Add files using upload-large-folder tool

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+wandb/run-20250506_192633-stxr6no4/run-stxr6no4.wandb filter=lfs diff=lfs merge=lfs -text

.hydra/config.yaml

@@ -0,0 +1,84 @@
+mode:
+  is_training: true
+model:
+  name: llava-7b
+  visual_encoder:
+    name: clip-vit-large-patch14-336
+    hf_name: openai/clip-vit-large-patch14-336
+    type: hf_visual_encoder
+    hidden_size: null
+    img_size: null
+    patch_size: null
+    output_layer: -2
+    use_cls_token: false
+  llm:
+    name: vicuna-7b-v1.5
+    hf_name: lmsys/vicuna-7b-v1.5
+    type: hf_llm
+    hidden_size: null
+    vocab_size: null
+    max_seq_length: 2048
+    image_token: <image>
+    use_start_end_tokens: false
+    use_image_patch_token: false
+    image_start_token: <im_start>
+    image_end_token: <im_end>
+    image_patch_token: <im_patch>
+    system_token: null
+    user_token: null
+    assistant_token: <s>
+    ignore_index: -100
+    image_token_index: -200
+    attn_implementation: flash_attention_2
+  connector:
+    name: mlp_2_gelu
+    type: mlp
+dataset:
+  name: llava-pretrain
+  path: /pasteur/u/yiming/.cache/huggingface/hub/datasets--liuhaotian--LLaVA-Pretrain/snapshots/70f9d1e5e1a697fe35830875cfc7de1dd590d727/blip_laion_cc_sbu_558k.json
+  type: json
+  lazy_preprocess: true
+  is_multimodal: true
+  image_folder: /pasteur/u/yiming/.cache/huggingface/hub/datasets--liuhaotian--LLaVA-Pretrain/snapshots/70f9d1e5e1a697fe35830875cfc7de1dd590d727/images
+trainer:
+  output_dir: .
+  unfreeze:
+    train_visual_encoder: false
+    train_language_model: false
+    train_connector: true
+  learning_rate:
+    visual_encoder_learning_rate: 0.001
+    language_model_learning_rate: 0.001
+    connector_learning_rate: 0.001
+    default_lr: 0.001
+  weight_decay:
+    visual_encoder_weight_decay: 0.0
+    language_model_weight_decay: 0.0
+    connector_weight_decay: 0.0
+    default_wd: 0.0
+  per_device_train_batch_size: 16
+  per_device_eval_batch_size: 4
+  bf16: true
+  fp16: false
+  tf32: true
+  deepspeed: /pasteur/u/yiming/small-vlm/src/vlm/config/deepspeed/zero2.json
+  num_train_epochs: 1
+  save_strategy: steps
+  save_steps: 5000
+  save_total_limit: 1
+  logging_steps: 1
+  warmup_ratio: 0.03
+  lr_scheduler_type: cosine
+  gradient_accumulation_steps: 4
+  report_to: wandb
+  dataloader_num_workers: 4
+  version: plain
+  group_by_modality_length: false
+  gradient_checkpointing: true
+  run_name: small-vlm
+  resume_from_checkpoint: null
+  seed: 42
+inference:
+  checkpoint_path: .
+  num_inference_samples: null
+  chat_template: chat

.hydra/hydra.yaml

@@ -0,0 +1,178 @@
+hydra:
+  run:
+    dir: outputs/${now:%Y-%m-%d}/${now:%H-%M-%S}
+  sweep:
+    dir: multirun/${now:%Y-%m-%d}/${now:%H-%M-%S}
+    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:
+        (): vlm.utils.logging_utils.ColoredFormatter
+        format: '%(message)s'
+      detailed:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    filters:
+      rank_zero_filter:
+        (): vlm.utils.logging_utils.RankZeroFilter
+    handlers:
+      console:
+        class: rich.logging.RichHandler
+        formatter: simple
+        rich_tracebacks: true
+        markup: true
+        show_time: false
+        show_level: true
+        show_path: true
+        filters:
+        - rank_zero_filter
+      file:
+        class: logging.FileHandler
+        formatter: detailed
+        filename: ${hydra:job.name}.log
+        mode: w
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task: []
+  job:
+    name: vlm
+    chdir: true
+    override_dirname: ''
+    id: ???
+    num: ???
+    config_name: pretrain_llava
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /pasteur/u/yiming/small-vlm
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /pasteur/u/yiming/small-vlm/src/vlm/config
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09
+    choices:
+      dataset: llava_pretrain
+      inference: default
+      trainer: pretrain
+      trainer/weight_decay: default
+      trainer/learning_rate: default
+      trainer/unfreeze: pretrain
+      model: llava-7b
+      model/connector: mlp-2-gelu
+      model/visual_encoder: clip-vit-large-patch14-336
+      model/llm: vicuna-7b-v1.5
+      mode: train
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: rich
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: basic
+      hydra/output: default
+  verbose: false

.hydra/overrides.yaml

@@ -0,0 +1 @@
+[]

checkpoint-2180/config.json

@@ -0,0 +1,30 @@
+{
+  "architectures": [
+    "VLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 4096,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "vocab_size": 32000
+}

checkpoint-2180/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.51.3"
+}

checkpoint-2180/global_step2180/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d30fdfc9a13f31bfacbfdaf453682bec58af0f8ccd20588c397259f80acbcad0
+size 62943088

checkpoint-2180/global_step2180/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cf0e7104cb231a074431c74b0471e41674d801e5cf15522b63c527d3710b9cf
+size 62942960

checkpoint-2180/global_step2180/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b35848834ae72481b444f88d09c634f45848a602388219fe420a3b0a7fd8d582
+size 28210259341

checkpoint-2180/latest

@@ -0,0 +1 @@
+global_step2180

checkpoint-2180/model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef58ddf136827c06918f0ed7d7dda0a0d2580a5a2d26a49fa5b50b244be63300
+size 4984005016

