss_llm": 1.764742374420166, "global_step": 226, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.790412425994873, "train_loss_dp": 0.7768189907073975, "train_loss_llm": 1.359344482421875, "global_step": 227, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.2196625471115112, "train_loss_dp": 1.2109549045562744, "train_loss_llm": 0.8707695007324219, "global_step": 228, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.0803388357162476, "train_loss_dp": 1.0543369054794312, "train_loss_llm": 2.600187301635742, "global_step": 229, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.9198564291000366, "train_loss_dp": 0.9066956043243408, "train_loss_llm": 1.3160839080810547, "global_step": 230, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.8806276321411133, "train_loss_dp": 0.8740400671958923, "train_loss_llm": 0.6587562561035156, "global_step": 231, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.7853339910507202, "train_loss_dp": 0.7752537727355957, "train_loss_llm": 1.0080232620239258, "global_step": 232, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.9136844873428345, "train_loss_dp": 0.829277753829956, "train_loss_llm": 8.440671920776367, "global_step": 233, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.8012468814849854, "train_loss_dp": 0.7941519021987915, "train_loss_llm": 0.7094950675964355, "global_step": 234, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.5320197343826294, "train_loss_dp": 1.488776445388794, "train_loss_llm": 4.324326515197754, "global_step": 235, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.2080738544464111, "train_loss_dp": 1.1741201877593994, "train_loss_llm": 3.395369529724121, "global_step": 236, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.4821357727050781, "train_loss_dp": 1.4457201957702637, "train_loss_llm": 3.6415634155273438, "global_step": 237, "epoch": 0, "lr": 3e-06}
+{"train_loss": 2.016885280609131, "train_loss_dp": 1.2957777976989746, "train_loss_llm": 72.11074829101562, "global_step": 238, "epoch": 0, "lr": 3e-06}
+{"train_loss": 0.9761942028999329, "train_loss_dp": 0.9635788202285767, "train_loss_llm": 1.2615375518798828, "global_step": 239, "epoch": 0, "lr": 3e-06}
+{"train_loss": 1.3200069665908813, "train_loss_dp": 1.125038743019104, "train_loss_llm": 19.4968204498291, "global_step": 240, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.2333816289901733, "train_loss_dp": 1.2297680377960205, "train_loss_llm": 0.3613567352294922, "global_step": 241, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0169427394866943, "train_loss_dp": 0.951324462890625, "train_loss_llm": 6.561830520629883, "global_step": 242, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.8573359847068787, "train_loss_dp": 0.8521511554718018, "train_loss_llm": 0.5184810161590576, "global_step": 243, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0643607378005981, "train_loss_dp": 1.0499974489212036, "train_loss_llm": 1.4363293647766113, "global_step": 244, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.251783847808838, "train_loss_dp": 1.220577359199524, "train_loss_llm": 3.120647430419922, "global_step": 245, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9586256742477417, "train_loss_dp": 0.952906608581543, "train_loss_llm": 0.5719079971313477, "global_step": 246, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9674288034439087, "train_loss_dp": 0.9626787900924683, "train_loss_llm": 0.4749996066093445, "global_step": 247, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1254202127456665, "train_loss_dp": 1.115891456604004, "train_loss_llm": 0.9528732299804688, "global_step": 248, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.1593478918075562, "train_loss_dp": 1.145406723022461, "train_loss_llm": 1.3941142559051514, "global_step": 249, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9617233276367188, "train_loss_dp": 0.9384018778800964, "train_loss_llm": 2.3321447372436523, "global_step": 250, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.0569924116134644, "train_loss_dp": 1.0468077659606934, "train_loss_llm": 1.0184690952301025, "global_step": 251, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9256843328475952, "train_loss_dp": 0.9084959030151367, "train_loss_llm": 1.7188434600830078, "global_step": 252, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.9216443300247192, "train_loss_dp": 0.893805742263794, "train_loss_llm": 2.78385591506958, "global_step": 253, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.6723043322563171, "train_loss_dp": 0.6642416715621948, "train_loss_llm": 0.8062641620635986, "global_step": 254, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 0.908981442451477, "train_loss_dp": 0.8921409249305725, "train_loss_llm": 1.6840534210205078, "global_step": 255, "epoch": 0, "lr": 3.2000000000000003e-06}
