Skip to content

Releases: tenstorrent/tt-metal

v0.56.0-rc45

03 Mar 02:28
@github-actions github-actions
v0.56.0-rc45
4bc3ae3
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc45 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13621564810

📦 Uncategorized

  • #0: Add programming examples for TT-Metalium multi-device native APIs
Loading

v0.56.0-rc44

01 Mar 05:31
@github-actions github-actions
v0.56.0-rc44
ce17124
Compare
Choose a tag to compare
v0.56.0-rc44 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13599594043

📦 Uncategorized

  • #0: Add programming examples for TT-Metalium multi-device native APIs
Loading

v0.56.0-rc43

28 Feb 05:40
@github-actions github-actions
v0.56.0-rc43
9b3d855
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc43 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13579442965

📦 Uncategorized

  • #0: Add programming examples for TT-Metalium multi-device native APIs
Loading

v0.56.0-rc42

27 Feb 06:26
@github-actions github-actions
v0.56.0-rc42
ed7d045
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc42 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13556960790

📦 Uncategorized

  • #0: Add programming examples for TT-Metalium multi-device native APIs
Loading

v0.56.0-rc41

26 Feb 03:56
@github-actions github-actions
v0.56.0-rc41
eb34ec8
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc41 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13534601244

📦 Uncategorized

  • #0: Add programming examples for TT-Metalium multi-device native APIs
Loading

v0.56.0-rc38

22 Feb 06:40
@github-actions github-actions
v0.56.0-rc38
4eb7c33
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc38 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13468231055

📦 Uncategorized

  • #17374: Add concurrency group for _produce-data.yaml
Loading

v0.56.0-rc37

21 Feb 01:46
@github-actions github-actions
v0.56.0-rc37
53f3d05
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc37 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13448174485

📦 Uncategorized

  • Implement JointAttention
  • #17374: Add concurrency group for _produce-data.yaml
  • Extra cleanup. With emphasis
  • [skip ci] Revert "disable-pr-gate"
Loading

v0.56.0-rc36

20 Feb 01:41
@github-actions github-actions
v0.56.0-rc36
e68353d
Compare
Choose a tag to compare
v0.56.0-rc36 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13425954330

📦 Uncategorized

  • Implement JointAttention
  • #17374: Add concurrency group for _produce-data.yaml
  • Extra cleanup. With emphasis
  • Add .test_durations for ttnn pytest splits
  • #13856: update gelu_bw op documentation on bfloat8_b limitations
  • #17134: Enable test_transformer_2d_model UT in SD
  • Add test for create_qkv_heads_decode on subcoregrids
  • Always flatten convolution input tensors
  • #0: Use kernel time instead of fw time for calculating device time
  • Fix golden function for ttnn.linear so it can be used with comparison mode
  • [skip ci] update codeowners with matmul+ team changes
  • #0: Fix NCRISC delay
  • Change create_tt_tensor_from_py_data to use from_vector - fixed branch
  • Revert "Change create_tt_tensor_from_py_data to use from_vector - fixed branch"
  • #0: Add missing flush in send_chunk_from_address
  • Pick up latest actions setup venv
  • #16364: split prefetch dependent config
  • #16339: let DispatchMemMap get created with custom DispatchSettings
  • #0: update CODEOWNERS based on eltwise team changes
  • [skip ci] add links and standardize pull request checklist text
  • [skip ci] Revert "disable-pr-gate"
  • #0: Simplify MeshDevice construction by eradicating mesh_type
  • #17541: Guard against passing 0 to counting zero intrinsics in program.cpp
  • Revamp install_dependencies.sh
  • Jjiang/fixed tensor creation and conversion
  • Create Tutorial - Adding a Model
  • [skip ci] Update README.md
  • #17201: Implement IDevice tracing APIs for MeshDevice
  • [skip ci] Update README.md
  • Optimize EDM-fabric flow-control protocols
Loading

v0.56.0-rc35

19 Feb 01:41
@github-actions github-actions
v0.56.0-rc35
686a4f0
This commit was created on github.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.56.0-rc35 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13403584877