checkpoint-2180/model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ea6e357f6235900d30d3a3ace1a6cc6471056ea38a6fcea17a51f492ff572d7
+size 4924325856

checkpoint-2180/model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d12b8fdb1499278a72a17140520bb0c23cdbcb1c5b28e8bae70895ffdd84a59
+size 4479718672

checkpoint-2180/model.safetensors.index.json

@@ -0,0 +1,694 @@
+{
+  "metadata": {
+    "total_size": 14387949568
+  },
+  "weight_map": {
+    "_connector.projection_layer.0.bias": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.0.weight": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.2.bias": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.2.weight": "model-00003-of-00003.safetensors",
+    "_language_model._embeddings.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.lm_head.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.embed_tokens.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.norm.weight": "model-00003-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.class_embedding": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.patch_embedding.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.position_embedding.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.post_layernorm.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.post_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.pre_layrnorm.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.pre_layrnorm.weight": "model-00001-of-00003.safetensors"
+  }
+}

checkpoint-2180/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f437c64f6d8ecf317df29b65ef8bbc2eda7522cdfd0cce0ba2f5ed892e40729
+size 14960

checkpoint-2180/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a21c1d04fb86211bfde9d8353d380d8f19f732c112c3fb8c230db017a3616e9a
+size 14960

checkpoint-2180/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:953edca3023325c557628c19acd1e2d515227cbdfec00cf350b3fdd9b853647f
+size 1064

checkpoint-2180/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-2180/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

checkpoint-2180/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 2048,
+  "pad_token": "<unk>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

checkpoint-2180/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48a689bb0beb39419413ccfff78cca6b5a4212394f31a837361ce56c5eaacf9a
+size 7160

checkpoint-2180/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

config.json

@@ -0,0 +1,46 @@
+{
+  "architectures": [
+    "VLM"
+  ],
+  "hidden_size": 4096,
+  "model_config_dict": {
+    "connector": {
+      "name": "mlp_2_gelu",
+      "type": "mlp"
+    },
+    "llm": {
+      "assistant_token": "<s>",
+      "hf_name": "lmsys/vicuna-7b-v1.5",
+      "hidden_size": null,
+      "ignore_index": -100,
+      "image_end_token": "<im_end>",
+      "image_patch_token": "<im_patch>",
+      "image_start_token": "<im_start>",
+      "image_token": "<image>",
+      "image_token_index": -200,
+      "max_seq_length": 2048,
+      "name": "vicuna-7b-v1.5",
+      "system_token": null,
+      "type": "hf_llm",
+      "use_image_patch_token": false,
+      "use_start_end_tokens": false,
+      "user_token": null,
+      "vocab_size": null
+    },
+    "name": "llava-7b",
+    "visual_encoder": {
+      "hf_name": "openai/clip-vit-large-patch14-336",
+      "hidden_size": null,
+      "img_size": null,
+      "name": "clip-vit-large-patch14-336",
+      "output_layer": -2,
+      "patch_size": null,
+      "type": "hf_visual_encoder",
+      "use_cls_token": false
+    }
+  },
+  "model_dtype": "bfloat16",
+  "model_type": "vlm",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3"
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.51.3"
+}

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef58ddf136827c06918f0ed7d7dda0a0d2580a5a2d26a49fa5b50b244be63300
+size 4984005016

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ea6e357f6235900d30d3a3ace1a6cc6471056ea38a6fcea17a51f492ff572d7
+size 4924325856

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d12b8fdb1499278a72a17140520bb0c23cdbcb1c5b28e8bae70895ffdd84a59
+size 4479718672

model.safetensors.index.json

@@ -0,0 +1,694 @@
+{
+  "metadata": {
+    "total_size": 14387949568
+  },
+  "weight_map": {
+    "_connector.projection_layer.0.bias": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.0.weight": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.2.bias": "model-00003-of-00003.safetensors",
+    "_connector.projection_layer.2.weight": "model-00003-of-00003.safetensors",
+    "_language_model._embeddings.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.lm_head.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.embed_tokens.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.23.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.24.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.26.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.27.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.28.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.29.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.30.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.31.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_language_model._language_model.model.norm.weight": "model-00003-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.class_embedding": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.patch_embedding.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.embeddings.position_embedding.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.layer_norm2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.post_layernorm.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.post_layernorm.weight": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.pre_layrnorm.bias": "model-00001-of-00003.safetensors",
+    "_visual_encoder._visual_encoder.pre_layrnorm.weight": "model-00001-of-00003.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 2048,
+  "pad_token": "<unk>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48a689bb0beb39419413ccfff78cca6b5a4212394f31a837361ce56c5eaacf9a
+size 7160

