API Reference

This page contains auto-generated API reference documentation [1].

• bridge
  • bridge.training
    • bridge.training.mlm_compat
      • bridge.training.mlm_compat.arguments
      • bridge.training.mlm_compat.activations
      • bridge.training.mlm_compat.model
    • bridge.training.tokenizers
      • bridge.training.tokenizers.tokenizer
      • bridge.training.tokenizers.config
    • bridge.training.utils
      • bridge.training.utils.theoretical_memory_utils
      • bridge.training.utils.pg_utils
      • bridge.training.utils.config_utils
      • bridge.training.utils.checkpoint_utils
      • bridge.training.utils.flop_utils
      • bridge.training.utils.moe_token_drop
      • bridge.training.utils.mlflow_utils
      • bridge.training.utils.log_utils
      • bridge.training.utils.train_utils
      • bridge.training.utils.wandb_utils
      • bridge.training.utils.sig_utils
      • bridge.training.utils.padding_utils
      • bridge.training.utils.visual_inputs
      • bridge.training.utils.omegaconf_utils
      • bridge.training.utils.batch_utils
      • bridge.training.utils.packed_seq_utils
    • bridge.training.post_training
      • bridge.training.post_training.distillation
      • bridge.training.post_training.checkpointing
    • bridge.training.losses
    • bridge.training.callbacks
    • bridge.training.inprocess_restart
    • bridge.training.profiling
    • bridge.training.flex_dispatcher_backend
    • bridge.training.eval
    • bridge.training.model_load_save
    • bridge.training.comm_overlap
    • bridge.training.pretrain
    • bridge.training.distill
    • bridge.training.setup
    • bridge.training.state
    • bridge.training.llava_step
    • bridge.training.train
    • bridge.training.config
    • bridge.training.tensor_inspect
    • bridge.training.nvrx_straggler
    • bridge.training.vlm_step
    • bridge.training.initialize
    • bridge.training.optim
    • bridge.training.mixed_precision
    • bridge.training.finetune
    • bridge.training.gpt_step
    • bridge.training.forward_step_func_types
    • bridge.training.checkpointing
    • bridge.training.fault_tolerance
  • bridge.peft
    • bridge.peft.dora_layers
    • bridge.peft.lora
    • bridge.peft.base
    • bridge.peft.dora
    • bridge.peft.utils
    • bridge.peft.module_matcher
    • bridge.peft.adapter_wrapper
    • bridge.peft.lora_layers
    • bridge.peft.canonical_lora
    • bridge.peft.walk_utils
    • bridge.peft.recompute
  • bridge.utils
    • bridge.utils.vocab_utils
    • bridge.utils.common_utils
    • bridge.utils.yaml_utils
    • bridge.utils.import_utils
    • bridge.utils.fusions
    • bridge.utils.instantiate_utils
    • bridge.utils.decorators
    • bridge.utils.slurm_utils
  • bridge.recipes
    • bridge.recipes.qwen_vl
      • bridge.recipes.qwen_vl.qwen3_vl
      • bridge.recipes.qwen_vl.qwen25_vl
      • bridge.recipes.qwen_vl.qwen25_vl_dataset
    • bridge.recipes.qwen
      • bridge.recipes.qwen.qwen2
      • bridge.recipes.qwen.qwen3_next
      • bridge.recipes.qwen.qwen3_moe
      • bridge.recipes.qwen.qwen3
    • bridge.recipes.gpt
      • bridge.recipes.gpt.gpt3_175b
    • bridge.recipes.gemma
      • bridge.recipes.gemma.gemma2
      • bridge.recipes.gemma.gemma3
    • bridge.recipes.glm
      • bridge.recipes.glm.glm45
    • bridge.recipes.deepseek
      • bridge.recipes.deepseek.deepseek_v2
      • bridge.recipes.deepseek.deepseek_v3
    • bridge.recipes.utils
      • bridge.recipes.utils.tokenizer_utils
      • bridge.recipes.utils.finetune_utils
