🦙🦙 Llama3.2 Release: Support for 1B, 3B, and 11B (Multimodal)

[Jack-Khuu]

This PR introduces support for the new Llama3.2 models announced at Connect 2024.

Llama3.2 1B and 3B are text-only models that now have access to all the same torchchat features as their predecesessors (Llama3.1 8B)

Llama3.2 11B is a Multimodal Model Capable of supporting both Image and Text inputs. For more information about what is currently supported, take a look at https://github.com/pytorch/torchchat/blob/llama32/docs/multimodal.md

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchchat/1204

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 22 New Failures, 3 Cancelled Jobs, 3 Unrelated Failures

As of commit 81b2ff2 with merge base 0b8ca05 ():

NEW FAILURES - The following jobs have failed:

• pull / compile-gguf (macos-14) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / runner-aoti (macos-14-xlarge) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / runner-et (macos-14-xlarge) (gh)
  RuntimeError: module was compiled against NumPy C-API version 0x10 (NumPy 1.23) but the running NumPy has C-API version 0xe. Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem.
• pull / test-cpu-aoti (aarch64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-aoti (x86_64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-compile (aarch64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-compile (x86_64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check (aarch64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check (x86_64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check-float16 (aarch64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check-float16 (x86_64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check-float32 (aarch64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-cpu-eval-sanity-check-float32 (x86_64, stories15M) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• pull / test-gpu-aoti-bfloat16 (cuda, stories15M) / linux-job (gh)
  RuntimeError: Command docker exec -t b170239fb655aeb7c0516aa3fb191b7cccf6d2bc7a9f870a582e82a4a1691548 /exec failed with exit code 1
• pull / test-gpu-aoti-float16 (cuda, stories15M) / linux-job (gh)
  RuntimeError: Command docker exec -t 9a8a5a129d6e807e3c99e03bb97de9c1c0617cdbd193f9710830465c0631878e /exec failed with exit code 1
• pull / test-gpu-aoti-float32 (cuda, stories15M) / linux-job (gh)
  RuntimeError: Command docker exec -t 1b5e0581a151c52731ab80fab8efd3350c6fabe06f28d3598c7c2e6678cc87df /exec failed with exit code 1
• pull / test-gpu-compile (cuda, stories15M) / linux-job (gh)
  RuntimeError: Command docker exec -t 354898eadafb2bec27948df56033af96e1fa2a05607939114d091281f5a0223b /exec failed with exit code 1
• pull / test-gpu-eval-sanity-check (cuda, stories15M) / linux-job (gh)
  RuntimeError: Command docker exec -t f711f91bca92f88ee432e47837a49690ad4110e54853c24ac6b844e4c47d9d61 /exec failed with exit code 1
• pull / test-tinystories-executorch (macos-14-xlarge) (gh)
  RuntimeError: module was compiled against NumPy C-API version 0x10 (NumPy 1.23) but the running NumPy has C-API version 0xe. Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem.
• pull / torchchat-command-load-test (macos-14) (gh)
  ModuleNotFoundError: No module named 'torchtune.models.llama3_2_vision'
• Run parallel prefill / test-cuda / linux-job (gh)
  RuntimeError: Command docker exec -t eb6e0c3259be8c9f5c257e3557c817557e1a7e51d7339d7172c0f391ce12bdb2 /exec failed with exit code 1
• Run the aoti runner with CUDA using stories / test-runner-aot-cuda / linux-job (gh)
  RuntimeError: Command docker exec -t 840e783545c8528312f97e200a6871af60cfa600203211e2469a0e2cd99a78c3 /exec failed with exit code 1

CANCELLED JOBS - The following jobs were cancelled. Please retry:

• pull / runner-aoti (16-core-ubuntu) (gh)
  ##[error]The operation was canceled.
• pull / runner-et (16-core-ubuntu) (gh)
  ##[error]The operation was canceled.
• pull / test-tinystories-executorch (16-core-ubuntu) (gh)
  ##[error]The operation was canceled.

FLAKY - The following jobs failed but were likely due to flakiness present on trunk:

• pull / test-mps / macos-job (gh) (detected as infra flaky with no log or failing log classifier)
• pull / test-mps-dtype / macos-job (gh) (detected as infra flaky with no log or failing log classifier)
• Run the README instructions - with stories - on MPS/MacOS / test-advanced-mps-macos / macos-job (gh) (detected as infra flaky with no log or failing log classifier)

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[iseeyuan]

Nice!