vlm.log

@@ -0,0 +1,31 @@
+[2025-05-06 19:26:09,133][vlm.vlm][INFO] - Global seed set to 42
+[2025-05-06 19:26:09,135][vlm.vlm][INFO] - Training mode
+[2025-05-06 19:26:09,670][vlm.vlm][INFO] - Loading model: [bold red][link=file:///pasteur/u/yiming/small-vlm/src/vlm/config/model/llava-7b.yaml]llava-7b[/link][/bold red]
+[2025-05-06 19:26:09,671][vlm.vlm][INFO] - Visual encoder: [bold cyan][link=file:///pasteur/u/yiming/small-vlm/src/vlm/config/model/visual_encoder/clip-vit-large-patch14-336.yaml]clip-vit-large-patch14-336[/link][/bold cyan]
+[2025-05-06 19:26:09,672][vlm.vlm][INFO] - LLM: [bold blue][link=file:///pasteur/u/yiming/small-vlm/src/vlm/config/model/llm/vicuna-7b-v1.5.yaml]vicuna-7b-v1.5[/link][/bold blue]
+[2025-05-06 19:26:09,673][vlm.vlm][INFO] - Connector: [bold yellow][link=file:///pasteur/u/yiming/small-vlm/src/vlm/config/model/connector/mlp_2_gelu.yaml]mlp_2_gelu[/link][/bold yellow]
+[2025-05-06 19:26:09,801][vlm.models.visual_encoders.base][INFO] - Using -2 layer as output layer
+[2025-05-06 19:26:09,803][vlm.models.visual_encoders.base][INFO] - Use cls token: False
+[2025-05-06 19:26:09,918][vlm.models.visual_encoders.base][INFO] - Hidden_size not found in config, using hf config: 1024
+[2025-05-06 19:26:09,918][vlm.models.visual_encoders.base][INFO] - Img_size not found in config, using hf config: 336
+[2025-05-06 19:26:09,919][vlm.models.visual_encoders.base][INFO] - Patch_size not found in config, using hf config: 14
+[2025-05-06 19:26:16,383][vlm.models.language_models.base][INFO] - Hidden_size not found in config, using hf config: 4096
+[2025-05-06 19:26:16,384][vlm.models.language_models.base][INFO] - Vocab_size not found in config, using hf config: 32000
+[2025-05-06 19:26:16,385][vlm.models.language_models.base][WARNING] - Max_seq_length mismatch: hf config: 4096 != config: 2048
+[2025-05-06 19:26:16,778][vlm.models.language_models.base][INFO] - Token '<s>' exists in tokenizer, ID: 1
+[2025-05-06 19:26:28,055][vlm.models.model][INFO] - Trainable parameters: 20,979,712 (0.30% of total)
+[2025-05-06 19:26:28,057][vlm.models.model][INFO] -   - visual_encoder: 0 (0.00% of 303,507,456)
+[2025-05-06 19:26:28,059][vlm.models.model][INFO] -   - language_model: 0 (0.00% of 6,738,415,616)
+[2025-05-06 19:26:28,059][vlm.models.model][INFO] -   - connector: 20,979,712 (100.00% of 20,979,712)
+[2025-05-06 19:26:28,060][vlm.train.train][INFO] - Using gradient checkpointing
+[2025-05-06 19:26:28,061][vlm.train.train][INFO] - Creating data module
+[2025-05-06 19:26:29,176][vlm.train.train][INFO] - Creating trainer
+[2025-05-06 19:26:29,186][accelerate.utils.other][WARNING] - Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommende[2025-05-06 19:26:29,561][vlm.train.optimizer][INFO] - con[2025-05-06 19:26:30,576][vlm.train.train][INFO] - Trainer: <vlm.train.vlm_trainer.VLMTrainer object at 0x7f37f15c8980>
+[2025-05-06 19:26:30,577][vlm.train.train][INFO] - Training from scratch
+[2025-05-06 19:26:31,114][vlm.train.optimizer][INFO] - configure_optimizers
+[2025-05-06 19:26:31,120][vlm.train.optim[2025-05-07 07:02:03,895][vlm.train.train][INFO] - Saving state
+[2025-05-07 07:02:03,896][vlm.train.train][INFO] - Saving model
+lr: 0.001, weight_decay: 0.0
+[2025-05-06 19:26:31,121][vlm.train.optimizer][INFO] - connector lr: 0.001, weight_decay: 0.0
+[2025-05-07 07:02:03,590][vlm.train.train][INFO] - Saving state
+[2025-05-07 07:02:03,843][vlm.train.train][INFO] - Saving model

wandb/debug-internal.log

@@ -0,0 +1,15 @@
+{"time":"2025-05-06T19:26:33.985204734-07:00","level":"INFO","msg":"stream: starting","core version":"0.19.10","symlink path":"/pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug-core.log"}
+{"time":"2025-05-06T19:26:34.170259167-07:00","level":"INFO","msg":"created new stream","id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170304588-07:00","level":"INFO","msg":"stream: started","id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170333408-07:00","level":"INFO","msg":"writer: Do: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170340018-07:00","level":"INFO","msg":"sender: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170387569-07:00","level":"INFO","msg":"handler: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.354345965-07:00","level":"INFO","msg":"Starting system monitor"}
+{"time":"2025-05-07T07:02:32.380834935-07:00","level":"INFO","msg":"stream: closing","id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.380877586-07:00","level":"INFO","msg":"Stopping system monitor"}
+{"time":"2025-05-07T07:02:32.381043048-07:00","level":"INFO","msg":"Stopped system monitor"}
+{"time":"2025-05-07T07:02:32.821863717-07:00","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-05-07T07:02:32.956426364-07:00","level":"INFO","msg":"handler: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.956481774-07:00","level":"INFO","msg":"sender: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.956471934-07:00","level":"INFO","msg":"writer: Close: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.957383798-07:00","level":"INFO","msg":"stream: closed","id":"stxr6no4"}