      • bridge.recipes.utils.optimizer_utils
      • bridge.recipes.utils.dataset_utils
    • bridge.recipes.olmoe
      • bridge.recipes.olmoe.olmoe_7b
    • bridge.recipes.gemma3_vl
      • bridge.recipes.gemma3_vl.gemma3_vl
    • bridge.recipes.glm_vl
      • bridge.recipes.glm_vl.glm_45v
    • bridge.recipes.moonlight
      • bridge.recipes.moonlight.moonlight_16b
    • bridge.recipes.nemotronh
      • bridge.recipes.nemotronh.nemotron_nano_v2
      • bridge.recipes.nemotronh.nemotron_3_nano
      • bridge.recipes.nemotronh.nemotronh
    • bridge.recipes.nemotron_vl
      • bridge.recipes.nemotron_vl.nemotron_nano_v2_vl
    • bridge.recipes.gpt_oss
      • bridge.recipes.gpt_oss.gpt_oss
    • bridge.recipes.llama
      • bridge.recipes.llama.llama2
      • bridge.recipes.llama.llama3
    • bridge.recipes.ministral3
      • bridge.recipes.ministral3.ministral3
    • bridge.recipes.common
    • bridge.recipes.run_plugins
  • bridge.data
    • bridge.data.datasets
      • bridge.data.datasets.utils
      • bridge.data.datasets.packing_utils
      • bridge.data.datasets.packed_sequence
      • bridge.data.datasets.fim_dataset
      • bridge.data.datasets.sft
    • bridge.data.hf_processors
      • bridge.data.hf_processors.squad
    • bridge.data.builders
      • bridge.data.builders.hf_dataset
      • bridge.data.builders.finetuning_dataset
    • bridge.data.vlm_datasets
      • bridge.data.vlm_datasets.token_utils
      • bridge.data.vlm_datasets.collate
      • bridge.data.vlm_datasets.conversation_dataset
      • bridge.data.vlm_datasets.hf_provider
      • bridge.data.vlm_datasets.mock_provider
      • bridge.data.vlm_datasets.preloaded_provider
      • bridge.data.vlm_datasets.hf_dataset_makers
    • bridge.data.iterator_utils
    • bridge.data.finetuning
    • bridge.data.utils
    • bridge.data.loaders
    • bridge.data.samplers
  • bridge.models
    • bridge.models.common
      • bridge.models.common.base
      • bridge.models.common.unimodal
    • bridge.models.qwen_vl
      • bridge.models.qwen_vl.modelling_qwen3_vl
        • bridge.models.qwen_vl.modelling_qwen3_vl.transformer_config
        • bridge.models.qwen_vl.modelling_qwen3_vl.utils
        • bridge.models.qwen_vl.modelling_qwen3_vl.attention
        • bridge.models.qwen_vl.modelling_qwen3_vl.vision_model
        • bridge.models.qwen_vl.modelling_qwen3_vl.model
        • bridge.models.qwen_vl.modelling_qwen3_vl.text_model
        • bridge.models.qwen_vl.modelling_qwen3_vl.rope
        • bridge.models.qwen_vl.modelling_qwen3_vl.transformer_block
      • bridge.models.qwen_vl.modeling_qwen25_vl
      • bridge.models.qwen_vl.qwen3_vl_step
      • bridge.models.qwen_vl.qwen25_vl_bridge
      • bridge.models.qwen_vl.qwen25_vl_provider
      • bridge.models.qwen_vl.qwen35_vl_bridge
      • bridge.models.qwen_vl.qwen35_vl_provider
      • bridge.models.qwen_vl.qwen3_vl_provider
      • bridge.models.qwen_vl.qwen3_vl_bridge
    • bridge.models.mamba
      • bridge.models.mamba.mamba_builder
      • bridge.models.mamba.mamba_provider
    • bridge.models.mimo
      • bridge.models.mimo.mimo_config
      • bridge.models.mimo.mimo_bridge
      • bridge.models.mimo.mimo_provider
      • bridge.models.mimo.llava_provider
      • bridge.models.mimo.mimo_builder
    • bridge.models.qwen
      • bridge.models.qwen.qwen3_bridge
      • bridge.models.qwen.qwen3_moe_bridge
      • bridge.models.qwen.qwen_provider
      • bridge.models.qwen.qwen2_bridge
      • bridge.models.qwen.qwen3_next_bridge
    • bridge.models.gemma
