Optimum-Habana docs re-org #1488
Conversation
Signed-off-by: Daniel Socek <[email protected]> Co-authored-by: Greg Serochi <[email protected]> Co-authored-by: Kiangpeng Lau <[email protected]> Co-authored-by: Seethong Vang <[email protected]>
| @@ -32,20 +34,8 @@ | |||
| title: How to use DeepSpeed | |||
| - local: usage_guides/multi_node_training | |||
| title: Multi-node Training | |||
| - local: usage_guides/quantization | |||
| title: Quantization | |||
There was a problem hiding this comment.
why do we change this way?
| <a class="!no-underline border dark:border-gray-700 p-5 rounded-lg shadow hover:shadow-lg" href="./package_reference/trainer" | ||
| ><div class="w-full text-center bg-gradient-to-br from-purple-400 to-purple-500 rounded-lg py-1.5 font-semibold mb-5 text-white text-lg leading-relaxed">Reference</div> | ||
| <p class="text-gray-700">Technical descriptions of how the Habana classes and methods of 🤗 Optimum Habana work.</p> | ||
| </a> |
There was a problem hiding this comment.
@libinta , please note this new, simplified organization was suggested by Greg. As for why, I believe the intension was to simplify and reduce number of sections
| # DistributedRunner | ||
|
|
||
| [[autodoc]] distributed.distributed_runner.DistributedRunner | ||
| - all |
There was a problem hiding this comment.
See Greg's vision plan below
| The `GaudiDDPMPipeline` is to enable unconditional image generations on HPUs. It has similar APIs as the regular `DiffusionPipeline`. | ||
| It shares a common parent class, `GaudiDiffusionPipeline`, with other existing Gaudi pipelines. It now supports both DDPM and DDIM scheduler. | ||
| It is recommended to use the optimized scheduler, `GaudiDDIMScheduler`, to obtain the best performance and image outputs. | ||
|
|
There was a problem hiding this comment.
Greg suggested to upgrade to SDXL as SD is kind of old model now. We can keep too, it may be a good intro diffuser model.
There was a problem hiding this comment.
@kplau1128 can you comment here?
There was a problem hiding this comment.
Greq was requested and suggested consolidate SD into one tutorials/stable_diffusion.mdx, and remove others
| ## GaudiSeq2SeqTrainingArguments | ||
|
|
||
| [[autodoc]] transformers.training_args_seq2seq.GaudiSeq2SeqTrainingArguments | ||
| - all |
There was a problem hiding this comment.
Greg's vision of new doc org:
| ``` | ||
|
|
||
| where `gaudi_config_name` is the name of a model from the [Hub](https://huggingface.co/Habana) (Gaudi configurations are stored in model repositories) or a path to a local Gaudi configuration file (you can see [here](./package_reference/gaudi_config) how to write your own). | ||
| where `gaudi_config_name` is the name of a model from the [Hub](https://huggingface.co/Habana) (Gaudi configurations are stored in model repositories) |
There was a problem hiding this comment.
why do we talk about gaudi_config_name here? it seems we are talking about docker before.
There was a problem hiding this comment.
This out of place. Good catch. Probably leftover from previous content.
There was a problem hiding this comment.
Good catch. This is out of place, probably leftover from previous content
Hugging Face Gaudi Documentation Update
Comprehensive update to HF Gaudi docs