wandb/debug.log

@@ -0,0 +1,26 @@
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Current SDK version is 0.19.10
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Configure stats pid to 3818578
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from /afs/cs.stanford.edu/u/yiming/.config/wandb/settings
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/settings
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from environment variables
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:setup_run_log_directory():724] Logging user logs to /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug.log
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:setup_run_log_directory():725] Logging internal logs to /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug-internal.log
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:init():852] calling init triggers
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():857] wandb.init called with sweep_config: {}
+config: {'_wandb': {}}
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():893] starting backend
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():897] sending inform_init request
+2025-05-06 19:26:33,982 INFO    MainThread:3818578 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
+2025-05-06 19:26:33,982 INFO    MainThread:3818578 [wandb_init.py:init():907] backend started and connected
+2025-05-06 19:26:33,983 INFO    MainThread:3818578 [wandb_init.py:init():1002] updated telemetry
+2025-05-06 19:26:33,995 INFO    MainThread:3818578 [wandb_init.py:init():1026] communicating run to backend with 90.0 second timeout
+2025-05-06 19:26:34,351 INFO    MainThread:3818578 [wandb_init.py:init():1101] starting run threads in backend
+2025-05-06 19:26:34,492 INFO    MainThread:3818578 [wandb_run.py:_console_start():2566] atexit reg
+2025-05-06 19:26:34,492 INFO    MainThread:3818578 [wandb_run.py:_redirect():2414] redirect: wrap_raw
+2025-05-06 19:26:34,493 INFO    MainThread:3818578 [wandb_run.py:_redirect():2483] Wrapping output streams.
+2025-05-06 19:26:34,493 INFO    MainThread:3818578 [wandb_run.py:_redirect():2506] Redirects installed.
+2025-05-06 19:26:34,496 INFO    MainThread:3818578 [wandb_init.py:init():1147] run started, returning control to user process
+2025-05-06 19:26:34,498 INFO    MainThread:3818578 [wandb_run.py:_config_callback():1429] config_cb None None {'vocab_size': 32000, 'max_position_embeddings': 4096, 'hidden_size': 4096, 'intermediate_size': 11008, 'num_hidden_layers': 32, 'num_attention_heads': 32, 'num_key_value_heads': 32, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'pretraining_tp': 1, 'use_cache': True, 'rope_theta': 10000.0, 'rope_scaling': None, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 128, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'float16', 'use_bfloat16': False, 'tf_legacy_loss': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'typical_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'architectures': ['LlamaForCausalLM'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'lmsys/vicuna-7b-v1.5', '_attn_implementation_autoset': True, 'transformers_version': '4.51.3', 'model_type': 'llama', 'output_dir': '.', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': 'no', 'prediction_loss_only': False, 'per_device_train_batch_size': 16, 'per_device_eval_batch_size': 4, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 4, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 0.001, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 1, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.03, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': './runs/May06_19-26-09_pasteur6.stanford.edu', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 5000, 'save_total_limit': 1, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': 42, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': True, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 4, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'small-vlm', 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'deepspeed': '/pasteur/u/yiming/small-vlm/src/vlm/config/deepspeed/zero2.json', 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'eval_use_gather_object': False, 'average_tokens_across_devices': False, 'unfreeze_language_model': False, 'unfreeze_connector': True, 'unfreeze_visual_encoder': False, 'language_model_lr': 0.001, 'connector_lr': 0.001, 'visual_encoder_lr': 0.001, 'group_by_modality_length': False, 'language_model_wd': 0.0, 'connector_wd': 0.0, 'visual_encoder_wd': 0.0, 'version': 'plain'}
+2025-05-06 19:26:34,503 INFO    MainThread:3818578 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 7062902784 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7f360b7d5a90>>
+2025-05-06 19:26:34,503 INFO    MainThread:3818578 [wandb_run.py:_config_callback():1429] config_cb model/num_parameters 7062902784 None
+2025-05-07 07:02:32,379 INFO    MsgRouterThr:3818578 [mailbox.py:close():129] [no run ID] Closing mailbox, abandoning 1 handles.