+{"train_loss": 1.204549789428711, "train_loss_dp": 1.198688268661499, "train_loss_llm": 0.5861563682556152, "global_step": 256, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2233939170837402, "train_loss_dp": 1.2152185440063477, "train_loss_llm": 0.817535400390625, "global_step": 257, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.104799747467041, "train_loss_dp": 1.0953290462493896, "train_loss_llm": 0.9470711946487427, "global_step": 258, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.3139615058898926, "train_loss_dp": 1.3055317401885986, "train_loss_llm": 0.8429725170135498, "global_step": 259, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.8189157247543335, "train_loss_dp": 0.7952377200126648, "train_loss_llm": 2.3677990436553955, "global_step": 260, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.7308699488639832, "train_loss_dp": 0.7195430994033813, "train_loss_llm": 1.1326828002929688, "global_step": 261, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.7054136395454407, "train_loss_dp": 0.6950370669364929, "train_loss_llm": 1.037658452987671, "global_step": 262, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.9875133633613586, "train_loss_dp": 0.9790518283843994, "train_loss_llm": 0.8461538553237915, "global_step": 263, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.1116515398025513, "train_loss_dp": 1.099230408668518, "train_loss_llm": 1.2421073913574219, "global_step": 264, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.26888108253479, "train_loss_dp": 1.2592241764068604, "train_loss_llm": 0.9656886458396912, "global_step": 265, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.076149344444275, "train_loss_dp": 1.0639097690582275, "train_loss_llm": 1.2239608764648438, "global_step": 266, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.063284993171692, "train_loss_dp": 1.0430526733398438, "train_loss_llm": 2.0232338905334473, "global_step": 267, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.9694548845291138, "train_loss_dp": 0.9406946897506714, "train_loss_llm": 2.8760199546813965, "global_step": 268, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.8402442336082458, "train_loss_dp": 0.8313390612602234, "train_loss_llm": 0.8905158042907715, "global_step": 269, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.0363178253173828, "train_loss_dp": 1.0363178253173828, "train_loss_llm": 0.0, "global_step": 270, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 1.2708722352981567, "train_loss_dp": 1.249863624572754, "train_loss_llm": 2.1008596420288086, "global_step": 271, "epoch": 0, "lr": 3.4000000000000005e-06}
+{"train_loss": 0.8101922273635864, "train_loss_dp": 0.8029913306236267, "train_loss_llm": 0.7200882434844971, "global_step": 272, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.270508050918579, "train_loss_dp": 1.2500911951065063, "train_loss_llm": 2.041682243347168, "global_step": 273, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.8519293069839478, "train_loss_dp": 0.8470282554626465, "train_loss_llm": 0.4901081323623657, "global_step": 274, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.206408143043518, "train_loss_dp": 1.1970850229263306, "train_loss_llm": 0.9323081970214844, "global_step": 275, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.2787306308746338, "train_loss_dp": 1.1633915901184082, "train_loss_llm": 11.533899307250977, "global_step": 276, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.938746452331543, "train_loss_dp": 0.9267393350601196, "train_loss_llm": 1.2007102966308594, "global_step": 277, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.0588295459747314, "train_loss_dp": 1.0540907382965088, "train_loss_llm": 0.473876953125, "global_step": 278, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.1287081241607666, "train_loss_dp": 1.1169074773788452, "train_loss_llm": 1.1800613403320312, "global_step": 279, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.1671936511993408, "train_loss_dp": 1.1598732471466064, "train_loss_llm": 0.7320423126220703, "global_step": 280, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9664711356163025, "train_loss_dp": 0.9661685824394226, "train_loss_llm": 0.0302581787109375, "global_step": 281, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.8956195712089539, "train_loss_dp": 0.8922961354255676, "train_loss_llm": 0.33234232664108276, "global_step": 282, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.0822522640228271, "train_loss_dp": 1.0808215141296387, "train_loss_llm": 0.14307308197021484, "global_step": 283, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9898412823677063, "train_loss_dp": 0.9549367427825928, "train_loss_llm": 3.4904537200927734, "global_step": 284, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.8137789964675903, "train_loss_dp": 0.8008929491043091, "train_loss_llm": 1.288604736328125, "global_step": 285, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 1.1449403762817383, "train_loss_dp": 1.1220868825912476, "train_loss_llm": 2.285353660583496, "global_step": 286, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.9789936542510986, "train_loss_dp": 0.9397944808006287, "train_loss_llm": 3.919917583465576, "global_step": 287, "epoch": 0, "lr": 3.6e-06}
