End-to-End LLM Model Development with Torchtitan and Torchtune

[KeitaW]

Issue #, if available:

Description of changes:

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[KeitaW]

Basic functionalities have been implemented. Allow me to iterate on the other PRs...

[pbelevich]

Suggested change

	for details of each step, refer the [overview documentation](../../README.md).
	for details of each step, refer to the [overview documentation](../../README.md).

[pbelevich]

Suggested change

One of the primary challenges during such training is memory consumption. A typical model trained in mixed precision with AdamW requires 18 bytes per model parameter plus activation memory (6 bytes for parameters in mixed precision training, 8 bytes for AdamW, and 4 bytes for other overheads). For more details on the anatomy, see the [Hugging Face blog post](https://huggingface.co/docs/transformers/model_memory_anatomy) blog post. This means that training a 70B parameter model would require more than 1.12 TB of accelerated memory, which far exceeds the 80 GB capacity of H100 accelerated memory. To address this issue, torchtune integrates PyTorch Fully Sharded Data Parallel (FSDP).

One of the primary challenges during such training is memory consumption. A typical model trained in mixed precision with AdamW requires 18 bytes per model parameter(6 bytes for parameter in mixed precision training, 4 bytes for gradient and 8 bytes for AdamW optimizer states) plus activation memory. For more details on the anatomy, see the [Hugging Face blog post](https://huggingface.co/docs/transformers/model_memory_anatomy) blog post. This means that training a 70B parameter model would require more than 1.12 TiB of accelerator's memory, which far exceeds the 80 GB capacity of H100 memory. To address this issue, torchtune integrates PyTorch Fully Sharded Data Parallel (FSDP).

[pbelevich]

How was 1.12 TiB calculated?
70_000_000_000 * 18 = 1_260_000_000_000
1_260_000_000_000 / 1024 / 1024 / 1024 / 1024 = 1.15TiB

[pbelevich]

memory is not accelerated itself

[pbelevich]

Suggested change

**FSDP** is a distributed training feature designed to efficiently handle large model training by sharding model parameters, gradients, and optimizer states across multiple devices. This approach significantly reduces memory consumption and optimizes resource utilization, making it possible to train models that are too large to fit on a single GPU. In `torchtune` users can launch FSDP training job with command `tune run full_finetune_distributed`.

**FSDP** is a distributed training technique designed to efficiently handle large model training by sharding model parameters, gradients, and optimizer states across multiple devices. This approach significantly reduces memory consumption and optimizes resource utilization, making it possible to train models that are too large to fit on a single GPU. In `torchtune` users can launch FSDP training job with command `tune run full_finetune_distributed`.

[pbelevich]

--nnodes twice

[pbelevich]

--nnodes twice

[pbelevich]

I don't see where two instances are specified by default, I see only --nnodes 1 --nnodes=1 in sbatch files