wandb/run-20250506_192633-stxr6no4/files/config.yaml

@@ -0,0 +1,511 @@
+_attn_implementation_autoset:
+    value: true
+_name_or_path:
+    value: lmsys/vicuna-7b-v1.5
+_wandb:
+    value:
+        cli_version: 0.19.10
+        m:
+            - "1": train/loss
+              "5": 2
+              "6":
+                - 1
+                - 3
+              "7": []
+            - "1": train/global_step
+              "6":
+                - 3
+              "7": []
+            - "1": train/grad_norm
+              "5": 2
+              "6":
+                - 1
+                - 3
+              "7": []
+            - "1": train/learning_rate
+              "5": 2
+              "6":
+                - 1
+                - 3
+              "7": []
+            - "1": train/epoch
+              "5": 2
+              "6":
+                - 1
+                - 3
+              "7": []
+        python_version: 3.13.3
+        t:
+            "1":
+                - 1
+                - 11
+                - 41
+                - 49
+                - 50
+                - 55
+                - 71
+            "2":
+                - 1
+                - 11
+                - 41
+                - 49
+                - 50
+                - 55
+                - 71
+            "3":
+                - 7
+                - 13
+                - 19
+                - 23
+                - 55
+                - 62
+                - 66
+            "4": 3.13.3
+            "5": 0.19.10
+            "6": 4.51.3
+            "8":
+                - 5
+            "9":
+                "1": transformers_trainer
+            "12": 0.19.10
+            "13": linux-x86_64
+accelerator_config:
+    value:
+        dispatch_batches: null
+        even_batches: true
+        gradient_accumulation_kwargs: null
+        non_blocking: false
+        split_batches: false
+        use_seedable_sampler: true
+adafactor:
+    value: false
+adam_beta1:
+    value: 0.9
+adam_beta2:
+    value: 0.999
+adam_epsilon:
+    value: 1e-08
+add_cross_attention:
+    value: false
+architectures:
+    value:
+        - LlamaForCausalLM
+attention_bias:
+    value: false
+attention_dropout:
+    value: 0
+auto_find_batch_size:
+    value: false
+average_tokens_across_devices:
+    value: false
+bad_words_ids:
+    value: null
+batch_eval_metrics:
+    value: false
+begin_suppress_tokens:
+    value: null
+bf16:
+    value: true
+bf16_full_eval:
+    value: false
+bos_token_id:
+    value: 1
+chunk_size_feed_forward:
+    value: 0
+connector_lr:
+    value: 0.001
+connector_wd:
+    value: 0
+cross_attention_hidden_size:
+    value: null
+data_seed:
+    value: 42
+dataloader_drop_last:
+    value: false
+dataloader_num_workers:
+    value: 4
+dataloader_persistent_workers:
+    value: false
+dataloader_pin_memory:
+    value: true
+dataloader_prefetch_factor:
+    value: null
+ddp_backend:
+    value: null
+ddp_broadcast_buffers:
+    value: null
+ddp_bucket_cap_mb:
+    value: null
+ddp_find_unused_parameters:
+    value: null
+ddp_timeout:
+    value: 1800
+debug:
+    value: []
+decoder_start_token_id:
+    value: null
+deepspeed:
+    value: /pasteur/u/yiming/small-vlm/src/vlm/config/deepspeed/zero2.json
+disable_tqdm:
+    value: false
+diversity_penalty:
+    value: 0
+do_eval:
+    value: false
+do_predict:
+    value: false
+do_sample:
+    value: false
+do_train:
+    value: false
+early_stopping:
+    value: false
+encoder_no_repeat_ngram_size:
+    value: 0
+eos_token_id:
+    value: 2
+eval_accumulation_steps:
+    value: null
+eval_delay:
+    value: 0
+eval_do_concat_batches:
+    value: true
+eval_on_start:
+    value: false
+eval_steps:
+    value: null
+eval_strategy:
+    value: "no"
+eval_use_gather_object:
+    value: false
+exponential_decay_length_penalty:
+    value: null
+finetuning_task:
+    value: null
+forced_bos_token_id:
+    value: null
+forced_eos_token_id:
+    value: null
+fp16:
+    value: false
+fp16_backend:
+    value: auto
+fp16_full_eval:
+    value: false
+fp16_opt_level:
+    value: O1
+fsdp:
+    value: []
+fsdp_config:
+    value:
+        min_num_params: 0
+        xla: false
+        xla_fsdp_grad_ckpt: false
+        xla_fsdp_v2: false
+fsdp_min_num_params:
+    value: 0
+fsdp_transformer_layer_cls_to_wrap:
+    value: null
+full_determinism:
+    value: false
+gradient_accumulation_steps:
+    value: 4
+gradient_checkpointing:
+    value: true
+gradient_checkpointing_kwargs:
+    value: null
+greater_is_better:
+    value: null
+group_by_length:
+    value: false
+group_by_modality_length:
+    value: false
+half_precision_backend:
+    value: auto
+head_dim:
+    value: 128
+hidden_act:
+    value: silu
+hidden_size:
+    value: 4096
+hub_always_push:
+    value: false
+hub_model_id:
+    value: null
+hub_private_repo:
+    value: null
+hub_strategy:
+    value: every_save
+hub_token:
+    value: <HUB_TOKEN>
+id2label:
+    value:
+        "0": LABEL_0
+        "1": LABEL_1
+ignore_data_skip:
+    value: false
+include_for_metrics:
+    value: []
+include_inputs_for_metrics:
+    value: false
+include_num_input_tokens_seen:
+    value: false
+include_tokens_per_second:
+    value: false
+initializer_range:
+    value: 0.02
+intermediate_size:
+    value: 11008
+is_decoder:
+    value: false
+is_encoder_decoder:
+    value: false
+jit_mode_eval:
+    value: false
+label_names:
+    value: null
+label_smoothing_factor:
+    value: 0
+label2id:
+    value:
+        LABEL_0: 0
+        LABEL_1: 1
+language_model_lr:
+    value: 0.001
+language_model_wd:
+    value: 0
+learning_rate:
+    value: 0.001
+length_column_name:
+    value: length
+length_penalty:
+    value: 1
+load_best_model_at_end:
+    value: false
+local_rank:
+    value: 0
+log_level:
+    value: passive
+log_level_replica:
+    value: warning
+log_on_each_node:
+    value: true
+logging_dir:
+    value: ./runs/May06_19-26-09_pasteur6.stanford.edu
+logging_first_step:
+    value: false
+logging_nan_inf_filter:
+    value: true
+logging_steps:
+    value: 1
+logging_strategy:
+    value: steps
+lr_scheduler_type:
+    value: cosine
+max_grad_norm:
+    value: 1
+max_length:
+    value: 20
+max_position_embeddings:
+    value: 4096
+max_steps:
+    value: -1
+metric_for_best_model:
+    value: null
+min_length:
+    value: 0
+mlp_bias:
+    value: false
+model/num_parameters:
+    value: 7062902784
+model_type:
+    value: llama
+mp_parameters:
+    value: ""
+neftune_noise_alpha:
+    value: null
+no_cuda:
+    value: false
+no_repeat_ngram_size:
+    value: 0
+num_attention_heads:
+    value: 32
+num_beam_groups:
+    value: 1
+num_beams:
+    value: 1
+num_hidden_layers:
+    value: 32
+num_key_value_heads:
+    value: 32
+num_return_sequences:
+    value: 1
+num_train_epochs:
+    value: 1
+optim:
+    value: adamw_torch
+optim_args:
+    value: null
+optim_target_modules:
+    value: null
+output_attentions:
+    value: false
+output_dir:
+    value: .
+output_hidden_states:
+    value: false
+output_scores:
+    value: false
+overwrite_output_dir:
+    value: false
+pad_token_id:
+    value: 0
+past_index:
+    value: -1
+per_device_eval_batch_size:
+    value: 4
+per_device_train_batch_size:
+    value: 16
+per_gpu_eval_batch_size:
+    value: null
+per_gpu_train_batch_size:
+    value: null
+prediction_loss_only:
+    value: false
+prefix:
+    value: null
+pretraining_tp:
+    value: 1
+problem_type:
+    value: null
+push_to_hub:
+    value: false
+push_to_hub_model_id:
+    value: null
+push_to_hub_organization:
+    value: null
+push_to_hub_token:
+    value: <PUSH_TO_HUB_TOKEN>
+ray_scope:
+    value: last
+remove_invalid_values:
+    value: false
+remove_unused_columns:
+    value: false
+repetition_penalty:
+    value: 1
+report_to:
+    value:
+        - wandb
+restore_callback_states_from_checkpoint:
+    value: false
+resume_from_checkpoint:
+    value: null
+return_dict:
+    value: true
+return_dict_in_generate:
+    value: false
+rms_norm_eps:
+    value: 1e-05
+rope_scaling:
+    value: null
+rope_theta:
+    value: 10000
+run_name:
+    value: small-vlm
+save_on_each_node:
+    value: false
+save_only_model:
+    value: false
+save_safetensors:
+    value: true
+save_steps:
+    value: 5000
+save_strategy:
+    value: steps
+save_total_limit:
+    value: 1
+seed:
+    value: 42
+sep_token_id:
+    value: null
+skip_memory_metrics:
+    value: true
+suppress_tokens:
+    value: null
+task_specific_params:
+    value: null
+temperature:
+    value: 1
+tf_legacy_loss:
+    value: false
+tf32:
+    value: true
+tie_encoder_decoder:
+    value: false
+tie_word_embeddings:
+    value: false
+tokenizer_class:
+    value: null
+top_k:
+    value: 50
+top_p:
+    value: 1
+torch_compile:
+    value: false
+torch_compile_backend:
+    value: null
+torch_compile_mode:
+    value: null
+torch_dtype:
+    value: float16
+torch_empty_cache_steps:
+    value: null
+torchdynamo:
+    value: null
+torchscript:
+    value: false
+tp_size:
+    value: 0
+tpu_metrics_debug:
+    value: false
+tpu_num_cores:
+    value: null
+transformers_version:
+    value: 4.51.3
+typical_p:
+    value: 1
+unfreeze_connector:
+    value: true
+unfreeze_language_model:
+    value: false
+unfreeze_visual_encoder:
+    value: false
+use_bfloat16:
+    value: false
+use_cache:
+    value: true
+use_cpu:
+    value: false
+use_ipex:
+    value: false
+use_legacy_prediction_loop:
+    value: false
+use_liger_kernel:
+    value: false
+use_mps_device:
+    value: false
+version:
+    value: plain
+visual_encoder_lr:
+    value: 0.001
+visual_encoder_wd:
+    value: 0
+vocab_size:
+    value: 32000
+warmup_ratio:
+    value: 0.03
+warmup_steps:
+    value: 0
+weight_decay:
+    value: 0

