Bump 3rdparty/NeMo from b685967 to 7d7a10c

[dependabot]

Bumps 3rdparty/NeMo from b685967 to 7d7a10c.

Commits

• 7d7a10c [fault tolerance] Add local checkpointing support (#12839)
• 37f2a69 Adds a built stage for the trt-llm wheel to reduce the overall test image siz...
• 95c9ede [NeMo2.0] Support MCore custom FSDP (#12547)
• 3da449b [automodel] Add FSDPv2-compatible context parallelism support. (#12821)
• bb9ff8b add accelerate to dependencies (#12871)
• d8ad2ed add apply_chat_template method to TokenizerSpec + AutoTokenizer (#12878)
• 87dc117 DeepseekV3 SFT finetuning perf config (#12829)
• 4e37ed2 Bump vllm 0.8.2 (#12753)
• f43a45f Support huggingface export to tensorrtllm (#12889)
• 57baba6 Enable in-fw deployment for eval with OAI compatible server (#12101)
• Additional commits viewable in compare view

You can trigger a rebase of this PR by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot merge will merge this PR after your CI passes on it
• @dependabot squash and merge will squash and merge this PR after your CI passes on it
• @dependabot cancel merge will cancel a previously requested merge and block automerging
• @dependabot reopen will reopen this PR if it is closed
• @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

Note
Automatic rebases have been disabled on this pull request as it has been open for over 30 days.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[codecov-commenter]

❌ 9 Tests Failed:

Tests completed	Failed	Passed	Skipped
980	9	971	22

View the top 3 failed test(s) by shortest run time

sub-packages/bionemo-evo2/tests/bionemo/evo2/test_hyena_operators.py::TestParallelShortHyenaOperator::test_gpu_forward

Stack Traces | 0.066s run time

self = <test_hyena_operators.TestParallelShortHyenaOperator object at 0x7f309f1b2810>
transformer_config = TransformerConfig(tensor_model_parallel_size=1, pipeline_model_parallel_comm_backend=None, pipeline_model_parallel_siz...g=True, disable_parameter_transpose_cache=False, config_logger_dir='', flash_decode=False, inference_rng_tracker=False)
hyena_config = HyenaConfig(tie_projection_weights=False, to_upper='normalized_weighted', lowercase_loss_reweighting=0.1, short_conv_L...onv_mixer=False, hyena_short_conv_pregate=True, hyena_short_conv_postgate=True, proj_groups=1, grouped_attention=False)

    @pytest.fixture
    def operator(self, transformer_config: TransformerConfig, hyena_config: HyenaConfig) -> ParallelShortHyenaOperator:
        with megatron_parallel_state_utils.distributed_model_parallel_state():
>           yield ParallelShortHyenaOperator(
                hidden_size=transformer_config.hidden_size,
                transformer_config=transformer_config,
                hyena_config=hyena_config,
                init_method="small_init",
                short_conv_class=ParallelCausalDepthwiseConv1d,
                use_fast_causal_conv=False,
                is_mlp=False,
                local_init=False,
                use_conv_bias=False,
            )
E           TypeError: ParallelShortHyenaOperator.__init__() got an unexpected keyword argument 'is_mlp'

.../bionemo/evo2/test_hyena_operators.py:85: TypeError

sub-packages/bionemo-evo2/tests/bionemo/evo2/test_hyena_operators.py::TestParallelShortHyenaOperator::test_initialization

Stack Traces | 0.066s run time

self = <test_hyena_operators.TestParallelShortHyenaOperator object at 0x7f309f1b1b50>
transformer_config = TransformerConfig(tensor_model_parallel_size=1, pipeline_model_parallel_comm_backend=None, pipeline_model_parallel_siz...g=True, disable_parameter_transpose_cache=False, config_logger_dir='', flash_decode=False, inference_rng_tracker=False)
hyena_config = HyenaConfig(tie_projection_weights=False, to_upper='normalized_weighted', lowercase_loss_reweighting=0.1, short_conv_L...onv_mixer=False, hyena_short_conv_pregate=True, hyena_short_conv_postgate=True, proj_groups=1, grouped_attention=False)

    @pytest.fixture
    def operator(self, transformer_config: TransformerConfig, hyena_config: HyenaConfig) -> ParallelShortHyenaOperator:
        with megatron_parallel_state_utils.distributed_model_parallel_state():
>           yield ParallelShortHyenaOperator(
                hidden_size=transformer_config.hidden_size,
                transformer_config=transformer_config,
                hyena_config=hyena_config,
                init_method="small_init",
                short_conv_class=ParallelCausalDepthwiseConv1d,
                use_fast_causal_conv=False,
                is_mlp=False,
                local_init=False,
                use_conv_bias=False,
            )
E           TypeError: ParallelShortHyenaOperator.__init__() got an unexpected keyword argument 'is_mlp'

.../bionemo/evo2/test_hyena_operators.py:85: TypeError

sub-packages/bionemo-evo2/tests/bionemo/evo2/test_hyena_operators.py::TestParallelHyenaOperator::test_gpu_forward

Stack Traces | 0.067s run time

self = <test_hyena_operators.TestParallelHyenaOperator object at 0x7f309f1b1be0>
operator = ParallelHyenaOperator(
  (filter): ExplicitSingleDecayFilter()
)

    def test_gpu_forward(self, operator: ParallelHyenaOperator):
        device = torch.device("cuda")
        operator = operator.to(device)
        batch_size = 2
        seq_len = operator.L  # operator.L maps to max_sequence_length
        g = operator.num_groups
        dg = operator.group_dim
    
        x1 = torch.ones((batch_size, seq_len, g, dg), device=device)
        x2 = torch.ones((batch_size, seq_len, g, dg), device=device)
        v = torch.ones((batch_size, seq_len, g, dg), device=device)
    
>       output = operator(x1, x2, v)

.../bionemo/evo2/test_hyena_operators.py:75: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
.../local/lib/python3.12.../nn/modules/module.py:1740: in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
.../local/lib/python3.12.../nn/modules/module.py:1751: in _call_impl
    return forward_call(*args, **kwargs)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = ParallelHyenaOperator(
  (filter): ExplicitSingleDecayFilter()
)
x1 = tensor([[[[1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]],

     ...1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]]]], device='cuda:0')
x2 = tensor([[[[1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]],

     ...1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]]]], device='cuda:0')
v = tensor([[[[1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]],

     ...1.],
          [1.],
          [1.],
          ...,
          [1.],
          [1.],
          [1.]]]], device='cuda:0')
_hyena_use_cp = True

    def forward(self, x1, x2, v, _hyena_use_cp=True):
        """Shape specification for inputs and outputs.
    
        Input shapes: bs, (num_groups, group_size), seq_length
        Output shapes: bs, (num_groups, group_size), seq_length
        """
>       B, GDG, L = x1.shape
E       ValueError: too many values to unpack (expected 3)

.../local/lib/python3.12.../megatron/hyena/hyena_utils.py:808: ValueError

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[dependabot]

A newer version of 3rdparty/NeMo exists, but since this PR has been edited by someone other than Dependabot I haven't updated it. You'll get a PR for the updated version as normal once this PR is merged.

[dependabot]

OK, I won't notify you again about this release, but will get in touch when a new version is available. You can also ignore all major, minor, or patch releases for a dependency by adding an ignore condition with the desired update_types to your config file.

If you change your mind, just re-open this PR and I'll resolve any conflicts on it.