      • bridge.models.gemma.gemma_provider
      • bridge.models.gemma.gemma_bridge
      • bridge.models.gemma.gemma3_bridge
      • bridge.models.gemma.gemma3_provider
      • bridge.models.gemma.gemma2_bridge
      • bridge.models.gemma.gemma2_provider
      • bridge.models.gemma.modules
    • bridge.models.gemma_vl
      • bridge.models.gemma_vl.gemma3_vl_bridge
      • bridge.models.gemma_vl.gemma3_vl_provider
      • bridge.models.gemma_vl.modeling_gemma3_vl
    • bridge.models.glm
      • bridge.models.glm.glm45_provider
      • bridge.models.glm.glm_moe_mappings
      • bridge.models.glm.glm45_bridge
    • bridge.models.deepseek
      • bridge.models.deepseek.deepseek_v3_bridge
      • bridge.models.deepseek.common
      • bridge.models.deepseek.deepseek_v2_bridge
      • bridge.models.deepseek.deepseek_provider
    • bridge.models.llama_nemotron
      • bridge.models.llama_nemotron.llama_nemotron_bridge
      • bridge.models.llama_nemotron.llama_nemotron_provider
    • bridge.models.olmoe
      • bridge.models.olmoe.olmoe_bridge
      • bridge.models.olmoe.olmoe_provider
    • bridge.models.conversion
      • bridge.models.conversion.param_mapping
      • bridge.models.conversion.mapping_registry
      • bridge.models.conversion.peft_bridge
      • bridge.models.conversion.utils
      • bridge.models.conversion.transformers_compat
      • bridge.models.conversion.model_bridge
      • bridge.models.conversion.auto_bridge
    • bridge.models.nemotron
      • bridge.models.nemotron.nemotron_provider
      • bridge.models.nemotron.nemotron_bridge
    • bridge.models.mistral
      • bridge.models.mistral.mistral_bridge
      • bridge.models.mistral.mistral_provider
    • bridge.models.decorators
      • bridge.models.decorators.dispatch
      • bridge.models.decorators.torchrun
    • bridge.models.glm_vl
      • bridge.models.glm_vl.glm_45v_bridge
      • bridge.models.glm_vl.modeling_glm_45v
      • bridge.models.glm_vl.glm_45v_provider
    • bridge.models.nemotronh
      • bridge.models.nemotronh.nemotron_h_bridge
      • bridge.models.nemotronh.nemotron_h_provider
    • bridge.models.nemotron_vl
      • bridge.models.nemotron_vl.nemotron_vl_bridge
      • bridge.models.nemotron_vl.modeling_nemotron_vl
      • bridge.models.nemotron_vl.nemotron_vl_provider
      • bridge.models.nemotron_vl.nemotron_vl_utils
    • bridge.models.hf_pretrained
      • bridge.models.hf_pretrained.vlm
      • bridge.models.hf_pretrained.base
      • bridge.models.hf_pretrained.utils
      • bridge.models.hf_pretrained.state
      • bridge.models.hf_pretrained.safe_config_loader
      • bridge.models.hf_pretrained.causal_lm
    • bridge.models.gpt_oss
      • bridge.models.gpt_oss.gpt_oss_bridge
      • bridge.models.gpt_oss.gpt_oss_provider
    • bridge.models.kimi
      • bridge.models.kimi.kimi_provider
    • bridge.models.llama
      • bridge.models.llama.llama_provider
      • bridge.models.llama.llama_bridge
      • bridge.models.llama.llama4_utils
    • bridge.models.ministral3
      • bridge.models.ministral3.modeling_ministral3
      • bridge.models.ministral3.ministral3_bridge
      • bridge.models.ministral3.ministral3_provider
    • bridge.models.model_provider
    • bridge.models.gpt_full_te_layer_autocast_spec
    • bridge.models.transformer_config
    • bridge.models.mla_provider
    • bridge.models.config
    • bridge.models.distillation_provider
    • bridge.models.gpt_provider
    • bridge.models.t5_provider
  • bridge.package_info

[1]

Created with sphinx-autodoc2