wandb/run-20250506_192633-stxr6no4/files/requirements.txt

@@ -0,0 +1,129 @@
+nest-asyncio==1.6.0
+typing-inspection==0.4.0
+ipython==9.2.0
+setproctitle==1.3.6
+hjson==3.1.0
+tokenizers==0.21.0
+deepspeed==0.16.7
+gitdb==4.0.12
+safetensors==0.5.3
+sentencepiece==0.2.0
+accelerate==1.6.0
+nvidia-nvjitlink-cu12==12.4.127
+jupyter_client==8.6.3
+pyzmq==26.4.0
+funlog==0.2.1
+docker-pycreds==0.4.0
+urllib3==2.4.0
+py-cpuinfo==9.0.0
+nvidia-cusolver-cu12==11.6.1.9
+asttokens==3.0.0
+triton==3.2.0
+antlr4-python3-runtime==4.9.3
+torchvision==0.21.0
+nvidia-cusparse-cu12==12.3.1.170
+mpi4py==4.0.3
+packaging==24.2
+ruff==0.11.8
+huggingface-hub==0.30.2
+Jinja2==3.1.6
+mpmath==1.3.0
+smmap==5.0.2
+MarkupSafe==2.1.5
+icecream==2.1.4
+lxml==5.4.0
+msgpack==1.1.0
+basedpyright==1.29.1
+stack-data==0.6.3
+nvidia-cudnn-cu12==9.1.0.70
+networkx==3.4.2
+torch==2.6.0
+PyYAML==6.0.2
+pycryptodomex==3.22.0
+regex==2024.11.6
+wandb==0.19.10
+nvidia-cusparselt-cu12==0.6.2
+blobfile==3.0.0
+jedi==0.19.2
+requests==2.32.3
+nvidia-nvtx-cu12==12.4.127
+psutil==7.0.0
+matplotlib-inline==0.1.7
+mdurl==0.1.2
+python-dateutil==2.9.0.post0
+small-vlm==0.5.3.dev13+90b5328
+platformdirs==4.3.7
+setuptools==80.3.1
+charset-normalizer==3.4.2
+nvidia-cuda-cupti-cu12==12.4.127
+tiktoken==0.9.0
+typing_extensions==4.13.2
+executing==2.2.0
+codespell==2.4.1
+hydra-core==1.3.2
+nvidia-cublas-cu12==12.4.5.8
+pure_eval==0.2.3
+nodejs-wheel-binaries==22.15.0
+colorama==0.4.6
+comm==0.2.2
+ipython_pygments_lexers==1.1.1
+filelock==3.18.0
+tornado==6.4.2
+six==1.17.0
+traitlets==5.14.3
+sympy==1.13.1
+pillow==11.2.1
+pytest==8.3.5
+pexpect==4.9.0
+iniconfig==2.1.0
+nvidia-cuda-nvrtc-cu12==12.4.127
+sentry-sdk==2.27.0
+pydantic==2.11.4
+click==8.1.8
+GitPython==3.1.44
+tqdm==4.67.1
+protobuf==6.30.2
+omegaconf==2.3.0
+uv==0.7.2
+nvidia-curand-cu12==10.3.5.147
+annotated-types==0.7.0
+jupyter_core==5.7.2
+ninja==1.11.1.4
+ptyprocess==0.7.0
+einops==0.6.1
+certifi==2025.4.26
+wcwidth==0.2.13
+prompt_toolkit==3.0.51
+Pygments==2.19.1
+parso==0.8.4
+numpy==1.26.4
+ipykernel==6.29.5
+transformers==4.51.3
+nvidia-cufft-cu12==11.2.1.3
+nvidia-cuda-runtime-cu12==12.4.127
+pluggy==1.5.0
+flash_attn==2.7.4.post1
+nvidia-nccl-cu12==2.21.5
+rich==14.0.0
+markdown-it-py==3.0.0
+idna==3.10
+pydantic_core==2.33.2
+fsspec==2025.3.2
+decorator==5.2.1
+debugpy==1.8.14
+inflect==7.3.1
+typing_extensions==4.12.2
+jaraco.text==3.12.1
+autocommand==2.2.2
+zipp==3.19.2
+more-itertools==10.3.0
+wheel==0.45.1
+backports.tarfile==1.2.0
+jaraco.context==5.3.0
+tomli==2.0.1
+jaraco.functools==4.0.1
+jaraco.collections==5.1.0
+importlib_metadata==8.0.0
+typeguard==4.3.0
+packaging==24.2
+platformdirs==4.2.2