+{"train_loss": 0.7996547818183899, "train_loss_dp": 0.7951436042785645, "train_loss_llm": 0.45111608505249023, "global_step": 288, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.1863093376159668, "train_loss_dp": 1.1545655727386475, "train_loss_llm": 3.1743812561035156, "global_step": 289, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.094476580619812, "train_loss_dp": 1.0789673328399658, "train_loss_llm": 1.5509250164031982, "global_step": 290, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.7286985516548157, "train_loss_dp": 0.695349395275116, "train_loss_llm": 3.3349151611328125, "global_step": 291, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.8538903594017029, "train_loss_dp": 0.8476069569587708, "train_loss_llm": 0.6283421516418457, "global_step": 292, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.096050500869751, "train_loss_dp": 1.081933617591858, "train_loss_llm": 1.4116897583007812, "global_step": 293, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.8915770649909973, "train_loss_dp": 0.880878210067749, "train_loss_llm": 1.069886565208435, "global_step": 294, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.1769020557403564, "train_loss_dp": 1.1708498001098633, "train_loss_llm": 0.6052254438400269, "global_step": 295, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.1362321376800537, "train_loss_dp": 1.1254382133483887, "train_loss_llm": 1.079390048980713, "global_step": 296, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0254043340682983, "train_loss_dp": 1.0254043340682983, "train_loss_llm": 0.0, "global_step": 297, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9390648007392883, "train_loss_dp": 0.9337210655212402, "train_loss_llm": 0.534372091293335, "global_step": 298, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0619200468063354, "train_loss_dp": 0.9854391813278198, "train_loss_llm": 7.648087501525879, "global_step": 299, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9273417592048645, "train_loss_dp": 0.9239466786384583, "train_loss_llm": 0.339508056640625, "global_step": 300, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.8679295778274536, "train_loss_dp": 0.8602054119110107, "train_loss_llm": 0.7724189758300781, "global_step": 301, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.2388479709625244, "train_loss_dp": 1.1727399826049805, "train_loss_llm": 6.61080265045166, "global_step": 302, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 0.9015404582023621, "train_loss_dp": 0.8882120847702026, "train_loss_llm": 1.3328378200531006, "global_step": 303, "epoch": 0, "lr": 3.8e-06}
+{"train_loss": 1.0708075761795044, "train_loss_dp": 1.0668206214904785, "train_loss_llm": 0.39869219064712524, "global_step": 304, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.9203828573226929, "train_loss_dp": 0.8907637596130371, "train_loss_llm": 2.961907386779785, "global_step": 305, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.0209006071090698, "train_loss_dp": 1.0103039741516113, "train_loss_llm": 1.059661865234375, "global_step": 306, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.0560258626937866, "train_loss_dp": 1.0481963157653809, "train_loss_llm": 0.7829568386077881, "global_step": 307, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.8696987628936768, "train_loss_dp": 0.853388786315918, "train_loss_llm": 1.6309967041015625, "global_step": 308, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.2426722049713135, "train_loss_dp": 1.226274013519287, "train_loss_llm": 1.639822006225586, "global_step": 309, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.0811240673065186, "train_loss_dp": 1.0267776250839233, "train_loss_llm": 5.434648513793945, "global_step": 310, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.268541932106018, "train_loss_dp": 1.2665159702301025, "train_loss_llm": 0.2025909423828125, "global_step": 311, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.1095284223556519, "train_loss_dp": 1.1064952611923218, "train_loss_llm": 0.3033146858215332, "global_step": 312, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.9223203659057617, "train_loss_dp": 0.9167141914367676, "train_loss_llm": 0.5606203079223633, "global_step": 313, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.2852535247802734, "train_loss_dp": 1.269548773765564, "train_loss_llm": 1.5704741477966309, "global_step": 314, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.7039891481399536, "train_loss_dp": 0.697694718837738, "train_loss_llm": 0.6294441223144531, "global_step": 315, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.4207193851470947, "train_loss_dp": 1.4016109704971313, "train_loss_llm": 1.9108467102050781, "global_step": 316, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.0881227254867554, "train_loss_dp": 1.0802347660064697, "train_loss_llm": 0.7887916564941406, "global_step": 317, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 1.4858653545379639, "train_loss_dp": 1.4580174684524536, "train_loss_llm": 2.784789800643921, "global_step": 318, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.7847555875778198, "train_loss_dp": 0.7751880884170532, "train_loss_llm": 0.9567484855651855, "global_step": 319, "epoch": 0, "lr": 4.000000000000001e-06}
