bridge.recipes.nemotronh.nemotronh

Module Contents

Functions

nemotronh_4b_pretrain_config	Return a pre-training config for NemotronH 4B.
nemotronh_8b_pretrain_config	Return a pre-training config for NemotronH 8B.
nemotronh_47b_pretrain_config	Return a pre-training config for NemotronH 47B.
nemotronh_56b_pretrain_config	Return a pre-training config for NemotronH 56B.
nemotronh_4b_sft_config	Return a full SFT config for NemotronH 4B.
nemotronh_8b_sft_config	Return a full SFT config for NemotronH 8B.
nemotronh_47b_sft_config	Return a full SFT config for NemotronH 47B.
nemotronh_56b_sft_config	Return a full SFT config for NemotronH 56B.
nemotronh_4b_peft_config	Return a PEFT config for NemotronH 4B.
nemotronh_8b_peft_config	Return a PEFT config for NemotronH 8B.
nemotronh_47b_peft_config	Return a PEFT config for NemotronH 47B.
nemotronh_56b_peft_config	Return a PEFT config for NemotronH 56B.

Data

__all__

API

bridge.recipes.nemotronh.nemotronh.nemotronh_4b_pretrain_config() → megatron.bridge.training.config.ConfigContainer

Return a pre-training config for NemotronH 4B.

This recipe is designed for single-node training (1 node). Default parallelism: TP=1, PP=1, SP=False.

bridge.recipes.nemotronh.nemotronh.nemotronh_8b_pretrain_config() → megatron.bridge.training.config.ConfigContainer

Return a pre-training config for NemotronH 8B.

This recipe is designed for single-node training (1 node). Default parallelism: TP=2, PP=1, SP=True.

bridge.recipes.nemotronh.nemotronh.nemotronh_47b_pretrain_config() → megatron.bridge.training.config.ConfigContainer

Return a pre-training config for NemotronH 47B.

This recipe is designed for single-node training (1 node with 8 GPUs). Default parallelism: TP=8, PP=1, SP=True.

Note: Uses FP8 precision by default.

bridge.recipes.nemotronh.nemotronh.nemotronh_56b_pretrain_config() → megatron.bridge.training.config.ConfigContainer

Return a pre-training config for NemotronH 56B.

This recipe is designed for single-node training (1 node with 8 GPUs). Default parallelism: TP=8, PP=1, SP=True.

Note: Uses FP8 precision by default.

bridge.recipes.nemotronh.nemotronh.nemotronh_4b_sft_config() → megatron.bridge.training.config.ConfigContainer

Return a full SFT config for NemotronH 4B.

Default parallelism: TP=1, PP=1, SP=False

Returns:

ConfigContainer with all settings pre-configured for NemotronH 4B SFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_8b_sft_config() → megatron.bridge.training.config.ConfigContainer

Return a full SFT config for NemotronH 8B.

Default parallelism: TP=2, PP=1, SP=True

Returns:

ConfigContainer with all settings pre-configured for NemotronH 8B SFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_47b_sft_config() → megatron.bridge.training.config.ConfigContainer

Return a full SFT config for NemotronH 47B.

Default parallelism: TP=8, PP=2, SP=True

Returns:

ConfigContainer with all settings pre-configured for NemotronH 47B SFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_56b_sft_config() → megatron.bridge.training.config.ConfigContainer

Return a full SFT config for NemotronH 56B.

Default parallelism: TP=8, PP=1, SP=True

Returns:

ConfigContainer with all settings pre-configured for NemotronH 56B SFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_4b_peft_config(

peft_scheme: str | megatron.bridge.peft.base.PEFT = 'lora',

) → megatron.bridge.training.config.ConfigContainer

Return a PEFT config for NemotronH 4B.

Default parallelism: TP=1, PP=1, SP=False

Parameters:

peft_scheme – PEFT scheme - “lora”, “dora”, or a custom PEFT instance.

Returns:

ConfigContainer with all settings pre-configured for NemotronH 4B PEFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_8b_peft_config(

peft_scheme: str | megatron.bridge.peft.base.PEFT = 'lora',

) → megatron.bridge.training.config.ConfigContainer

Return a PEFT config for NemotronH 8B.

Default parallelism: TP=1, PP=1, SP=False

Parameters:

peft_scheme – PEFT scheme - “lora”, “dora”, or a custom PEFT instance.

Returns:

ConfigContainer with all settings pre-configured for NemotronH 8B PEFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_47b_peft_config(

peft_scheme: str | megatron.bridge.peft.base.PEFT = 'lora',

) → megatron.bridge.training.config.ConfigContainer

Return a PEFT config for NemotronH 47B.

Default parallelism: TP=4, PP=1, SP=False

Parameters:

peft_scheme – PEFT scheme - “lora”, “dora”, or a custom PEFT instance.

Returns:

ConfigContainer with all settings pre-configured for NemotronH 47B PEFT.

bridge.recipes.nemotronh.nemotronh.nemotronh_56b_peft_config(

peft_scheme: str | megatron.bridge.peft.base.PEFT = 'lora',

) → megatron.bridge.training.config.ConfigContainer

Return a PEFT config for NemotronH 56B.

Default parallelism: TP=4, PP=1, SP=False

Parameters:

peft_scheme – PEFT scheme - “lora”, “dora”, or a custom PEFT instance.

Returns:

ConfigContainer with all settings pre-configured for NemotronH 56B PEFT.

bridge.recipes.nemotronh.nemotronh.__all__

[‘nemotronh_4b_pretrain_config’, ‘nemotronh_8b_pretrain_config’, ‘nemotronh_47b_pretrain_config’, ‘n…