wandb/run-20250506_192633-stxr6no4/files/wandb-metadata.json

@@ -0,0 +1,65 @@
+{
+  "os": "Linux-5.4.0-156-generic-x86_64-with-glibc2.31",
+  "python": "CPython 3.13.3",
+  "startedAt": "2025-05-07T02:26:33.982740Z",
+  "args": [
+    "-cn",
+    "pretrain_llava"
+  ],
+  "program": "-m vlm.__main__",
+  "git": {
+    "remote": "git@github.com:leo1oel/small-vlm.git",
+    "commit": "6ea10092d37b8118317afe26722e383941b23d44"
+  },
+  "email": "liuym13579@gmail.com",
+  "root": "/pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09",
+  "host": "pasteur6.stanford.edu",
+  "executable": "/pasteur/u/yiming/small-vlm/.venv/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA RTX A6000",
+  "gpu_count": 4,
+  "disk": {
+    "/": {
+      "total": "470306373632",
+      "used": "119354068992"
+    }
+  },
+  "memory": {
+    "total": "811303198720"
+  },
+  "cpu": {
+    "count": 64,
+    "countLogical": 128
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA RTX A6000",
+      "memoryTotal": "51527024640",
+      "cudaCores": 10752,
+      "architecture": "Ampere"
+    },
+    {
+      "name": "NVIDIA RTX A6000",
+      "memoryTotal": "51527024640",
+      "cudaCores": 10752,
+      "architecture": "Ampere"
+    },
+    {
+      "name": "NVIDIA RTX A6000",
+      "memoryTotal": "51527024640",
+      "cudaCores": 10752,
+      "architecture": "Ampere"
+    },
+    {
+      "name": "NVIDIA RTX A6000",
+      "memoryTotal": "51527024640",
+      "cudaCores": 10752,
+      "architecture": "Ampere"
+    }
+  ],
+  "slurm": {
+    "job_id": "10312591"
+  },
+  "cudaVersion": "12.2"
+}

wandb/run-20250506_192633-stxr6no4/files/wandb-summary.json

@@ -0,0 +1 @@
+{"train/grad_norm":0.3432144820690155,"train/global_step":2180,"train_runtime":41729.1956,"train_samples_per_second":13.375,"_wandb":{"runtime":41758},"train/epoch":0.9998853342506593,"_timestamp":1.7466265227935104e+09,"_runtime":41728.810855759,"_step":2180,"train/learning_rate":5.521152057763601e-10,"train_steps_per_second":0.052,"total_flos":6.815296424813527e+17,"train_loss":2.1650498506672884,"train/loss":1.9774}

wandb/run-20250506_192633-stxr6no4/logs/debug-core.log

@@ -0,0 +1,13 @@
+{"time":"2025-05-06T19:26:33.634523111-07:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/user/24772/tmp_fgvwime/port-3818578.txt","pid":3818578,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false}
+{"time":"2025-05-06T19:26:33.635248223-07:00","level":"INFO","msg":"Will exit if parent process dies.","ppid":3818578}
+{"time":"2025-05-06T19:26:33.635251453-07:00","level":"INFO","msg":"server is running","addr":{"IP":"127.0.0.1","Port":39483,"Zone":""}}
+{"time":"2025-05-06T19:26:33.822819056-07:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"127.0.0.1:49678"}
+{"time":"2025-05-06T19:26:33.984266749-07:00","level":"INFO","msg":"handleInformInit: received","streamId":"stxr6no4","id":"127.0.0.1:49678"}
+{"time":"2025-05-06T19:26:34.170311618-07:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"stxr6no4","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.380436789-07:00","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.380805794-07:00","level":"INFO","msg":"connection: closing","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.380833945-07:00","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-05-07T07:02:32.380960997-07:00","level":"INFO","msg":"connection: closed successfully","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.95751064-07:00","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.957543771-07:00","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"127.0.0.1:49678"}
+{"time":"2025-05-07T07:02:32.957559001-07:00","level":"INFO","msg":"server is closed"}