📦 Uncategorized

  • Implement JointAttention
  • #17374: Add concurrency group for _produce-data.yaml
  • Extra cleanup. With emphasis
  • #17324: port Padding CPP Unit OP Tests to TTNN
  • Add dprint/watcher support for BH second erisc
  • Fix outdated call to ttnn.open_device in ttnn tutorial notebooks
  • Use three arguments for distro, version, architecture
  • update deprecated API use
  • Fix pgm dispatch sweep to use new args
  • Fix nanogpt failure in tt-train after shape changes
  • Fix init calls for binary sfpu pow and quant
  • Aliu/quanta bringup
  • Split ttnn pytests into more groups
  • Remove MeshCQ's dependence on HardwareCommandQueue for loading SubDevices
  • Fix two broken workflows by updating arguments
  • Aliu/quanta bringup
  • #17203: Remove set_worker_queue_mode method from IDevice
  • [skip ci] Update metal-api-surface.yaml
  • Fix workflow call
  • [skip-ci] Missed one
  • #0: Fix T3000 unit tests and suppress chatty logs
  • Use transaction id for eth receiver write to worker
  • [skip ci] Fix name bug
  • Update data_formats.md with mantissa rounding description
  • #0: Use stream autoinc register to send word credits over ethernet.
  • #17134: Fix SD resnet block ut parametrizations
  • Add .test_durations for ttnn pytest splits
  • #13856: update gelu_bw op documentation on bfloat8_b limitations
  • #17134: Enable test_transformer_2d_model UT in SD
  • Add test for create_qkv_heads_decode on subcoregrids
  • Always flatten convolution input tensors
  • #0: Use kernel time instead of fw time for calculating device time
  • Fix golden function for ttnn.linear so it can be used with comparison mode
  • [skip ci] update codeowners with matmul+ team changes
  • #0: Fix NCRISC delay
  • Change create_tt_tensor_from_py_data to use from_vector - fixed branch
  • Revert "Change create_tt_tensor_from_py_data to use from_vector - fixed branch"
  • #0: Add missing flush in send_chunk_from_address
  • Pick up latest actions setup venv
  • #16364: split prefetch dependent config
  • #16339: let DispatchMemMap get created with custom DispatchSettings
  • #0: update CODEOWNERS based on eltwise team changes
  • [skip ci] add links and standardize pull request checklist text
  • [skip ci] Revert "disable-pr-gate"
  • #0: Simplify MeshDevice construction by eradicating mesh_type
  • #17541: Guard against passing 0 to counting zero intrinsics in program.cpp
  • Revamp install_dependencies.sh
  • Jjiang/fixed tensor creation and conversion
  • Create Tutorial - Adding a Model
  • [skip ci] Update README.md
  • #17201: Implement IDevice tracing APIs for MeshDevice
  • [skip ci] Update README.md
  • Optimize EDM-fabric flow-control protocols
Loading

v0.56.0-rc34

18 Feb 22:52
@github-actions github-actions
v0.56.0-rc34
a477396
Compare
Choose a tag to compare
v0.56.0-rc34 Pre-release
Pre-release

Note

If you are installing from a release, please refer to the README, INSTALLATION instructions, and any other documentation packaged with the release, not on the main branch. There may be differences between the latest main and the previous release.

The changelog will now follow, showing the changes from last release.

This release was generated by the CI workflow https://github.com/tenstorrent/tt-metal/actions/runs/13401445732

📦 Uncategorized

  • Implement JointAttention
  • #17374: Add concurrency group for _produce-data.yaml
  • Extra cleanup. With emphasis
  • #17324: port Padding CPP Unit OP Tests to TTNN
  • #17013: Update add_bw doc
  • #0: Fix produce data crash when testcase is missing properties block
  • #0: Update Llama3 PERF.md and llama3 vision PCC after cache regeneration
  • #17483: Fix untilize with unpad for 1d tensors
  • Add dprint/watcher support for BH second erisc
  • Fix outdated call to ttnn.open_device in ttnn tutorial notebooks
  • Use three arguments for distro, version, architecture
  • update deprecated API use
  • Fix pgm dispatch sweep to use new args
  • Fix nanogpt failure in tt-train after shape changes
  • Fix init calls for binary sfpu pow and quant
  • Aliu/quanta bringup
  • Split ttnn pytests into more groups
  • Remove MeshCQ's dependence on HardwareCommandQueue for loading SubDevices
  • Fix two broken workflows by updating arguments
  • Aliu/quanta bringup
  • #17203: Remove set_worker_queue_mode method from IDevice
  • [skip ci] Update metal-api-surface.yaml
  • Fix workflow call
  • [skip-ci] Missed one
  • #0: Fix T3000 unit tests and suppress chatty logs
  • Use transaction id for eth receiver write to worker
  • [skip ci] Fix name bug
  • Update data_formats.md with mantissa rounding description
  • #0: Use stream autoinc register to send word credits over ethernet.
  • #17134: Fix SD resnet block ut parametrizations
  • Add .test_durations for ttnn pytest splits
  • #13856: update gelu_bw op documentation on bfloat8_b limitations
  • #17134: Enable test_transformer_2d_model UT in SD
  • Add test for create_qkv_heads_decode on subcoregrids
  • Always flatten convolution input tensors
  • #0: Use kernel time instead of fw time for calculating device time
  • Fix golden function for ttnn.linear so it can be used with comparison mode
  • [skip ci] update codeowners with matmul+ team changes
  • #0: Fix NCRISC delay
  • Change create_tt_tensor_from_py_data to use from_vector - fixed branch
  • Revert "Change create_tt_tensor_from_py_data to use from_vector - fixed branch"
  • #0: Add missing flush in send_chunk_from_address
  • Pick up latest actions setup venv
  • #16364: split prefetch dependent config
  • #16339: let DispatchMemMap get created with custom DispatchSettings
  • #0: update CODEOWNERS based on eltwise team changes
  • [skip ci] add links and standardize pull request checklist text
  • [skip ci] Revert "disable-pr-gate"
  • #0: Simplify MeshDevice construction by eradicating mesh_type
  • #17541: Guard against passing 0 to counting zero intrinsics in program.cpp
  • Revamp install_dependencies.sh
  • Jjiang/fixed tensor creation and conversion
  • Create Tutorial - Adding a Model
  • [skip ci] Update README.md
  • #17201: Implement IDevice tracing APIs for MeshDevice
  • [skip ci] Update README.md
  • Optimize EDM-fabric flow-control protocols
Loading