+{"train_loss": 0.9809284806251526, "train_loss_dp": 0.9737030267715454, "train_loss_llm": 0.7225475311279297, "global_step": 320, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1537349224090576, "train_loss_dp": 1.1510746479034424, "train_loss_llm": 0.266021728515625, "global_step": 321, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 0.835584819316864, "train_loss_dp": 0.8233403563499451, "train_loss_llm": 1.2244462966918945, "global_step": 322, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1909435987472534, "train_loss_dp": 1.170578956604004, "train_loss_llm": 2.036468505859375, "global_step": 323, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.04839289188385, "train_loss_dp": 1.0379482507705688, "train_loss_llm": 1.0444637537002563, "global_step": 324, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1169261932373047, "train_loss_dp": 1.113403558731079, "train_loss_llm": 0.35226768255233765, "global_step": 325, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 0.9538261890411377, "train_loss_dp": 0.900374174118042, "train_loss_llm": 5.3451995849609375, "global_step": 326, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.8533211946487427, "train_loss_dp": 1.1018012762069702, "train_loss_llm": 75.15199279785156, "global_step": 327, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.2160426378250122, "train_loss_dp": 1.1747119426727295, "train_loss_llm": 4.133065223693848, "global_step": 328, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.5272111892700195, "train_loss_dp": 1.4971003532409668, "train_loss_llm": 3.011080741882324, "global_step": 329, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.198259949684143, "train_loss_dp": 1.1762311458587646, "train_loss_llm": 2.2028770446777344, "global_step": 330, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 0.9671088457107544, "train_loss_dp": 0.9570469856262207, "train_loss_llm": 1.0061886310577393, "global_step": 331, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.1578236818313599, "train_loss_dp": 1.090648889541626, "train_loss_llm": 6.717479705810547, "global_step": 332, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.2753055095672607, "train_loss_dp": 1.2596111297607422, "train_loss_llm": 1.569443702697754, "global_step": 333, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.244830846786499, "train_loss_dp": 1.2119802236557007, "train_loss_llm": 3.2850570678710938, "global_step": 334, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.2530181407928467, "train_loss_dp": 1.2431042194366455, "train_loss_llm": 0.9913945198059082, "global_step": 335, "epoch": 0, "lr": 4.2000000000000004e-06}
+{"train_loss": 1.028684139251709, "train_loss_dp": 1.0209881067276, "train_loss_llm": 0.7696084976196289, "global_step": 336, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2703415155410767, "train_loss_dp": 1.2386189699172974, "train_loss_llm": 3.172260284423828, "global_step": 337, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9917907118797302, "train_loss_dp": 0.9633889198303223, "train_loss_llm": 2.840177536010742, "global_step": 338, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.2144412994384766, "train_loss_dp": 1.1925339698791504, "train_loss_llm": 2.19073486328125, "global_step": 339, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.1099491119384766, "train_loss_dp": 1.0976893901824951, "train_loss_llm": 1.225975513458252, "global_step": 340, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9008839130401611, "train_loss_dp": 0.8591746687889099, "train_loss_llm": 4.170927047729492, "global_step": 341, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.1456151008605957, "train_loss_dp": 1.136354923248291, "train_loss_llm": 0.9260196685791016, "global_step": 342, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.3753620386123657, "train_loss_dp": 1.3299955129623413, "train_loss_llm": 4.536657333374023, "global_step": 343, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.505616307258606, "train_loss_dp": 1.4726543426513672, "train_loss_llm": 3.296200752258301, "global_step": 344, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.036675214767456, "train_loss_dp": 1.0265394449234009, "train_loss_llm": 1.0135817527770996, "global_step": 345, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 1.1253588199615479, "train_loss_dp": 1.1137542724609375, "train_loss_llm": 1.1604557037353516, "global_step": 346, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9643039107322693, "train_loss_dp": 0.9567458629608154, "train_loss_llm": 0.7558053731918335, "global_step": 347, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.920651912689209, "train_loss_dp": 0.9127265214920044, "train_loss_llm": 0.7925376892089844, "global_step": 348, "epoch": 0, "lr": 4.4e-06}
+{"train_loss": 0.9552338719367981, "train_loss_dp": 0.9501639008522034, "train_loss_llm": 0.5069942474365234, "global_step": 349, "epoch": 0, "lr": 4.4e-06}