wandb/run-20250506_192633-stxr6no4/logs/debug-internal.log

@@ -0,0 +1,15 @@
+{"time":"2025-05-06T19:26:33.985204734-07:00","level":"INFO","msg":"stream: starting","core version":"0.19.10","symlink path":"/pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug-core.log"}
+{"time":"2025-05-06T19:26:34.170259167-07:00","level":"INFO","msg":"created new stream","id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170304588-07:00","level":"INFO","msg":"stream: started","id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170333408-07:00","level":"INFO","msg":"writer: Do: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170340018-07:00","level":"INFO","msg":"sender: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.170387569-07:00","level":"INFO","msg":"handler: started","stream_id":"stxr6no4"}
+{"time":"2025-05-06T19:26:34.354345965-07:00","level":"INFO","msg":"Starting system monitor"}
+{"time":"2025-05-07T07:02:32.380834935-07:00","level":"INFO","msg":"stream: closing","id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.380877586-07:00","level":"INFO","msg":"Stopping system monitor"}
+{"time":"2025-05-07T07:02:32.381043048-07:00","level":"INFO","msg":"Stopped system monitor"}
+{"time":"2025-05-07T07:02:32.821863717-07:00","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-05-07T07:02:32.956426364-07:00","level":"INFO","msg":"handler: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.956481774-07:00","level":"INFO","msg":"sender: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.956471934-07:00","level":"INFO","msg":"writer: Close: closed","stream_id":"stxr6no4"}
+{"time":"2025-05-07T07:02:32.957383798-07:00","level":"INFO","msg":"stream: closed","id":"stxr6no4"}

wandb/run-20250506_192633-stxr6no4/logs/debug.log

@@ -0,0 +1,26 @@
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Current SDK version is 0.19.10
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Configure stats pid to 3818578
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from /afs/cs.stanford.edu/u/yiming/.config/wandb/settings
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/settings
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_setup.py:_flush():68] Loading settings from environment variables
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:setup_run_log_directory():724] Logging user logs to /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug.log
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:setup_run_log_directory():725] Logging internal logs to /pasteur/u/yiming/small-vlm/outputs/2025-05-06/19-26-09/wandb/run-20250506_192633-stxr6no4/logs/debug-internal.log
+2025-05-06 19:26:33,978 INFO    MainThread:3818578 [wandb_init.py:init():852] calling init triggers
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():857] wandb.init called with sweep_config: {}
+config: {'_wandb': {}}
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():893] starting backend
+2025-05-06 19:26:33,979 INFO    MainThread:3818578 [wandb_init.py:init():897] sending inform_init request
+2025-05-06 19:26:33,982 INFO    MainThread:3818578 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
+2025-05-06 19:26:33,982 INFO    MainThread:3818578 [wandb_init.py:init():907] backend started and connected
+2025-05-06 19:26:33,983 INFO    MainThread:3818578 [wandb_init.py:init():1002] updated telemetry
+2025-05-06 19:26:33,995 INFO    MainThread:3818578 [wandb_init.py:init():1026] communicating run to backend with 90.0 second timeout
+2025-05-06 19:26:34,351 INFO    MainThread:3818578 [wandb_init.py:init():1101] starting run threads in backend
+2025-05-06 19:26:34,492 INFO    MainThread:3818578 [wandb_run.py:_console_start():2566] atexit reg
+2025-05-06 19:26:34,492 INFO    MainThread:3818578 [wandb_run.py:_redirect():2414] redirect: wrap_raw
+2025-05-06 19:26:34,493 INFO    MainThread:3818578 [wandb_run.py:_redirect():2483] Wrapping output streams.
+2025-05-06 19:26:34,493 INFO    MainThread:3818578 [wandb_run.py:_redirect():2506] Redirects installed.
+2025-05-06 19:26:34,496 INFO    MainThread:3818578 [wandb_init.py:init():1147] run started, returning control to user process
+2025-05-06 19:26:34,498 INFO    MainThread:3818578 [wandb_run.py:_config_callback():1429] config_cb None None {'vocab_size': 32000, 'max_position_embeddings': 4096, 'hidden_size': 4096, 'intermediate_size': 11008, 'num_hidden_layers': 32, 'num_attention_heads': 32, 'num_key_value_heads': 32, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'pretraining_tp': 1, 'use_cache': True, 'rope_theta': 10000.0, 'rope_scaling': None, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 128, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'float16', 'use_bfloat16': False, 'tf_legacy_loss': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'typical_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'architectures': ['LlamaForCausalLM'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'lmsys/vicuna-7b-v1.5', '_attn_implementation_autoset': True, 'transformers_version': '4.51.3', 'model_type': 'llama', 'output_dir': '.', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': False, 'do_predict': False, 'eval_strategy': 'no', 'prediction_loss_only': False, 'per_device_train_batch_size': 16, 'per_device_eval_batch_size': 4, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 4, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 0.001, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 1, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': {}, 'warmup_ratio': 0.03, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': './runs/May06_19-26-09_pasteur6.stanford.edu', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 5000, 'save_total_limit': 1, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': 42, 'jit_mode_eval': False, 'use_ipex': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': True, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': None, 'dataloader_num_workers': 4, 'dataloader_prefetch_factor': None, 'past_index': -1, 'run_name': 'small-vlm', 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'tp_size': 0, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'deepspeed': '/pasteur/u/yiming/small-vlm/src/vlm/config/deepspeed/zero2.json', 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'gradient_checkpointing': True, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': False, 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'eval_use_gather_object': False, 'average_tokens_across_devices': False, 'unfreeze_language_model': False, 'unfreeze_connector': True, 'unfreeze_visual_encoder': False, 'language_model_lr': 0.001, 'connector_lr': 0.001, 'visual_encoder_lr': 0.001, 'group_by_modality_length': False, 'language_model_wd': 0.0, 'connector_wd': 0.0, 'visual_encoder_wd': 0.0, 'version': 'plain'}
+2025-05-06 19:26:34,503 INFO    MainThread:3818578 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 7062902784 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7f360b7d5a90>>
+2025-05-06 19:26:34,503 INFO    MainThread:3818578 [wandb_run.py:_config_callback():1429] config_cb model/num_parameters 7062902784 None
+2025-05-07 07:02:32,379 INFO    MsgRouterThr:3818578 [mailbox.py:close():129] [no run ID] Closing mailbox, abandoning 1 handles.

wandb/run-20250506_192633-stxr6no4/run-stxr6no4.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d5ad125ad3ee65f243c5b574d4775e05a4fab340d94f16ca9210b17eb1182ce
+size 9041199