Skip to content

exporters - #4672

Open
micwill755 wants to merge 3 commits into
mainfrom
exporters
Open

exporters#4672
micwill755 wants to merge 3 commits into
mainfrom
exporters

Conversation

@micwill755

Copy link
Copy Markdown
Collaborator

Description

Please include a summary of the change and which issue is fixed. Please also include relevant motivation and context. List any dependencies that are required for this change.

Fixes # (issue)

Type of change

Please delete options that are not relevant and/or add your own.

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)
  • This change requires a documentation update

Checklist:

  • My code follows the style guidelines of this project (You can use the linters)
  • I have performed a self-review of my own code
  • I have commented my code, particularly in hard-to-understand areas and hacks
  • I have made corresponding changes to the documentation
  • I have added tests to verify my fix or my feature
  • New and existing unit tests pass locally with my changes
  • I have added the relevant labels to my PR in so that relevant reviewers are notified

Stack created with GitHub Stacks CLIGive Feedback 💬

Drop the Test/trt import path so PI05, GR00T, and Nemotron export is self-contained, and rename GR00T action_context to context_projection.
Each family now has spec, helpers, and patches in one folder, with shared vision/language/action-step code in models/common.
@meta-cla meta-cla Bot added the cla signed label Sep 3, 2026
@github-actions github-actions Bot added component: tests Issues re: Tests component: build system Issues re: Build system component: api [Python] Issues re: Python API labels Sep 3, 2026
@github-actions
github-actions Bot requested a review from narendasan September 3, 2026 21:35

@narendasan narendasan left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is cool, for sure much more clean. Do you have like a guide on how to bring up your own model?

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You can just put the example of the exporter tool with the tool unless you want them rendered. Then we need to use the sphinx gallery format

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you have the inference side of these examples? (Ideally both python and C++ deployments)

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we start this out of tree for now like //tools/hf_exporters?

num_kv = int(cfg.num_key_value_heads)
head_dim = int(getattr(cfg, "head_dim", cfg.hidden_size // cfg.num_attention_heads))
try:
from torch_tensorrt.hf.exporters.rope import make_rope_rotary_cos_sin

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do we need to guard around code we ship?



@dynamo_tensorrt_converter(
torch.ops.trt.attention_plugin.default,

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

make this namespace torch.ops.tensorrt_edge_llm.*



@dynamo_tensorrt_converter(
torch.ops.trt.vit_attention_plugin.default,

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here



@dynamo_tensorrt_converter(
torch.ops.trt.causal_conv1d.default,

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Here as well



@dynamo_tensorrt_converter(
torch.ops.trt.update_ssm_state.default,

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

""



@dynamo_tensorrt_converter(
torch.ops.trt.nvfp4_moe_plugin.default,

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

""


import torch

_THOR_CUDA_LIB = Path("/usr/local/cuda-13.0/thor/targets/aarch64-linux/lib")

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should have some more robust discovery method

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cla signed component: api [Python] Issues re: Python API component: build system Issues re: Build system component: tests Issues re: Tests

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants