diff --git a/docs/source/bring_your_own_policies.mdx b/docs/source/bring_your_own_policies.mdx index c3cc040e3..bf71efb7e 100644 --- a/docs/source/bring_your_own_policies.mdx +++ b/docs/source/bring_your_own_policies.mdx @@ -150,14 +150,14 @@ class MyPolicy(PreTrainedPolicy): The methods called by the train/eval loops: -| Method | Used by | What it does | -| ----------------------------------------------------------------- | ----------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -| `reset() -> None` | `lerobot-eval` | Clear per-episode state at the start of each episode. | -| `select_action(batch, **kwargs) -> Tensor` | `lerobot-eval` | Return the next action `(B, action_dim)`. Called every step. | -| `predict_action_chunk(batch, **kwargs) -> Tensor` | the policy itself | Return an action chunk `(B, chunk_size, action_dim)`. Currently abstract on the base class — raise `NotImplementedError` if your policy doesn't chunk. | -| `forward(batch, reduction="mean") -> tuple[Tensor, dict \| None]` | `lerobot-train` | Return `(loss, output_dict)`. Accept `reduction="none"` if you want to support per-sample weighting. | -| `get_optim_params() -> dict` | the optimizer | Return `self.parameters()` for simple policies; return a named parameter dict for [multi-optimizer policies](https://github.com/huggingface/lerobot/blob/ecd38c50d7d15b4184cf42649ff1185ee2e11eeb/src/lerobot/policies/sac/modeling_sac.py#L61-L73). | -| `update() -> None` _(optional)_ | `lerobot-train` | Called after each optimizer step _if defined_. Use for EMA, target nets, replay buffers (TDMPC uses this). | +| Method | Used by | What it does | +| ----------------------------------------------------------------- | ----------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `reset() -> None` | `lerobot-eval` | Clear per-episode state at the start of each episode. | +| `select_action(batch, **kwargs) -> Tensor` | `lerobot-eval` | Return the next action `(B, action_dim)`. Called every step. | +| `predict_action_chunk(batch, **kwargs) -> Tensor` | the policy itself | Return an action chunk `(B, chunk_size, action_dim)`. Currently abstract on the base class — raise `NotImplementedError` if your policy doesn't chunk. | +| `forward(batch, reduction="mean") -> tuple[Tensor, dict \| None]` | `lerobot-train` | Return `(loss, output_dict)`. Accept `reduction="none"` if you want to support per-sample weighting. | +| `get_optim_params() -> dict` | the optimizer | Return `self.parameters()` for simple policies; return a named parameter dict for multi-optimizer policies (see `get_optim_params` in [`modeling_act.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/act/modeling_act.py) for a per-group learning-rate example). | +| `update() -> None` _(optional)_ | `lerobot-train` | Called after each optimizer step _if defined_. Use for EMA, target nets, replay buffers (TDMPC uses this). | Batches are flat dictionaries keyed by the constants in [`lerobot.utils.constants`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/utils/constants.py): `OBS_STATE` (`observation.state.`), `OBS_IMAGES` (`observation.images.`), `OBS_LANGUAGE`, `ACTION`, etc. Reuse the constants — don't invent new prefixes. @@ -295,12 +295,10 @@ The file names are load-bearing: the factory does lazy imports by name, and the ### Wiring -Four places need to know about your policy. All by name. +Two places need to know about your policy. All by name. -1. **`policies/__init__.py`** — re-export `MyPolicyConfig` and add it to `__all__`. **Don't** re-export the modeling class; it loads lazily through the factory (so `import lerobot` stays fast). -2. **`factory.py:get_policy_class`** — add a branch returning `MyPolicy` from a lazy import. -3. **`factory.py:make_policy_config`** and **`factory.py:make_pre_post_processors`** — same idea, two more branches. -4. **`templates/lerobot_modelcard_template.md` and the root `README.md`** — the template is what `push_model_to_hub` renders into the model card of every checkpoint trained with your policy: add a one-line description of your policy in the `model_name` branches, map it in `policy_docs` so cards link to your MDX guide, and optionally add an architecture image to `diagrams`. Then add your policy to the models table in the root `README.md`, under the right category, linking to your doc page. +1. **`policies/__init__.py`** — re-export `MyPolicyConfig` and add it to `__all__`. This import is what registers your policy: `@PreTrainedConfig.register_subclass("my_policy")` runs, and from then on the factory resolves everything by convention. **Don't** re-export the modeling class; it loads lazily through the factory (so `import lerobot` stays fast). +2. **`templates/lerobot_modelcard_template.md` and the root `README.md`** — the template is what `push_model_to_hub` renders into the model card of every checkpoint trained with your policy: add a one-line description of your policy in the `model_name` branches, map it in `policy_docs` so cards link to your MDX guide, and optionally add an architecture image to `diagrams`. Then add your policy to the models table in the root `README.md`, under the right category, linking to your doc page. Mirror an existing policy that's structurally similar to yours; the diff is small. @@ -332,6 +330,10 @@ This way: Add a matching extra to [`pyproject.toml`](https://github.com/huggingface/lerobot/blob/main/pyproject.toml) `[project.optional-dependencies]` and include it in the `all` extra so `pip install 'lerobot[all]'` keeps installing everything. +### Avoid copying a modeling file — subclass it + +If your policy needs to modify a backbone that already exists in `transformers` (custom conditioning, extra inputs, a swapped sub-module), **do not vendor a copy of its `modeling_*.py`**. Instead, subclass the smallest upstream unit and override only what changes. [`pi_gemma.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/pi_gemma.py) is the canonical reference: it injects AdaRMS conditioning into PaliGemma/Gemma in ~370 lines by subclassing `GemmaModel`/`PaliGemmaModel` and overriding the decoder-layer forward, instead of forking the ~2,000-line modeling file. Model surgery on a _loaded_ native model is also fine (layer truncation, tokenizer expansion, hidden-state capture — see `evo1/internvl3_embedder.py`, `eo1/modeling_eo1.py`, `groot/groot_n1_7.py` for working examples). Reviewers will ask for this pattern when a PR arrives with a copied modeling file; the only accepted exception is a model that does not exist in `transformers` at all. + ### Benchmarks and a published checkpoint A new policy is much easier to review — and far more useful — when it ships with a working checkpoint and at least one number you can reproduce. @@ -367,7 +369,7 @@ If your policy is real-robot-only and no sim benchmark applies, swap the sim eva The general expectations are in [`CONTRIBUTING.md`](https://github.com/huggingface/lerobot/blob/main/CONTRIBUTING.md) and the [PR template](https://github.com/huggingface/lerobot/blob/main/.github/PULL_REQUEST_TEMPLATE.md). On top of those, reviewers will look for: - [ ] `MyPolicy` and `MyPolicyConfig` cover the surface above; `__init_subclass__` accepts the class. -- [ ] `factory.py` and `policies/__init__.py` are wired (lazy imports for modeling). +- [ ] `policies/__init__.py` re-exports the config (this registers the policy; the factory resolves modeling/processor by naming convention). - [ ] `make_my_policy_pre_post_processors` follows the naming convention. - [ ] Optional deps live behind a `[project.optional-dependencies]` extra and the `TYPE_CHECKING + require_package` guard. - [ ] `tests/policies/` updated; backward-compat artifact committed & policy-specific tests. diff --git a/examples/annotations/run_hf_job.py b/examples/annotations/run_hf_job.py index a77e22f14..506a45a12 100644 --- a/examples/annotations/run_hf_job.py +++ b/examples/annotations/run_hf_job.py @@ -46,8 +46,11 @@ CMD = ( "apt-get update -qq && apt-get install -y -qq git ffmpeg && " "pip install --no-deps " "'lerobot @ git+https://github.com/huggingface/lerobot.git@main' && " + # Pins mirror pyproject.toml — unpinned installs pull av 18 / datasets 5 / + # draccus 0.11, which break lerobot at import time. "pip install --upgrade-strategy only-if-needed " - "datasets pyarrow av jsonlines draccus gymnasium torchcodec mergedeep pyyaml-include toml typing-inspect " + "'datasets>=4.7.0,<5.0.0' 'pyarrow>=21.0.0,<30.0.0' 'av>=15.0.0,<16.0.0' 'draccus==0.10.0' " + "'pandas>=2.0.0,<3.0.0' jsonlines gymnasium torchcodec mergedeep pyyaml-include toml typing-inspect " "openai && " "export VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0 && " "export VLLM_VIDEO_BACKEND=pyav && " diff --git a/pyproject.toml b/pyproject.toml index de028e41a..90c1653f7 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -417,8 +417,6 @@ ignore = [ "__init__.py" = ["F401", "F403", "E402"] # E402: conditional-import guards (TYPE_CHECKING / is_package_available) must precede the imports they protect "src/lerobot/scripts/convert_dataset_v21_to_v30.py" = ["E402"] -"src/lerobot/policies/wall_x/**" = ["N801", "N812", "SIM102", "SIM108", "SIM210", "SIM211", "B006", "B007", "SIM118"] # Supprese these as they are coming from original Qwen2_5_vl code TODO(pepijn): refactor original - [tool.ruff.lint.isort] combine-as-imports = true known-first-party = ["lerobot"] diff --git a/src/lerobot/configs/policies.py b/src/lerobot/configs/policies.py index b0f003519..4da0fb9e8 100644 --- a/src/lerobot/configs/policies.py +++ b/src/lerobot/configs/policies.py @@ -205,24 +205,30 @@ class PreTrainedConfig(draccus.ChoiceRegistry, HubMixin, abc.ABC): # type: igno f"{CONFIG_NAME} not found on the HuggingFace Hub in {model_id}" ) from e - # HACK: Parse the original config to get the config subclass, so that we can - # apply cli overrides. - # This is very ugly, ideally we'd like to be able to do that natively with draccus - # something like --policy.path (in addition to --policy.type) - with draccus.config_type("json"): - orig_config = draccus.parse(cls, config_file, args=[]) - if config_file is None: raise FileNotFoundError(f"{CONFIG_NAME} not found in {model_id}") with open(config_file) as f: config = json.load(f) - config.pop("type") + # Resolve the concrete config subclass from the serialized "type" tag, then parse + # the config (with CLI overrides) directly for that class. The "type" key is + # stripped because draccus only consumes it when parsing the registry base class. + policy_type = config.pop("type", None) + if policy_type is None: + raise ValueError(f"Missing 'type' field in {CONFIG_NAME} of {model_id}") + try: + config_cls = cls.get_choice_class(policy_type) + except Exception as e: + raise ValueError( + f"Policy type '{policy_type}' (from {CONFIG_NAME} of {model_id}) is not registered. " + f"Available policy types: {cls.get_known_choices()}" + ) from e + with tempfile.NamedTemporaryFile("w+", delete=False, suffix=".json") as f: json.dump(config, f) config_file = f.name cli_overrides = policy_kwargs.pop("cli_overrides", []) with draccus.config_type("json"): - return draccus.parse(orig_config.__class__, config_file, args=cli_overrides) + return draccus.parse(config_cls, config_file, args=cli_overrides) diff --git a/src/lerobot/policies/__init__.py b/src/lerobot/policies/__init__.py index 7f0bed2e0..a95d23b91 100644 --- a/src/lerobot/policies/__init__.py +++ b/src/lerobot/policies/__init__.py @@ -32,6 +32,7 @@ from .pretrained import PreTrainedPolicy as PreTrainedPolicy from .smolvla.configuration_smolvla import SmolVLAConfig as SmolVLAConfig from .tdmpc.configuration_tdmpc import TDMPCConfig as TDMPCConfig from .utils import make_robot_action, prepare_observation_for_inference +from .vla_jepa.configuration_vla_jepa import VLAJEPAConfig as VLAJEPAConfig from .vqbet.configuration_vqbet import VQBeTConfig as VQBeTConfig from .wall_x.configuration_wall_x import WallXConfig as WallXConfig from .xvla.configuration_xvla import XVLAConfig as XVLAConfig @@ -57,6 +58,7 @@ __all__ = [ "PI05Config", "SmolVLAConfig", "TDMPCConfig", + "VLAJEPAConfig", "VQBeTConfig", "WallXConfig", "XVLAConfig", diff --git a/src/lerobot/policies/act/processor_act.py b/src/lerobot/policies/act/processor_act.py index d87ade900..d3109aeb3 100644 --- a/src/lerobot/policies/act/processor_act.py +++ b/src/lerobot/policies/act/processor_act.py @@ -18,17 +18,10 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_pre_post_processors, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_act import ACTConfig @@ -54,34 +47,4 @@ def make_act_pre_post_processors( tuple[PolicyProcessorPipeline[dict[str, Any], dict[str, Any]], PolicyProcessorPipeline[PolicyAction, PolicyAction]]: A tuple containing the pre-processor pipeline and the post-processor pipeline. """ - - input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - device=config.device, - ), - ] - output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), - ] - - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_default_pre_post_processors(config, dataset_stats, normalizer_device=config.device) diff --git a/src/lerobot/policies/common/flow_matching.py b/src/lerobot/policies/common/flow_matching.py new file mode 100644 index 000000000..f66b0aacf --- /dev/null +++ b/src/lerobot/policies/common/flow_matching.py @@ -0,0 +1,122 @@ +#!/usr/bin/env python + +# Copyright 2026 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Flow-matching sampling primitives shared across policies. + +Canonical versions of the beta-distributed timestep sampler and the forward-Euler +denoising loop (with its real-time-chunking hook) that the openpi-derived policies +(pi0, pi05, smolvla, eo1) historically each carried a copy of. All functions are +stateless; adopting them does not affect checkpoints. +""" + +from collections.abc import Callable +from typing import TYPE_CHECKING + +import torch +from torch import Tensor + +if TYPE_CHECKING: + from lerobot.policies.rtc.modeling_rtc import RTCProcessor + + +def sample_beta(alpha: float, beta: float, bsize: int, device) -> Tensor: # see openpi (exact copy) + # Beta sampling uses _sample_dirichlet which isn't implemented for MPS, so sample on CPU + alpha_t = torch.tensor(alpha, dtype=torch.float32) + beta_t = torch.tensor(beta, dtype=torch.float32) + dist = torch.distributions.Beta(alpha_t, beta_t) + return dist.sample((bsize,)).to(device) + + +def sample_noise(shape, device) -> Tensor: + """Standard-normal float32 noise, the flow-matching x_1 sample.""" + return torch.normal( + mean=0.0, + std=1.0, + size=shape, + dtype=torch.float32, + device=device, + ) + + +def sample_time_beta(bsize: int, device, *, alpha: float, beta: float, scale: float, offset: float) -> Tensor: + """Beta-distributed flow-matching timesteps: ``Beta(alpha, beta) * scale + offset`` (openpi convention).""" + time_beta = sample_beta(alpha, beta, bsize, device) + time = time_beta * scale + offset + return time.to(dtype=torch.float32, device=device) + + +def euler_integrate( + denoise_fn: Callable[[Tensor, Tensor], Tensor], + noise: Tensor, + num_steps: int, + *, + rtc_processor: "RTCProcessor | None" = None, + rtc_enabled: bool = False, + inference_delay: int | None = None, + prev_chunk_left_over: Tensor | None = None, + execution_horizon: int | None = None, +) -> Tensor: + """Forward-Euler integration of a velocity field from t=1 (noise) to t=0 (actions). + + This is the openpi sampling loop: ``dt = -1/num_steps``, ``time = 1.0 + step*dt``, + ``x_t <- x_t + dt * v_t``, with the optional real-time-chunking (RTC) guidance hook + wrapping the velocity computation and debug tracking after each step. + + Args: + denoise_fn: Computes the velocity ``v_t`` from ``(x_t, time_tensor)`` where + ``time_tensor`` is a float32 tensor of shape ``(batch_size,)``. The returned + velocity must have the same shape and dtype as ``x_t``. + noise: Initial sample ``x_1`` of shape ``(batch_size, ...)``. + num_steps: Number of Euler steps. + rtc_processor: Optional RTC processor. Debug tracking fires whenever it is set and + has debugging enabled, even if RTC guidance itself is disabled (this mirrors + the historical per-policy loops). + rtc_enabled: Whether to route the velocity computation through + ``rtc_processor.denoise_step`` (requires ``rtc_processor``). + inference_delay: RTC guidance parameter, forwarded verbatim. + prev_chunk_left_over: RTC guidance parameter, forwarded verbatim. + execution_horizon: RTC guidance parameter, forwarded verbatim. + """ + bsize = noise.shape[0] + device = noise.device + + dt = -1.0 / num_steps + x_t = noise + for step in range(num_steps): + time = 1.0 + step * dt + time_tensor = torch.tensor(time, dtype=torch.float32, device=device).expand(bsize) + + def denoise_step_partial_call(input_x_t, current_timestep=time_tensor): + return denoise_fn(input_x_t, current_timestep) + + if rtc_enabled: + v_t = rtc_processor.denoise_step( + x_t=x_t, + prev_chunk_left_over=prev_chunk_left_over, + inference_delay=inference_delay, + time=time, + original_denoise_step_partial=denoise_step_partial_call, + execution_horizon=execution_horizon, + ) + else: + v_t = denoise_step_partial_call(x_t) + + x_t = x_t + dt * v_t + + if rtc_processor is not None and rtc_processor.is_debug_enabled(): + rtc_processor.track(time=time, x_t=x_t, v_t=v_t) + + return x_t diff --git a/src/lerobot/policies/common/vla_utils.py b/src/lerobot/policies/common/vla_utils.py new file mode 100644 index 000000000..b728a49f4 --- /dev/null +++ b/src/lerobot/policies/common/vla_utils.py @@ -0,0 +1,243 @@ +#!/usr/bin/env python + +# Copyright 2026 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Helpers shared by the openpi-derived VLA policies (pi0, pi05, pi0_fast, smolvla, eo1, xvla). + +These are the canonical versions of functions that historically were copy-pasted per +policy. They are pure (no parameters, no module state), so importing them from here +instead of a policy-local copy has no effect on checkpoints. +""" + +import math +from typing import TYPE_CHECKING + +import torch +import torch.nn.functional as F # noqa: N812 +from torch import Tensor + +from lerobot.utils.constants import OPENPI_ATTENTION_MASK_VALUE +from lerobot.utils.device_utils import get_safe_dtype +from lerobot.utils.import_utils import _transformers_available, require_package + +if TYPE_CHECKING or _transformers_available: + from transformers import DynamicCache +else: + DynamicCache = None + + +def create_sinusoidal_pos_embedding( # see openpi `create_sinusoidal_pos_embedding` (exact copy) + time: torch.Tensor, dimension: int, min_period: float, max_period: float, device="cpu" +) -> Tensor: + """Computes sine-cosine positional embedding vectors for scalar positions.""" + if dimension % 2 != 0: + raise ValueError(f"dimension ({dimension}) must be divisible by 2") + + if time.ndim != 1: + raise ValueError("The time tensor is expected to be of shape `(batch_size, )`.") + + dtype = get_safe_dtype(torch.float64, device.type) + fraction = torch.linspace(0.0, 1.0, dimension // 2, dtype=dtype, device=device) + period = min_period * (max_period / min_period) ** fraction + + # Compute the outer product + scaling_factor = 1.0 / period * 2 * math.pi + sin_input = scaling_factor[None, :] * time[:, None] + return torch.cat([torch.sin(sin_input), torch.cos(sin_input)], dim=1) + + +def make_att_2d_masks(pad_masks: Tensor, att_masks: Tensor) -> Tensor: # see openpi (exact copy) + """Copied from big_vision. + + Tokens can attend to valid inputs tokens which have a cumulative mask_ar + smaller or equal to theirs. This way `mask_ar` int[B, N] can be used to + setup several types of attention, for example: + + [[1 1 1 1 1 1]]: pure causal attention. + + [[0 0 0 1 1 1]]: prefix-lm attention. The first 3 tokens can attend between + themselves and the last 3 tokens have a causal attention. The first + entry could also be a 1 without changing behaviour. + + [[1 0 1 0 1 0 0 1 0 0]]: causal attention between 4 blocks. Tokens of a + block can attend all previous blocks and all tokens on the same block. + + Args: + input_mask: bool[B, N] true if its part of the input, false if padding. + mask_ar: int32[B, N] mask that's 1 where previous tokens cannot depend on + it and 0 where it shares the same attention mask as the previous token. + """ + if att_masks.ndim != 2: + raise ValueError(att_masks.ndim) + if pad_masks.ndim != 2: + raise ValueError(pad_masks.ndim) + + cumsum = torch.cumsum(att_masks, dim=1) + att_2d_masks = cumsum[:, None, :] <= cumsum[:, :, None] + pad_2d_masks = pad_masks[:, None, :] * pad_masks[:, :, None] + return att_2d_masks & pad_2d_masks + + +def prepare_attention_masks_4d(att_2d_masks: Tensor, dtype: torch.dtype | None = None) -> Tensor: + """Expand boolean 2D attention masks to the additive 4D layout expected by transformers. + + Valid positions become 0.0 and masked positions the large negative openpi constant. + """ + att_2d_masks_4d = att_2d_masks[:, None, :, :] + result = torch.where(att_2d_masks_4d, 0.0, OPENPI_ATTENTION_MASK_VALUE) + if dtype is not None: + result = result.to(dtype=dtype) + return result + + +def clone_past_key_values(past_key_values): + """Clone the DynamicCache returned by prefix prefill for compiled denoising.""" + if DynamicCache is None: + require_package("transformers", extra="transformers-dep") + + return DynamicCache( + tuple( + (keys.clone(), values.clone(), sliding_window) for keys, values, sliding_window in past_key_values + ) + ) + + +def pad_vector(vector: Tensor, new_dim: int, *, truncate: bool = False) -> Tensor: + """Pad the last dimension of a vector to new_dim with zeros. + + Can be (batch_size x sequence_length x features_dimension) + or (batch_size x features_dimension) + + With ``truncate=False`` (openpi behavior), vectors whose last dimension is already + >= new_dim are returned unchanged. With ``truncate=True`` (xVLA behavior), the last + dimension is truncated to exactly ``new_dim`` (which may be 0). + """ + if vector.shape[-1] == new_dim: + return vector + if not truncate: + if vector.shape[-1] >= new_dim: + return vector + return F.pad(vector, (0, new_dim - vector.shape[-1])) + shape = list(vector.shape) + current_dim = shape[-1] + shape[-1] = new_dim + new_vector = vector.new_zeros(*shape) + length = min(current_dim, new_dim) + new_vector[..., :length] = vector[..., :length] + return new_vector + + +def resize_with_pad_torch( # see openpi `resize_with_pad_torch` (exact copy) + images: torch.Tensor, + height: int, + width: int, + mode: str = "bilinear", +) -> torch.Tensor: + """PyTorch version of resize_with_pad. Resizes an image to a target height and width without distortion + by padding with black. If the image is float32, it must be in the range [-1, 1]. + + Padding is centered (openpi convention). For the top-left-padding variant used by + smolvla/xvla, see :func:`resize_with_pad`. + + Args: + images: Tensor of shape [*b, h, w, c] or [*b, c, h, w] + height: Target height + width: Target width + mode: Interpolation mode ('bilinear', 'nearest', etc.) + + Returns: + Resized and padded tensor with same shape format as input + """ + # Check if input is in channels-last format [*b, h, w, c] or channels-first [*b, c, h, w] + if images.shape[-1] <= 4: # Assume channels-last format + channels_last = True + if images.dim() == 3: + images = images.unsqueeze(0) # Add batch dimension + images = images.permute(0, 3, 1, 2) # [b, h, w, c] -> [b, c, h, w] + else: + channels_last = False + if images.dim() == 3: + images = images.unsqueeze(0) # Add batch dimension + + batch_size, channels, cur_height, cur_width = images.shape + + # Calculate resize ratio + ratio = max(cur_width / width, cur_height / height) + resized_height = int(cur_height / ratio) + resized_width = int(cur_width / ratio) + + # Resize + resized_images = F.interpolate( + images, + size=(resized_height, resized_width), + mode=mode, + align_corners=False if mode == "bilinear" else None, + ) + + # Handle dtype-specific clipping + if images.dtype == torch.uint8: + resized_images = torch.round(resized_images).clamp(0, 255).to(torch.uint8) + elif images.dtype == torch.float32: + resized_images = resized_images.clamp(0.0, 1.0) + else: + raise ValueError(f"Unsupported image dtype: {images.dtype}") + + # Calculate padding + pad_h0, remainder_h = divmod(height - resized_height, 2) + pad_h1 = pad_h0 + remainder_h + pad_w0, remainder_w = divmod(width - resized_width, 2) + pad_w1 = pad_w0 + remainder_w + + # Pad + constant_value = 0 if images.dtype == torch.uint8 else 0.0 + padded_images = F.pad( + resized_images, + (pad_w0, pad_w1, pad_h0, pad_h1), # left, right, top, bottom + mode="constant", + value=constant_value, + ) + + # Convert back to original format if needed + if channels_last: + padded_images = padded_images.permute(0, 2, 3, 1) # [b, c, h, w] -> [b, h, w, c] + + return padded_images + + +def resize_with_pad(img: torch.Tensor, height: int, width: int, *, pad_value: float) -> torch.Tensor: + """Resize a (b, c, h, w) image without distortion, padding on the LEFT and TOP. + + This is the smolvla/xvla convention. For the centered-padding openpi variant, see + :func:`resize_with_pad_torch`. ``pad_value`` is keyword-only on purpose: callers + historically used different values (0, -1) and must state their choice explicitly. + """ + if img.ndim != 4: + raise ValueError(f"(b,c,h,w) expected, but got {img.shape}") + + current_height, current_width = img.shape[2:] + if current_height == height and current_width == width: + return img + + ratio = max(current_width / width, current_height / height) + resized_height = int(current_height / ratio) + resized_width = int(current_width / ratio) + resized_img = F.interpolate( + img, size=(resized_height, resized_width), mode="bilinear", align_corners=False + ) + + pad_height = max(0, height - resized_height) + pad_width = max(0, width - resized_width) + padded_img = F.pad(resized_img, (pad_width, 0, pad_height, 0), value=pad_value) + return padded_img diff --git a/src/lerobot/policies/diffusion/processor_diffusion.py b/src/lerobot/policies/diffusion/processor_diffusion.py index c4bc17680..28d77dc83 100644 --- a/src/lerobot/policies/diffusion/processor_diffusion.py +++ b/src/lerobot/policies/diffusion/processor_diffusion.py @@ -19,17 +19,10 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_pre_post_processors, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_diffusion import DiffusionConfig @@ -63,32 +56,4 @@ def make_diffusion_pre_post_processors( Returns: A tuple containing the configured pre-processor and post-processor pipelines. """ - - input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - ] - output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), - ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_default_pre_post_processors(config, dataset_stats) diff --git a/src/lerobot/policies/eo1/processor_eo1.py b/src/lerobot/policies/eo1/processor_eo1.py index b1f32756a..1d3dc0515 100644 --- a/src/lerobot/policies/eo1/processor_eo1.py +++ b/src/lerobot/policies/eo1/processor_eo1.py @@ -23,24 +23,16 @@ import torch from lerobot.configs.types import FeatureType, PipelineFeatureType, PolicyFeature from lerobot.processor import ( - AddBatchDimensionProcessorStep, ComplementaryDataProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.processor.converters import policy_action_to_transition, transition_to_policy_action from lerobot.types import TransitionKey -from lerobot.utils.constants import ( - OBS_STATE, - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, -) +from lerobot.utils.constants import OBS_STATE from lerobot.utils.import_utils import _transformers_available, require_package from .configuration_eo1 import EO1Config @@ -242,14 +234,12 @@ def make_eo1_pre_post_processors( ]: """Build pre/post processor pipelines for EO1.""" + steps = make_default_policy_processor_steps(config, dataset_stats) + input_steps: list[ProcessorStep] = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.rename_observations, + steps.add_batch_dim, + steps.normalize, EO1ConversationTemplateStep(input_features=config.input_features, chunk_size=config.chunk_size), EO1QwenProcessorStep( processor_name=config.vlm_base, @@ -257,27 +247,12 @@ def make_eo1_pre_post_processors( image_max_pixels=config.image_max_pixels, use_fast_processor=config.use_fast_processor, ), - DeviceProcessorStep(device=config.device), + steps.to_device, ] output_steps: list[ProcessorStep] = [ - UnnormalizerProcessorStep( - features=config.output_features, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - DeviceProcessorStep(device="cpu"), + steps.unnormalize, + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/factory.py b/src/lerobot/policies/factory.py index 73fd9455f..36a0de7ca 100644 --- a/src/lerobot/policies/factory.py +++ b/src/lerobot/policies/factory.py @@ -17,6 +17,7 @@ from __future__ import annotations import importlib +import inspect import logging from typing import TYPE_CHECKING, Any, TypedDict, Unpack @@ -44,26 +45,10 @@ from lerobot.utils.constants import ( ) from lerobot.utils.feature_utils import dataset_to_policy_features -from .act.configuration_act import ACTConfig -from .diffusion.configuration_diffusion import DiffusionConfig -from .eo1.configuration_eo1 import EO1Config from .evo1.configuration_evo1 import Evo1Config -from .fastwam.configuration_fastwam import FastWAMConfig -from .gaussian_actor.configuration_gaussian_actor import GaussianActorConfig from .groot.configuration_groot import GrootConfig -from .lingbot_va.configuration_lingbot_va import LingBotVAConfig -from .molmoact2.configuration_molmoact2 import MolmoAct2Config -from .multi_task_dit.configuration_multi_task_dit import MultiTaskDiTConfig -from .pi0.configuration_pi0 import PI0Config -from .pi05.configuration_pi05 import PI05Config from .pretrained import PreTrainedPolicy -from .smolvla.configuration_smolvla import SmolVLAConfig -from .tdmpc.configuration_tdmpc import TDMPCConfig from .utils import validate_visual_features_consistency -from .vla_jepa.configuration_vla_jepa import VLAJEPAConfig -from .vqbet.configuration_vqbet import VQBeTConfig -from .wall_x.configuration_wall_x import WallXConfig -from .xvla.configuration_xvla import XVLAConfig def _reconnect_relative_absolute_steps( @@ -88,100 +73,23 @@ def get_policy_class(name: str) -> type[PreTrainedPolicy]: """ Retrieves a policy class by its registered name. - This function uses dynamic imports to avoid loading all policy classes into memory - at once, improving startup time and reducing dependencies. + Resolution is convention-based: the draccus-registered config class of ``name`` is + looked up, its ``configuration_*`` module path is rewritten to ``modeling_*``, and + the ``Policy`` class is imported from there. The modeling module is only imported + at call time, keeping heavy optional dependencies lazy. This works for both built-in + policies and third-party lerobot plugins (anything registered via + ``@PreTrainedConfig.register_subclass``). Args: - name: The name of the policy. Supported names are "tdmpc", "diffusion", "act", - "multi_task_dit", "vqbet", "pi0", "pi05", "gaussian_actor", "smolvla", "wall_x", - "molmoact2", "eo1", "evo1". + name: The registered name of the policy (e.g. "act", "diffusion", "pi0"). Returns: The policy class corresponding to the given name. Raises: - NotImplementedError: If the policy name is not recognized. + ValueError: If the policy name is not registered. + ImportError: If the policy's optional dependencies are not installed. """ - if name == "tdmpc": - from .tdmpc.modeling_tdmpc import TDMPCPolicy - - return TDMPCPolicy - elif name == "diffusion": - from .diffusion.modeling_diffusion import DiffusionPolicy - - return DiffusionPolicy - elif name == "act": - from .act.modeling_act import ACTPolicy - - return ACTPolicy - elif name == "multi_task_dit": - from .multi_task_dit.modeling_multi_task_dit import MultiTaskDiTPolicy - - return MultiTaskDiTPolicy - elif name == "vqbet": - from .vqbet.modeling_vqbet import VQBeTPolicy - - return VQBeTPolicy - elif name == "pi0": - from .pi0.modeling_pi0 import PI0Policy - - return PI0Policy - elif name == "pi0_fast": - from .pi0_fast.modeling_pi0_fast import PI0FastPolicy - - return PI0FastPolicy - elif name == "pi05": - from .pi05.modeling_pi05 import PI05Policy - - return PI05Policy - elif name == "gaussian_actor": - from .gaussian_actor.modeling_gaussian_actor import GaussianActorPolicy - - return GaussianActorPolicy - elif name == "smolvla": - from .smolvla.modeling_smolvla import SmolVLAPolicy - - return SmolVLAPolicy - elif name == "groot": - from .groot.modeling_groot import GrootPolicy - - return GrootPolicy - elif name == "xvla": - from .xvla.modeling_xvla import XVLAPolicy - - return XVLAPolicy - elif name == "wall_x": - from .wall_x.modeling_wall_x import WallXPolicy - - return WallXPolicy - elif name == "eo1": - from .eo1.modeling_eo1 import EO1Policy - - return EO1Policy - elif name == "molmoact2": - from .molmoact2.modeling_molmoact2 import MolmoAct2Policy - - return MolmoAct2Policy - elif name == "vla_jepa": - from .vla_jepa.modeling_vla_jepa import VLAJEPAPolicy - - return VLAJEPAPolicy - elif name == "lingbot_va": - from .lingbot_va.modeling_lingbot_va import LingBotVAPolicy - - return LingBotVAPolicy - elif name == "fastwam": - from .fastwam.modeling_fastwam import FastWAMPolicy - - return FastWAMPolicy - elif name == "evo1": - from .evo1.modeling_evo1 import Evo1Policy - - return Evo1Policy - else: - try: - return _get_policy_cls_from_policy_name(name=name) - except Exception as e: - raise ValueError(f"Policy type '{name}' is not available.") from e + return _get_policy_cls_from_policy_name(name=name) def make_policy_config(policy_type: str, **kwargs) -> PreTrainedConfig: @@ -192,9 +100,8 @@ def make_policy_config(policy_type: str, **kwargs) -> PreTrainedConfig: mapping a string identifier to the corresponding config class. Args: - policy_type: The type of the policy. Supported types include "tdmpc", - "multi_task_dit", "diffusion", "act", "vqbet", "pi0", "pi05", "gaussian_actor", - "smolvla", "wall_x", "molmoact2", "eo1", "evo1". + policy_type: The registered type of the policy (any name registered via + ``@PreTrainedConfig.register_subclass``, e.g. "act", "diffusion", "pi0"). **kwargs: Keyword arguments to be passed to the configuration class constructor. Returns: @@ -203,48 +110,11 @@ def make_policy_config(policy_type: str, **kwargs) -> PreTrainedConfig: Raises: ValueError: If the `policy_type` is not recognized. """ - if policy_type == "tdmpc": - return TDMPCConfig(**kwargs) - elif policy_type == "diffusion": - return DiffusionConfig(**kwargs) - elif policy_type == "act": - return ACTConfig(**kwargs) - elif policy_type == "multi_task_dit": - return MultiTaskDiTConfig(**kwargs) - elif policy_type == "vqbet": - return VQBeTConfig(**kwargs) - elif policy_type == "pi0": - return PI0Config(**kwargs) - elif policy_type == "pi05": - return PI05Config(**kwargs) - elif policy_type == "gaussian_actor": - return GaussianActorConfig(**kwargs) - elif policy_type == "smolvla": - return SmolVLAConfig(**kwargs) - elif policy_type == "groot": - return GrootConfig(**kwargs) - elif policy_type == "xvla": - return XVLAConfig(**kwargs) - elif policy_type == "wall_x": - return WallXConfig(**kwargs) - elif policy_type == "eo1": - return EO1Config(**kwargs) - elif policy_type == "molmoact2": - return MolmoAct2Config(**kwargs) - elif policy_type == "vla_jepa": - return VLAJEPAConfig(**kwargs) - elif policy_type == "lingbot_va": - return LingBotVAConfig(**kwargs) - elif policy_type == "fastwam": - return FastWAMConfig(**kwargs) - elif policy_type == "evo1": - return Evo1Config(**kwargs) - else: - try: - config_cls = PreTrainedConfig.get_choice_class(policy_type) - return config_cls(**kwargs) - except Exception as e: - raise ValueError(f"Policy type '{policy_type}' is not available.") from e + try: + config_cls = PreTrainedConfig.get_choice_class(policy_type) + except Exception as e: + raise ValueError(f"Policy type '{policy_type}' is not available.") from e + return config_cls(**kwargs) class ProcessorConfigKwargs(TypedDict, total=False): @@ -298,8 +168,7 @@ def make_pre_post_processors( A tuple containing the input (pre-processor) and output (post-processor) pipelines. Raises: - NotImplementedError: If a processor factory is not implemented for the given - policy configuration type. + ValueError: If no processor factory exists for the given policy configuration type. """ if pretrained_path: if isinstance(policy_cfg, GrootConfig): @@ -351,166 +220,13 @@ def make_pre_post_processors( ) return preprocessor, postprocessor - # Create a new processor based on policy type - if isinstance(policy_cfg, TDMPCConfig): - from .tdmpc.processor_tdmpc import make_tdmpc_pre_post_processors - - processors = make_tdmpc_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, DiffusionConfig): - from .diffusion.processor_diffusion import make_diffusion_pre_post_processors - - processors = make_diffusion_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, ACTConfig): - from .act.processor_act import make_act_pre_post_processors - - processors = make_act_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, MultiTaskDiTConfig): - from .multi_task_dit.processor_multi_task_dit import ( - make_multi_task_dit_pre_post_processors, - ) - - processors = make_multi_task_dit_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, VQBeTConfig): - from .vqbet.processor_vqbet import make_vqbet_pre_post_processors - - processors = make_vqbet_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, PI0Config): - from .pi0.processor_pi0 import make_pi0_pre_post_processors - - processors = make_pi0_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, PI05Config): - from .pi05.processor_pi05 import make_pi05_pre_post_processors - - processors = make_pi05_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, GaussianActorConfig): - from .gaussian_actor.processor_gaussian_actor import make_gaussian_actor_pre_post_processors - - processors = make_gaussian_actor_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, SmolVLAConfig): - from .smolvla.processor_smolvla import make_smolvla_pre_post_processors - - processors = make_smolvla_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, GrootConfig): - from .groot.processor_groot import make_groot_pre_post_processors - - processors = make_groot_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - dataset_meta=kwargs.get("dataset_meta"), - ) - - elif isinstance(policy_cfg, XVLAConfig): - from .xvla.processor_xvla import ( - make_xvla_pre_post_processors, - ) - - processors = make_xvla_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, WallXConfig): - from .wall_x.processor_wall_x import make_wall_x_pre_post_processors - - processors = make_wall_x_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, EO1Config): - from .eo1.processor_eo1 import make_eo1_pre_post_processors - - processors = make_eo1_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - elif isinstance(policy_cfg, Evo1Config): - from .evo1.processor_evo1 import make_evo1_pre_post_processors - - processors = make_evo1_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, MolmoAct2Config): - from .molmoact2.processor_molmoact2 import make_molmoact2_pre_post_processors - - processors = make_molmoact2_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - dataset_meta=kwargs.get("dataset_meta"), - ) - - elif isinstance(policy_cfg, VLAJEPAConfig): - from .vla_jepa.processor_vla_jepa import make_vla_jepa_pre_post_processors - - processors = make_vla_jepa_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, LingBotVAConfig): - from .lingbot_va.processor_lingbot_va import make_lingbot_va_pre_post_processors - - processors = make_lingbot_va_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - elif isinstance(policy_cfg, FastWAMConfig): - from .fastwam.processor_fastwam import make_fastwam_pre_post_processors - - processors = make_fastwam_pre_post_processors( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - - else: - try: - processors = _make_processors_from_policy_config( - config=policy_cfg, - dataset_stats=kwargs.get("dataset_stats"), - ) - except Exception as e: - raise ValueError(f"Processor for policy type '{policy_cfg.type}' is not implemented.") from e - - return processors + # Create new processors from the policy config, resolving the per-policy factory + # function by naming convention (lazy import keeps optional dependencies optional). + return _make_processors_from_policy_config( + config=policy_cfg, + dataset_stats=kwargs.get("dataset_stats"), + dataset_meta=kwargs.get("dataset_meta"), + ) def make_policy( @@ -654,10 +370,12 @@ def make_policy( return policy -def _get_policy_cls_from_policy_name(name: str) -> type[PreTrainedConfig]: +def _get_policy_cls_from_policy_name(name: str) -> type[PreTrainedPolicy]: """Get policy class from its registered name using dynamic imports. - This is used as a helper function to import policies from 3rd party lerobot plugins. + Works for built-in policies and 3rd party lerobot plugins alike: the config class + registered under ``name`` is resolved via the draccus ChoiceRegistry, and the policy + class is imported from the sibling ``modeling_*`` module by naming convention. Args: name: The name of the policy. @@ -683,22 +401,39 @@ def _get_policy_cls_from_policy_name(name: str) -> type[PreTrainedConfig]: "configuration_", "modeling_" ) # e.g., configuration_diffusion -> modeling_diffusion - module = importlib.import_module(module_path) - policy_cls = getattr(module, cls_name) + try: + module = importlib.import_module(module_path) + except ModuleNotFoundError as e: + if e.name == module_path: + # The modeling_* module itself does not exist for this policy type. A missing + # optional dependency inside an existing module propagates unchanged instead, + # so its actionable install hint stays visible. + raise ValueError(f"Policy class for '{name}' is not implemented.") from e + raise + policy_cls = getattr(module, cls_name, None) + if policy_cls is None: + raise ValueError( + f"Policy class '{cls_name}' not found in '{module_path}'. " + f"Policies must expose 'Policy' in the sibling 'modeling_*' module by naming convention." + ) return policy_cls def _make_processors_from_policy_config( config: PreTrainedConfig, dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None, + dataset_meta: Any | None = None, ) -> tuple[Any, Any]: """Create pre- and post-processors from a policy configuration using dynamic imports. - This is used as a helper function to import processor factories from 3rd party lerobot plugins. + Resolves ``make_{type}_pre_post_processors`` from the policy's ``processor_*`` module + by naming convention. Works for built-in policies and 3rd party lerobot plugins. Args: config: The policy configuration object. dataset_stats: Dataset statistics for normalization. + dataset_meta: Dataset metadata, forwarded only to factories that declare a + ``dataset_meta`` parameter (e.g. groot, molmoact2). Returns: A tuple containing the input (pre-processor) and output (post-processor) pipelines. """ @@ -711,6 +446,19 @@ def _make_processors_from_policy_config( logging.debug( f"Instantiating pre/post processors using function '{function_name}' from module '{module_path}'" ) - module = importlib.import_module(module_path) - function = getattr(module, function_name) - return function(config, dataset_stats=dataset_stats) + try: + module = importlib.import_module(module_path) + except ModuleNotFoundError as e: + if e.name == module_path: + # The processor_* module itself does not exist for this policy type. A missing + # optional dependency inside an existing module propagates unchanged instead, + # so its actionable install hint stays visible. + raise ValueError(f"Processor for policy type '{policy_type}' is not implemented.") from e + raise + function = getattr(module, function_name, None) + if function is None: + raise ValueError(f"Processor for policy type '{policy_type}' is not implemented.") + call_kwargs: dict[str, Any] = {"dataset_stats": dataset_stats} + if "dataset_meta" in inspect.signature(function).parameters: + call_kwargs["dataset_meta"] = dataset_meta + return function(config, **call_kwargs) diff --git a/src/lerobot/policies/fastwam/processor_fastwam.py b/src/lerobot/policies/fastwam/processor_fastwam.py index 31f3b9277..318c7ff97 100644 --- a/src/lerobot/policies/fastwam/processor_fastwam.py +++ b/src/lerobot/policies/fastwam/processor_fastwam.py @@ -22,20 +22,11 @@ import torch from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( ActionProcessorStep, - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStepRegistry, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, -) -from lerobot.utils.constants import ( - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) from .configuration_fastwam import FastWAMConfig @@ -105,38 +96,20 @@ def make_fastwam_pre_post_processors( # anyway) and unsafe across fine-tuning: its `resize_size` would be inherited from the base # checkpoint's camera geometry, not this dataset's, making the concatenation N_cameras x too wide. + steps = make_default_policy_processor_steps(config, normalization_stats, normalizer_device=config.device) + input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=normalization_stats, - device=config.device, - ), + steps.rename_observations, + steps.add_batch_dim, + steps.to_device, + steps.normalize, ] output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, - norm_map=config.normalization_mapping, - stats=normalization_stats, - ), + steps.unnormalize, ] if config.toggle_action_dimensions: output_steps.append( FastWAMActionToggleProcessorStep(toggle_dimensions=config.toggle_action_dimensions) ) - output_steps.append(DeviceProcessorStep(device="cpu")) - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + output_steps.append(steps.to_cpu) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/gaussian_actor/processor_gaussian_actor.py b/src/lerobot/policies/gaussian_actor/processor_gaussian_actor.py index 1e930d178..fede487ff 100644 --- a/src/lerobot/policies/gaussian_actor/processor_gaussian_actor.py +++ b/src/lerobot/policies/gaussian_actor/processor_gaussian_actor.py @@ -20,17 +20,10 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_pre_post_processors, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_gaussian_actor import GaussianActorConfig @@ -62,33 +55,4 @@ def make_gaussian_actor_pre_post_processors( Returns: A tuple containing the configured pre-processor and post-processor pipelines. """ - - # Add remaining processors - input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - ] - output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), - ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_default_pre_post_processors(config, dataset_stats) diff --git a/src/lerobot/policies/lingbot_va/processor_lingbot_va.py b/src/lerobot/policies/lingbot_va/processor_lingbot_va.py index 119f77d5b..d183b8c34 100644 --- a/src/lerobot/policies/lingbot_va/processor_lingbot_va.py +++ b/src/lerobot/policies/lingbot_va/processor_lingbot_va.py @@ -25,19 +25,12 @@ import torch from lerobot.configs.types import FeatureType, NormalizationMode from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, - RenameObservationsProcessorStep, UnnormalizerProcessorStep, -) -from lerobot.processor.converters import policy_action_to_transition, transition_to_policy_action -from lerobot.utils.constants import ( - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) from .configuration_lingbot_va import LingBotVAConfig @@ -52,15 +45,13 @@ def make_lingbot_va_pre_post_processors( ]: """Build the pre/post processor pipelines for LingBot-VA.""" + steps = make_default_policy_processor_steps(config, dataset_stats) + input_steps: list[ProcessorStep] = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - DeviceProcessorStep(device=config.device), + steps.rename_observations, + steps.add_batch_dim, + steps.normalize, + steps.to_device, ] # Unnormalize actions from [-1, 1] to physical units (QUANTILES) using q01/q99 restored from the checkpoint. @@ -70,18 +61,7 @@ def make_lingbot_va_pre_post_processors( norm_map={FeatureType.ACTION: NormalizationMode.QUANTILES}, stats=dataset_stats, ), - DeviceProcessorStep(device="cpu"), + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/multi_task_dit/processor_multi_task_dit.py b/src/lerobot/policies/multi_task_dit/processor_multi_task_dit.py index 5f5b9994e..7cb49d4ea 100644 --- a/src/lerobot/policies/multi_task_dit/processor_multi_task_dit.py +++ b/src/lerobot/policies/multi_task_dit/processor_multi_task_dit.py @@ -19,18 +19,12 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_multi_task_dit import MultiTaskDiTConfig @@ -66,9 +60,11 @@ def make_multi_task_dit_pre_post_processors( A tuple containing the configured pre-processor and post-processor pipelines. """ + steps = make_default_policy_processor_steps(config, dataset_stats, normalizer_device=config.device) + input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), + steps.rename_observations, + steps.add_batch_dim, TokenizerProcessorStep( tokenizer_name=config.text_encoder_name, padding=config.tokenizer_padding, @@ -76,32 +72,12 @@ def make_multi_task_dit_pre_post_processors( max_length=config.tokenizer_max_length, truncation=config.tokenizer_truncation, ), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - device=config.device, - ), + steps.to_device, + steps.normalize, ] output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - DeviceProcessorStep(device="cpu"), + steps.unnormalize, + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/pi0/processor_pi0.py b/src/lerobot/policies/pi0/processor_pi0.py index ad861f85b..ea6d662bd 100644 --- a/src/lerobot/policies/pi0/processor_pi0.py +++ b/src/lerobot/policies/pi0/processor_pi0.py @@ -21,22 +21,16 @@ import torch from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( AbsoluteActionsProcessorStep, - AddBatchDimensionProcessorStep, ComplementaryDataProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, RelativeActionsProcessorStep, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_pi0 import PI0Config @@ -136,10 +130,12 @@ def make_pi0_pre_post_processors( action_names=getattr(config, "action_feature_names", None), ) + steps = make_default_policy_processor_steps(config, dataset_stats) + # OpenPI order: raw → relative → normalize → model → unnormalize → absolute input_steps: list[ProcessorStep] = [ - RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one - AddBatchDimensionProcessorStep(), + steps.rename_observations, # To mimic the same processor as pretrained one + steps.add_batch_dim, Pi0NewLineProcessor(), # Add newlines before tokenization for PaliGemma TokenizerProcessorStep( tokenizer_name="google/paligemma-3b-pt-224", @@ -147,32 +143,15 @@ def make_pi0_pre_post_processors( padding_side="right", padding="max_length", ), - DeviceProcessorStep(device=config.device), + steps.to_device, relative_step, - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.normalize, ] output_steps: list[ProcessorStep] = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), + steps.unnormalize, AbsoluteActionsProcessorStep(enabled=config.use_relative_actions, relative_step=relative_step), - DeviceProcessorStep(device="cpu"), + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/pi05/processor_pi05.py b/src/lerobot/policies/pi05/processor_pi05.py index 2d015b24f..ac16269a0 100644 --- a/src/lerobot/policies/pi05/processor_pi05.py +++ b/src/lerobot/policies/pi05/processor_pi05.py @@ -24,26 +24,17 @@ import torch from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( AbsoluteActionsProcessorStep, - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, RelativeActionsProcessorStep, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) from lerobot.types import EnvTransition, TransitionKey -from lerobot.utils.constants import ( - OBS_STATE, - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, -) +from lerobot.utils.constants import OBS_STATE from .configuration_pi05 import PI05Config @@ -135,18 +126,16 @@ def make_pi05_pre_post_processors( action_names=getattr(config, "action_feature_names", None), ) + steps = make_default_policy_processor_steps(config, dataset_stats) + # OpenPI order: raw → relative → normalize → model → unnormalize → absolute input_steps: list[ProcessorStep] = [ - RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one - AddBatchDimensionProcessorStep(), + steps.rename_observations, # To mimic the same processor as pretrained one + steps.add_batch_dim, relative_step, # NOTE: NormalizerProcessorStep MUST come before Pi05PrepareStateTokenizerProcessorStep # because the tokenizer step expects normalized state in [-1, 1] range for discretization - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.normalize, Pi05PrepareStateTokenizerProcessorStep(max_state_dim=config.max_state_dim), TokenizerProcessorStep( tokenizer_name="google/paligemma-3b-pt-224", @@ -154,26 +143,13 @@ def make_pi05_pre_post_processors( padding_side="right", padding="max_length", ), - DeviceProcessorStep(device=config.device), + steps.to_device, ] output_steps: list[ProcessorStep] = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), + steps.unnormalize, AbsoluteActionsProcessorStep(enabled=config.use_relative_actions, relative_step=relative_step), - DeviceProcessorStep(device="cpu"), + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/pi0_fast/processor_pi0_fast.py b/src/lerobot/policies/pi0_fast/processor_pi0_fast.py index 60a519786..fdf759e29 100644 --- a/src/lerobot/policies/pi0_fast/processor_pi0_fast.py +++ b/src/lerobot/policies/pi0_fast/processor_pi0_fast.py @@ -25,26 +25,17 @@ from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( AbsoluteActionsProcessorStep, ActionTokenizerProcessorStep, - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, RelativeActionsProcessorStep, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) from lerobot.types import EnvTransition, TransitionKey -from lerobot.utils.constants import ( - OBS_STATE, - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, -) +from lerobot.utils.constants import OBS_STATE from .configuration_pi0_fast import PI0FastConfig @@ -135,6 +126,8 @@ def make_pi0_fast_pre_post_processors( action_names=getattr(config, "action_feature_names", None), ) + steps = make_default_policy_processor_steps(config, dataset_stats) + # Pi0Fast order: relative → normalize → tokenize → model → unnormalize → absolute # This matches pi0/pi0.5: RelativeActionsProcessorStep runs first on raw absolute actions, # caching the raw state. NormalizerProcessorStep then normalizes the raw relative actions, @@ -144,14 +137,10 @@ def make_pi0_fast_pre_post_processors( # before Pi0FastPrepareStateAndLanguageTokenizerProcessorStep, so the state tokenizer # continues to receive normalized state in [-1, 1] as expected. input_steps: list[ProcessorStep] = [ - RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one - AddBatchDimensionProcessorStep(), + steps.rename_observations, # To mimic the same processor as pretrained one + steps.add_batch_dim, relative_step, - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.normalize, Pi0FastPrepareStateAndLanguageTokenizerProcessorStep(max_state_dim=config.max_state_dim), TokenizerProcessorStep( tokenizer_name=config.text_tokenizer_name, @@ -165,26 +154,13 @@ def make_pi0_fast_pre_post_processors( fast_skip_tokens=config.fast_skip_tokens, paligemma_tokenizer_name=config.text_tokenizer_name, ), - DeviceProcessorStep(device=config.device), + steps.to_device, ] output_steps: list[ProcessorStep] = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), + steps.unnormalize, AbsoluteActionsProcessorStep(enabled=config.use_relative_actions, relative_step=relative_step), - DeviceProcessorStep(device="cpu"), + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/pretrained.py b/src/lerobot/policies/pretrained.py index 03a624fef..90d894e30 100644 --- a/src/lerobot/policies/pretrained.py +++ b/src/lerobot/policies/pretrained.py @@ -32,6 +32,7 @@ from torch import Tensor, nn from lerobot.__version__ import __version__ from lerobot.configs import PreTrainedConfig from lerobot.configs.train import TrainPipelineConfig +from lerobot.utils.device_utils import resolve_safetensors_device from lerobot.utils.hub import HubMixin from .utils import log_model_loading_keys @@ -220,7 +221,7 @@ class PreTrainedPolicy(nn.Module, HubMixin, abc.ABC): @classmethod def _load_as_safetensor(cls, model: T, model_file: str, map_location: str, strict: bool) -> T: missing_keys, unexpected_keys = load_model_as_safetensor( - model, model_file, strict=strict, device=map_location + model, model_file, strict=strict, device=resolve_safetensors_device(map_location) ) log_model_loading_keys(missing_keys, unexpected_keys) return model diff --git a/src/lerobot/policies/smolvla/processor_smolvla.py b/src/lerobot/policies/smolvla/processor_smolvla.py index 8d6c8aca4..e924780cd 100644 --- a/src/lerobot/policies/smolvla/processor_smolvla.py +++ b/src/lerobot/policies/smolvla/processor_smolvla.py @@ -19,19 +19,13 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, NewLineTaskProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_smolvla import SmolVLAConfig @@ -66,9 +60,11 @@ def make_smolvla_pre_post_processors( A tuple containing the configured pre-processor and post-processor pipelines. """ + steps = make_default_policy_processor_steps(config, dataset_stats) + input_steps = [ - RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one - AddBatchDimensionProcessorStep(), + steps.rename_observations, # To mimic the same processor as pretrained one + steps.add_batch_dim, NewLineTaskProcessorStep(), TokenizerProcessorStep( tokenizer_name=config.vlm_model_name, @@ -76,28 +72,11 @@ def make_smolvla_pre_post_processors( padding_side="right", max_length=config.tokenizer_max_length, ), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.to_device, + steps.normalize, ] output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), + steps.unnormalize, + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/tdmpc/processor_tdmpc.py b/src/lerobot/policies/tdmpc/processor_tdmpc.py index 7afe956dc..08756e819 100644 --- a/src/lerobot/policies/tdmpc/processor_tdmpc.py +++ b/src/lerobot/policies/tdmpc/processor_tdmpc.py @@ -19,17 +19,10 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_pre_post_processors, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_tdmpc import TDMPCConfig @@ -61,32 +54,4 @@ def make_tdmpc_pre_post_processors( Returns: A tuple containing the configured pre-processor and post-processor pipelines. """ - - input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - ] - output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), - ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_default_pre_post_processors(config, dataset_stats) diff --git a/src/lerobot/policies/vla_jepa/processor_vla_jepa.py b/src/lerobot/policies/vla_jepa/processor_vla_jepa.py index b59cc0e90..aed2b754b 100644 --- a/src/lerobot/policies/vla_jepa/processor_vla_jepa.py +++ b/src/lerobot/policies/vla_jepa/processor_vla_jepa.py @@ -20,20 +20,16 @@ import torch from lerobot.policies.vla_jepa.configuration_vla_jepa import VLAJEPAConfig from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, EnvTransition, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, - RenameObservationsProcessorStep, TransitionKey, UnnormalizerProcessorStep, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.processor.converters import policy_action_to_transition, transition_to_policy_action -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME @ProcessorStepRegistry.register(name="vla_jepa_clip_actions") @@ -112,15 +108,12 @@ def make_vla_jepa_pre_post_processors( PolicyProcessorPipeline[PolicyAction, PolicyAction], ]: features = {**config.input_features, **config.output_features} + steps = make_default_policy_processor_steps(config, dataset_stats) input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features=features, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), + steps.rename_observations, + steps.add_batch_dim, + steps.to_device, + steps.normalize, ] output_steps: list[ProcessorStep] = [] if config.clip_normalized_actions: @@ -129,6 +122,8 @@ def make_vla_jepa_pre_post_processors( output_steps.append( PreSnapGripperProcessorStep(gripper_dim=config.gripper_dim, threshold=config.gripper_threshold) ) + # NOTE: unlike the default policy unnormalizer (output features only), VLA-JEPA + # unnormalizes over BOTH input and output features. output_steps.append( UnnormalizerProcessorStep( features=features, @@ -140,16 +135,5 @@ def make_vla_jepa_pre_post_processors( output_steps.append( BinarizeGripperProcessorStep(gripper_dim=config.gripper_dim, threshold=config.gripper_threshold) ) - output_steps.append(DeviceProcessorStep(device="cpu")) - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + output_steps.append(steps.to_cpu) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) diff --git a/src/lerobot/policies/vqbet/processor_vqbet.py b/src/lerobot/policies/vqbet/processor_vqbet.py index f7b6a061e..68182db3f 100644 --- a/src/lerobot/policies/vqbet/processor_vqbet.py +++ b/src/lerobot/policies/vqbet/processor_vqbet.py @@ -20,17 +20,10 @@ from typing import Any import torch from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_pre_post_processors, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_vqbet import VQBeTConfig @@ -62,32 +55,4 @@ def make_vqbet_pre_post_processors( Returns: A tuple containing the configured pre-processor and post-processor pipelines. """ - - input_steps = [ - RenameObservationsProcessorStep(rename_map={}), # Let the possibility to the user to rename the keys - AddBatchDimensionProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - ] - output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), - ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_default_pre_post_processors(config, dataset_stats) diff --git a/src/lerobot/policies/wall_x/configuration_wall_x.py b/src/lerobot/policies/wall_x/configuration_wall_x.py index 70576a46b..f26bf61e2 100644 --- a/src/lerobot/policies/wall_x/configuration_wall_x.py +++ b/src/lerobot/policies/wall_x/configuration_wall_x.py @@ -58,10 +58,14 @@ class WallXConfig(PreTrainedConfig): # Action prediction mode: "diffusion" or "fast" prediction_mode: str = "diffusion" - # Attention Implementation, options: "eager", "flash_attention_2", "sdpa" - # NOTE: flash-attn==2.7.4.post1 is required for flash_attention_2 implementation + # Wall-X's bidirectional action-token islands currently require eager attention. attn_implementation: str = "eager" + # Vision attention is independent from the text action-token mask. ``auto`` uses + # PyTorch's packed variable-length attention when the runtime supports it and + # otherwise falls back to the native per-chunk SDPA implementation. + vision_attn_implementation: str = "auto" + # ==================== Optimizer Presets ==================== optimizer_lr: float = 2e-5 optimizer_betas: tuple[float, float] = (0.9, 0.95) @@ -86,6 +90,18 @@ class WallXConfig(PreTrainedConfig): if self.prediction_mode not in ["diffusion", "fast"]: raise ValueError(f"prediction_mode must be 'diffusion' or 'fast', got {self.prediction_mode}") + if self.attn_implementation != "eager": + raise ValueError( + "Wall-X currently supports only attn_implementation='eager' because its " + "bidirectional action-token islands require an explicit attention mask." + ) + + if self.vision_attn_implementation not in {"auto", "sdpa", "varlen"}: + raise ValueError( + "vision_attn_implementation must be one of 'auto', 'sdpa', or 'varlen', got " + f"{self.vision_attn_implementation!r}" + ) + # Assign use_fast_tokenizer based on prediction_mode if self.prediction_mode == "fast": self.use_fast_tokenizer = True diff --git a/src/lerobot/policies/wall_x/modeling_wall_x.py b/src/lerobot/policies/wall_x/modeling_wall_x.py index bfecf3852..ee8532a4c 100644 --- a/src/lerobot/policies/wall_x/modeling_wall_x.py +++ b/src/lerobot/policies/wall_x/modeling_wall_x.py @@ -43,11 +43,14 @@ from typing import TYPE_CHECKING, Any import numpy as np import torch import torch.nn as nn -import torch.nn.functional as F -from PIL import Image +import torch.nn.functional as functional +from safetensors import SafetensorError +from safetensors.torch import load_file from torch import Tensor from torch.distributions import Beta from torch.nn import CrossEntropyLoss +from torchvision.transforms import InterpolationMode +from torchvision.transforms.v2 import functional as tv_functional from lerobot.utils.constants import ACTION, OBS_STATE from lerobot.utils.import_utils import ( @@ -74,17 +77,17 @@ if TYPE_CHECKING or _wallx_deps_available: from qwen_vl_utils.vision_process import smart_resize from torchdiffeq import odeint from transformers import AutoProcessor, BatchFeature - from transformers.cache_utils import StaticCache from transformers.models.qwen2_5_vl.modeling_qwen2_5_vl import ( + Qwen2_5_VisionTransformerPretrainedModel, Qwen2_5_VLForConditionalGeneration, ) - from transformers.utils import is_torchdynamo_compiling + from transformers.utils import cached_file, is_torchdynamo_compiling - from .qwen_model.configuration_qwen2_5_vl import Qwen2_5_VLConfig - from .qwen_model.qwen2_5_vl_moe import ( - Qwen2_5_VisionTransformerPretrainedModel, + from .qwen_model import ( Qwen2_5_VLACausalLMOutputWithPast, + Qwen2_5_VLConfig, Qwen2_5_VLMoEModel, + configure_wall_x_vision_attention, ) else: LoraConfig = None @@ -93,13 +96,14 @@ else: odeint = None AutoProcessor = None BatchFeature = None - StaticCache = None Qwen2_5_VLForConditionalGeneration = None + cached_file = None is_torchdynamo_compiling = None Qwen2_5_VLConfig = None Qwen2_5_VisionTransformerPretrainedModel = None Qwen2_5_VLACausalLMOutputWithPast = None Qwen2_5_VLMoEModel = None + configure_wall_x_vision_attention = None from .utils import ( get_wallx_normal_text, @@ -111,6 +115,75 @@ from .utils import ( logger = logging.getLogger(__name__) +def _wall_x_resize_dimensions(height: int, width: int) -> tuple[int, int, int, int]: + """Return the intermediate and final Wall-X resize dimensions as ``(H, W, H, W)``.""" + if RESOLUTION == -1: + intermediate_height, intermediate_width = height, width + elif width > height: + intermediate_width = RESOLUTION + intermediate_height = int(RESOLUTION * height / width) + else: + intermediate_height = RESOLUTION + intermediate_width = int(RESOLUTION * width / height) + + resized_height, resized_width = smart_resize( + intermediate_height, + intermediate_width, + factor=IMAGE_FACTOR, + min_pixels=MIN_PIXELS, + max_pixels=MAX_PIXELS, + ) + return intermediate_height, intermediate_width, resized_height, resized_width + + +def _resize_wall_x_image_batch(images: Tensor) -> tuple[Tensor, tuple[int, int, int, int]]: + """Quantize and resize a BCHW camera batch without leaving its current device.""" + if images.ndim != 4: + raise ValueError(f"Wall-X images must be BCHW tensors, got shape {tuple(images.shape)}") + + original_height, original_width = images.shape[-2:] + intermediate_height, intermediate_width, resized_height, resized_width = _wall_x_resize_dimensions( + original_height, original_width + ) + + if images.is_floating_point(): + # Match the previous PIL path, which quantized via `(image * 255).to(torch.uint8)`. + images = (images * 255).to(torch.uint8) + elif images.dtype != torch.uint8: + raise TypeError(f"Wall-X images must be floating point or uint8, got {images.dtype}") + + if images.shape[-2:] != (intermediate_height, intermediate_width): + images = tv_functional.resize( + images, + [intermediate_height, intermediate_width], + interpolation=InterpolationMode.BICUBIC, + antialias=True, + ) + if images.shape[-2:] != (resized_height, resized_width): + images = tv_functional.resize( + images, + [resized_height, resized_width], + interpolation=InterpolationMode.BICUBIC, + antialias=True, + ) + + return images, (original_height, original_width, resized_height, resized_width) + + +def _prepare_wall_x_image_inputs( + batch: dict[str, Any], img_keys: list[str] +) -> tuple[list[list[Tensor]], dict[str, tuple[int, int, int, int]]]: + """Resize each camera as a batch, then restore sample-major/camera-minor ordering.""" + resized_by_key: dict[str, Tensor] = {} + dimensions_by_key: dict[str, tuple[int, int, int, int]] = {} + for key in img_keys: + resized_by_key[key], dimensions_by_key[key] = _resize_wall_x_image_batch(batch[key]) + + batch_size = batch[img_keys[0]].shape[0] + image_inputs = [[resized_by_key[key][i] for key in img_keys] for i in range(batch_size)] + return image_inputs, dimensions_by_key + + class SinusoidalPosEmb(nn.Module): """Sinusoidal positional embedding for diffusion timesteps.""" @@ -246,7 +319,7 @@ class ActionHead(nn.Module): flow = flow.to(torch.float32) action_pred = self.action_proj_back(action_hidden_states) - loss = F.mse_loss(action_pred, flow, reduction="none") + loss = functional.mse_loss(action_pred, flow, reduction="none") if dof_mask is not None: dof_mask = dof_mask.reshape(-1, dof_mask.shape[-1]).to(torch.float32) @@ -254,7 +327,7 @@ class ActionHead(nn.Module): return loss - def proprioception_proj(self, proprioception, dof_mask=None, use_history=False): + def proprioception_proj(self, proprioception, dof_mask=None): """Project proprioceptive data to hidden space.""" # Ensure proper device and dtype alignment proprioception = proprioception.to(device=self.propri_proj.weight.device).to( @@ -264,10 +337,7 @@ class ActionHead(nn.Module): if dof_mask is not None: # Concatenate proprioception with DOF mask # TODO: Use variable-based dimension checking for better flexibility - if use_history: - proprioception = torch.cat([proprioception, dof_mask], dim=-1) - else: - proprioception = torch.cat([proprioception, dof_mask], dim=-1) + proprioception = torch.cat([proprioception, dof_mask], dim=-1) proprioception = proprioception.to(device=self.propri_proj.weight.device).to( dtype=self.propri_proj.weight.dtype @@ -281,7 +351,7 @@ class ActionHead(nn.Module): _Qwen2_5_VLForAction_Base = Qwen2_5_VLForConditionalGeneration if _wallx_deps_available else nn.Module -class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): +class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # noqa: N801 """ Qwen2.5 Vision-Language Mixture of Experts model for action processing. @@ -305,6 +375,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): config=None, action_tokenizer_path=None, attn_implementation: str = "eager", + vision_attn_implementation: str = "auto", cache_dir: str | PathLike | None = None, force_download: bool = False, local_files_only: bool = False, @@ -321,11 +392,14 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): config_path (str, optional): Configuration file path, if None will look for qwen25_config.json in pretrained_model_path action_tokenizer_path (str, optional): Action tokenizer path, if None will load from default config attn_implementation (str, optional): Attention implementation, if None will load from default config + vision_attn_implementation (str, optional): Vision attention backend. ``auto`` uses packed + variable-length attention when supported and otherwise falls back to SDPA. **kwargs: Additional arguments Returns: Qwen2_5_VLMoEForAction: Loaded model instance """ + Qwen2_5_VLMoEModel._require_eager_attention(attn_implementation) if config is None: config = cls.config_class.from_pretrained( pretrained_name_or_path, @@ -339,7 +413,15 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): ) if attn_implementation is not None: config._attn_implementation = attn_implementation - processor = AutoProcessor.from_pretrained(pretrained_name_or_path, use_fast=True) + processor = AutoProcessor.from_pretrained( + pretrained_name_or_path, + cache_dir=cache_dir, + force_download=force_download, + local_files_only=local_files_only, + token=token, + revision=revision, + use_fast=True, + ) if action_tokenizer_path is not None: action_tokenizer = AutoProcessor.from_pretrained(action_tokenizer_path, trust_remote_code=True) processor.action_processor = action_tokenizer @@ -351,41 +433,41 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): config.text_config.pad_token_id = processor.tokenizer.pad_token_id # Initialize model with configuration and processor - model = cls(config, processor=processor, action_tokenizer=action_tokenizer, **kwargs) + model = cls( + config, + processor=processor, + action_tokenizer=action_tokenizer, + vision_attn_implementation=vision_attn_implementation, + **kwargs, + ) # Resize token embeddings to match processor tokenizer vocabulary size model.resize_token_embeddings(len(processor.tokenizer)) - # Try to load the model.safetensors file - print(f"Loading model from: {pretrained_name_or_path}") + logger.info("Loading Wall-X model from %s", pretrained_name_or_path) try: - from transformers.utils import cached_file - - # Try safetensors first resolved_file = cached_file( pretrained_name_or_path, "model.safetensors", - cache_dir=kwargs.get("cache_dir"), - force_download=kwargs.get("force_download", False), + cache_dir=cache_dir, + force_download=force_download, resume_download=kwargs.get("resume_download"), proxies=kwargs.get("proxies"), - token=kwargs.get("token"), - revision=kwargs.get("revision"), - local_files_only=kwargs.get("local_files_only", False), + token=token, + revision=revision, + local_files_only=local_files_only, ) - from safetensors.torch import load_file - sd = load_file(resolved_file) - print("✓ Loaded state dict from model.safetensors") - except Exception as e: - print(f"Could not load state dict from remote files: {e}") - print("Returning model without loading pretrained weights") - return model + except (OSError, SafetensorError) as error: + raise OSError( + f"Failed to load pretrained Wall-X weights from {pretrained_name_or_path!r}" + ) from error + logger.info("Loaded Wall-X state dict from model.safetensors") state_dict = {} # filter normalizer statistic params del_keys = [] - for key in sd.keys(): + for key in sd: if "action_preprocessor.normalizer" in key: del_keys.append(key) for key in del_keys: @@ -404,6 +486,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): action_tokenizer=None, action_mapper=None, flow_loss_weight=1.0, + vision_attn_implementation: str = "auto", ): """ Initialize the Qwen2.5 VLMoE model for action processing. @@ -416,10 +499,16 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): action_mapper: Action mapping utility flow_loss_weight (float): Weight for flow loss computation """ + Qwen2_5_VLMoEModel._require_eager_attention(config._attn_implementation) + config._attn_implementation = "eager" + # Text needs eager attention for action-token islands. Vision has no such + # constraint, so keep its portable native fallback on SDPA. + config.vision_config._attn_implementation = "sdpa" super().__init__(config) # Initialize vision transformer and language model components self.visual = Qwen2_5_VisionTransformerPretrainedModel._from_config(config.vision_config) + configure_wall_x_vision_attention(self.visual, vision_attn_implementation) self.model = Qwen2_5_VLMoEModel(config) self.vocab_size = config.vocab_size self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False) @@ -457,7 +546,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): params_to_keep_float32 = [] - for name, param in self.named_parameters(): + for name, _param in self.named_parameters(): if "input_layernorm" in name or "post_attention_layernorm" in name or "model.norm" in name: params_to_keep_float32.append(name) if "action_preprocessor" in name: @@ -491,7 +580,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): "action_token_id": action_token_id, } - def add_lora(self, r=8, lora_alpha=32, target_modules=["q_proj", "v_proj"], lora_dropout=0.1): + def add_lora(self, r=8, lora_alpha=32, target_modules=None, lora_dropout=0.1): """ Add LoRA (Low-Rank Adaptation) adapters to the model. @@ -501,6 +590,9 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): target_modules (list): List of module names to apply LoRA to lora_dropout (float): Dropout probability for LoRA layers """ + if target_modules is None: + target_modules = ["q_proj", "v_proj"] + config = LoraConfig( r=r, lora_alpha=lora_alpha, @@ -795,6 +887,9 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): ) return_dict = return_dict if return_dict is not None else self.config.use_return_dict + if rope_deltas is not None: + self.rope_deltas = rope_deltas + # Calculate RoPE position IDs if not provided # Note: Cannot calculate rope deltas with 4D attention mask. TODO: Fix this limitation if position_ids is None and (attention_mask is None or attention_mask.ndim == 2): @@ -833,7 +928,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # Process image embeddings if pixel_values is not None: pixel_values = pixel_values.type(self.visual.dtype) - image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw) + image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw).pooler_output mask = input_ids == self.config.image_token_id mask_unsqueezed = mask.unsqueeze(-1) mask_expanded = mask_unsqueezed.expand_as(inputs_embeds) @@ -845,7 +940,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # Process video embeddings if pixel_values_videos is not None: pixel_values_videos = pixel_values_videos.type(self.visual.dtype) - video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw) + video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw).pooler_output n_video_tokens = (input_ids == self.config.video_token_id).sum().item() n_video_features = video_embeds.shape[0] @@ -869,7 +964,6 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): proprioception = self.action_preprocessor.proprioception_proj( proprioception, agent_pos_mask, - use_history=proprioception.shape[1] > 1, ) mask = input_ids == self.action_token_id_set["propri_token_id"] mask_unsqueezed = mask.unsqueeze(-1) @@ -919,6 +1013,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): output_attentions=output_attentions, output_hidden_states=output_hidden_states, return_dict=return_dict, + cache_position=cache_position, ) hidden_states = outputs[0] @@ -1107,7 +1202,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # Process image embeddings if pixel_values is not None: pixel_values = pixel_values.type(self.visual.dtype) - image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw) + image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw).pooler_output n_image_tokens = (input_ids == self.config.image_token_id).sum().item() n_image_features = image_embeds.shape[0] @@ -1128,7 +1223,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # Process video embeddings if pixel_values_videos is not None: pixel_values_videos = pixel_values_videos.type(self.visual.dtype) - video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw) + video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw).pooler_output n_video_tokens = (input_ids == self.config.video_token_id).sum().item() n_video_features = video_embeds.shape[0] @@ -1153,7 +1248,6 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): proprio_embed = self.action_preprocessor.proprioception_proj( proprioception, agent_pos_mask, - use_history=proprioception.shape[1] > 1, ) proprioception_mask = input_ids == self.action_token_id_set["propri_token_id"] proprio_embed = proprio_embed.to(torch.bfloat16) @@ -1202,25 +1296,37 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): # Split input sequence for text and fast modes (not needed for diffusion) if predict_mode == "text" or predict_mode == "fast": - # Look for generation prompt tokens: <|im_start|>assistant + generation_prompt = "<|im_start|>assistant\n" generation_prompt_ids = torch.tensor( - [151644, 77091], device=input_ids.device, dtype=input_ids.dtype - ) - matches = (input_ids[0, :-1] == generation_prompt_ids[0]) & ( - input_ids[0, 1:] == generation_prompt_ids[1] + self.processor.tokenizer.encode(generation_prompt, add_special_tokens=False), + device=input_ids.device, + dtype=input_ids.dtype, ) + prompt_length = generation_prompt_ids.numel() + if prompt_length == 0: + raise ValueError(f"Tokenizer produced no tokens for generation prompt {generation_prompt!r}") + if input_ids.shape[1] < prompt_length: + matches = torch.empty(0, device=input_ids.device, dtype=torch.bool) + else: + matches = ( + input_ids[0] + .unfold(dimension=0, size=prompt_length, step=1) + .eq(generation_prompt_ids) + .all(dim=-1) + ) if matches.any(): split_pos = torch.nonzero(matches, as_tuple=True)[0][0].item() + prompt_end = split_pos + prompt_length # Extract ground truth output tokens (including newline) - gt_output_ids = input_ids[:, split_pos + 3 :] + gt_output_ids = input_ids[:, prompt_end:] # Remove output part from input, keeping prompt - input_ids = input_ids[:, : split_pos + 3] - inputs_embeds = inputs_embeds[:, : split_pos + 3, :] + input_ids = input_ids[:, :prompt_end] + inputs_embeds = inputs_embeds[:, :prompt_end, :] if attention_mask is not None: - attention_mask = attention_mask[:, : split_pos + 3] + attention_mask = attention_mask[:, :prompt_end] if labels is not None: - labels = labels[:, split_pos + 3 :] + labels = labels[:, prompt_end:] else: raise ValueError( "input_ids does not contain the generation prompt tokens <|im_start|>assistant" @@ -1255,7 +1361,7 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): use_cache=True, pad_token_id=self.processor.tokenizer.pad_token_id, temperature=(1.0 if not re_generate else 0.7), # Higher temperature for regeneration - do_sample=(False if not re_generate else True), # Enable sampling for regeneration + do_sample=re_generate, # Enable sampling for regeneration ) # Decode generated and ground truth text @@ -1524,27 +1630,6 @@ class Qwen2_5_VLMoEForAction(_Qwen2_5_VLForAction_Base): else: model_inputs = {"input_ids": input_ids, "inputs_embeds": None} - # Prepare 4D causal attention mask for static cache - if isinstance(past_key_values, StaticCache) and attention_mask.ndim == 2: - if model_inputs["inputs_embeds"] is not None: - batch_size, sequence_length, _ = inputs_embeds.shape - device = inputs_embeds.device - else: - batch_size, sequence_length = input_ids.shape - device = input_ids.device - - attention_mask = self.model._prepare_4d_causal_attention_mask_with_cache_position( - attention_mask, - sequence_length=sequence_length, - target_length=past_key_values.get_max_cache_shape(), - dtype=self.lm_head.weight.dtype, - device=device, - cache_position=cache_position, - batch_size=batch_size, - config=self.config, - past_key_values=past_key_values, - ) - # Assemble all model inputs for generation model_inputs.update( { @@ -1749,6 +1834,7 @@ class WallXPolicy(PreTrainedPolicy): pretrained_name_or_path=config.pretrained_name_or_path, action_tokenizer_path=config.action_tokenizer_path, attn_implementation=config.attn_implementation, + vision_attn_implementation=config.vision_attn_implementation, ) self.model.to(config.device) self.model.to_bfloat16_for_selected_params() @@ -1768,6 +1854,8 @@ class WallXPolicy(PreTrainedPolicy): def preprocess_inputs( self, batch: dict[str, Any], + *, + compute_position_ids: bool = False, ) -> BatchFeature: """ Convert a batch of LeRobot dataset items to Wall-X model input format. @@ -1789,50 +1877,21 @@ class WallXPolicy(PreTrainedPolicy): # Get batch size from state tensor batch_size = batch[OBS_STATE].shape[0] - # ==================== PROCESS ALL SAMPLES ==================== - all_image_inputs = [] - all_texts = [] - # Find image keys in batch img_keys = [key for key in self.config.image_features if key in batch] + if not img_keys: + raise ValueError("Wall-X requires at least one image feature in each batch") + + # Resize one camera batch at a time on the tensors' current device. Reassembling + # sample-major keeps image_grid_thw aligned with each sample's image placeholders. + all_image_inputs, dimensions_by_key = _prepare_wall_x_image_inputs(batch, img_keys) + all_texts = [] + + # Preserve the existing grounding behavior for multi-camera inputs: the old camera + # loop left these values set to the final configured camera's dimensions. + orig_height, orig_width, resized_height, resized_width = dimensions_by_key[img_keys[-1]] for i in range(batch_size): - # Vision preprocessing per sample - processed_frames = [] - orig_height, orig_width = None, None - resized_height, resized_width = None, None - - for key in img_keys: - current_obs = batch[key][i].clone() # (C, H, W) - if current_obs.dim() == 3: - current_obs = current_obs.permute(1, 2, 0) # (H, W, C) - - img_pil = Image.fromarray((current_obs * 255).to(torch.uint8).cpu().numpy()) - orig_width, orig_height = img_pil.size - - target_size = RESOLUTION - if target_size != -1: - if orig_width > orig_height: - new_width = target_size - new_height = int(target_size * orig_height / orig_width) - else: - new_height = target_size - new_width = int(target_size * orig_width / orig_height) - img_pil = img_pil.resize((new_width, new_height)) - - current_width, current_height = img_pil.size - resized_height, resized_width = smart_resize( - current_height, - current_width, - factor=IMAGE_FACTOR, - min_pixels=MIN_PIXELS, - max_pixels=MAX_PIXELS, - ) - resized_img = img_pil.resize((resized_width, resized_height)) - processed_frames.append(resized_img) - - all_image_inputs.append(processed_frames) - # Text preprocessing task_text = batch["task"][i] if isinstance(batch["task"], list) else batch["task"] instruction_info = {"instruction": task_text} @@ -1859,8 +1918,8 @@ class WallXPolicy(PreTrainedPolicy): agent_pos_mask = (~torch.isnan(agent_pos)).float() agent_pos = agent_pos.nan_to_num(nan=0.0) - if agent_pos.shape[-1] != 20: - pad_size = 20 - agent_pos.shape[-1] + if agent_pos.shape[-1] < self.config.max_state_dim: + pad_size = self.config.max_state_dim - agent_pos.shape[-1] agent_pos = torch.cat( [ agent_pos, @@ -1880,6 +1939,10 @@ class WallXPolicy(PreTrainedPolicy): ], dim=-1, ) + elif agent_pos.shape[-1] > self.config.max_state_dim: + raise ValueError( + f"State dimension {agent_pos.shape[-1]} exceeds max_state_dim {self.config.max_state_dim}" + ) # ==================== PROCESS ACTIONS ==================== action = batch.get(ACTION) # (batch_size, chunk_size, action_dim) @@ -1889,8 +1952,8 @@ class WallXPolicy(PreTrainedPolicy): dof_mask = (~torch.isnan(action)).float() action = action.nan_to_num(nan=0.0) - if action.shape[-1] != 20: - pad_size = 20 - action.shape[-1] + if action.shape[-1] < self.config.max_action_dim: + pad_size = self.config.max_action_dim - action.shape[-1] action = torch.cat( [action, torch.zeros(action.shape[0], action.shape[1], pad_size, device=action.device)], dim=-1, @@ -1902,6 +1965,10 @@ class WallXPolicy(PreTrainedPolicy): ], dim=-1, ) + elif action.shape[-1] > self.config.max_action_dim: + raise ValueError( + f"Action dimension {action.shape[-1]} exceeds max_action_dim {self.config.max_action_dim}" + ) else: action_dim = self.config.output_features[ACTION].shape[0] dof_mask = torch.cat( @@ -1910,7 +1977,10 @@ class WallXPolicy(PreTrainedPolicy): batch_size, self.config.chunk_size, action_dim, device=batch[OBS_STATE].device ), torch.zeros( - batch_size, self.config.chunk_size, 20 - action_dim, device=batch[OBS_STATE].device + batch_size, + self.config.chunk_size, + self.config.max_action_dim - action_dim, + device=batch[OBS_STATE].device, ), ], dim=-1, @@ -1930,12 +2000,26 @@ class WallXPolicy(PreTrainedPolicy): text=all_texts, images=all_image_inputs, videos=None, + device=batch[OBS_STATE].device, padding=True, truncation=True, return_tensors="pt", max_length=TOKENIZER_MAX_LENGTH, ) + if compute_position_ids: + # Qwen's RoPE indexing uses Python list/scalar conversions. Run it while the + # tokenizer and grid metadata are still on CPU, then move the compact result. + position_ids, rope_deltas = self.model.get_rope_index( + inputs.input_ids, + inputs.get("image_grid_thw"), + inputs.get("video_grid_thw"), + inputs.get("second_per_grid_ts"), + inputs.attention_mask, + ) + inputs["position_ids"] = position_ids + inputs["rope_deltas"] = rope_deltas + # ==================== ADDITIONAL INPUTS ==================== action_token_id = self.model.processor.tokenizer.convert_tokens_to_ids("<|action|>") moe_token_types = inputs.input_ids == action_token_id @@ -1952,7 +2036,7 @@ class WallXPolicy(PreTrainedPolicy): ) # Move all tensors to the correct device - device = self.config.device + device = batch[OBS_STATE].device for key, value in inputs.items(): if isinstance(value, torch.Tensor): inputs[key] = value.to(device) @@ -1972,9 +2056,7 @@ class WallXPolicy(PreTrainedPolicy): Returns: tuple: (loss, loss_dict) """ - batch = self.preprocess_inputs( - batch, - ) + batch = self.preprocess_inputs(batch, compute_position_ids=True) # Call the underlying model's forward with mode="train" outputs = self.model(**batch, mode="train") @@ -1982,19 +2064,19 @@ class WallXPolicy(PreTrainedPolicy): # Extract losses from output loss = outputs.loss loss_dict = { - "loss": loss.item() if loss is not None else 0.0, + "loss": loss.detach() if loss is not None else 0.0, } if outputs.flow_loss is not None: - loss_dict["flow_loss"] = outputs.flow_loss.item() + loss_dict["flow_loss"] = outputs.flow_loss.detach() if outputs.cross_entropy_loss is not None: - loss_dict["cross_entropy_loss"] = outputs.cross_entropy_loss.item() + loss_dict["cross_entropy_loss"] = outputs.cross_entropy_loss.detach() # Add channel losses if available if outputs.channel_loss_dict is not None: for key, value in outputs.channel_loss_dict.items(): if isinstance(value, torch.Tensor): - loss_dict[f"channel_{key}"] = value.item() + loss_dict[f"channel_{key}"] = value.detach() return loss, loss_dict diff --git a/src/lerobot/policies/wall_x/processor_wall_x.py b/src/lerobot/policies/wall_x/processor_wall_x.py index 069cef5d6..e8cec6ccb 100644 --- a/src/lerobot/policies/wall_x/processor_wall_x.py +++ b/src/lerobot/policies/wall_x/processor_wall_x.py @@ -20,19 +20,13 @@ import torch from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( - AddBatchDimensionProcessorStep, ComplementaryDataProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStepRegistry, - RenameObservationsProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) -from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME from .configuration_wall_x import WallXConfig @@ -65,37 +59,22 @@ def make_wall_x_pre_post_processors( A tuple containing the configured pre-processor and post-processor pipelines """ + steps = make_default_policy_processor_steps(config, dataset_stats) + input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), + steps.rename_observations, + steps.add_batch_dim, WallXTaskProcessor(), # Process task description - NormalizerProcessorStep( - features={**config.input_features, **config.output_features}, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - DeviceProcessorStep(device=config.device), + steps.normalize, + steps.to_device, ] output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats - ), - DeviceProcessorStep(device="cpu"), + steps.unnormalize, + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) @ProcessorStepRegistry.register(name="wall_x_task_processor") diff --git a/src/lerobot/policies/wall_x/qwen_model/__init__.py b/src/lerobot/policies/wall_x/qwen_model/__init__.py new file mode 100644 index 000000000..6269c247d --- /dev/null +++ b/src/lerobot/policies/wall_x/qwen_model/__init__.py @@ -0,0 +1,45 @@ +#!/usr/bin/env python + +# Copyright 2025 HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from .configuration_qwen2_5_vl import ( + Qwen2_5_VLConfig, + Qwen2_5_VLTextConfig, + Qwen2_5_VLVisionConfig, +) +from .qwen2_5_vl_moe import ( + BlockSparseMLP, + Qwen2_5_VLACausalLMOutputWithPast, + Qwen2_5_VLDecoderLayer_with_MoE, + Qwen2_5_VLMoEModel, + SparseMoeBlock, +) +from .vision_attention import ( + WallXVisionAttention, + configure_wall_x_vision_attention, +) + +__all__ = [ + "BlockSparseMLP", + "Qwen2_5_VLACausalLMOutputWithPast", + "Qwen2_5_VLConfig", + "Qwen2_5_VLDecoderLayer_with_MoE", + "Qwen2_5_VLMoEModel", + "Qwen2_5_VLTextConfig", + "Qwen2_5_VLVisionConfig", + "SparseMoeBlock", + "WallXVisionAttention", + "configure_wall_x_vision_attention", +] diff --git a/src/lerobot/policies/wall_x/qwen_model/configuration_qwen2_5_vl.py b/src/lerobot/policies/wall_x/qwen_model/configuration_qwen2_5_vl.py index 19874b6ff..6f59e752f 100644 --- a/src/lerobot/policies/wall_x/qwen_model/configuration_qwen2_5_vl.py +++ b/src/lerobot/policies/wall_x/qwen_model/configuration_qwen2_5_vl.py @@ -1,250 +1,114 @@ -from transformers.configuration_utils import PretrainedConfig -from transformers.modeling_rope_utils import rope_config_validation +#!/usr/bin/env python + +# Copyright 2025 HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Wall-X configuration extensions for the native Transformers Qwen2.5-VL config.""" + +from dataclasses import dataclass +from typing import TYPE_CHECKING + +from huggingface_hub.dataclasses import strict + +from lerobot.utils.import_utils import _transformers_available + +if TYPE_CHECKING or _transformers_available: + from transformers.models.qwen2_5_vl.configuration_qwen2_5_vl import ( + Qwen2_5_VLConfig as TransformersQwen2_5_VLConfig, + Qwen2_5_VLTextConfig as TransformersQwen2_5_VLTextConfig, + Qwen2_5_VLVisionConfig, + ) +else: + + @dataclass + class _TransformersConfigFallback: + """Import-safe stand-in used only when Transformers is unavailable.""" + + TransformersQwen2_5_VLConfig = _TransformersConfigFallback + TransformersQwen2_5_VLTextConfig = _TransformersConfigFallback + Qwen2_5_VLVisionConfig = None + +# Wall-X checkpoints pre0.6.0 use the legacy, flat Qwen2.5-VL config layout. The native +# ``Qwen2_5_VLConfig`` accepts that layout and moves text-model fields into its +# ``text_config`` sub-config, so only the Wall-X-specific MoE fields need to be +# declared here. +_LEGACY_TEXT_ATTRIBUTES = { + "attention_dropout", + "attention_moe", + "dim_inputs", + "dof_config", + "experts", + "hidden_act", + "hidden_size", + "initializer_range", + "intermediate_size", + "layer_types", + "max_position_embeddings", + "max_window_layers", + "mlp_moe", + "noise_scheduler", + "num_attention_heads", + "num_experts", + "num_hidden_layers", + "num_key_value_heads", + "pad_token_id", + "rms_norm_eps", + "sliding_window", + "use_cache", + "use_sliding_window", + "vocab_size", +} -class Qwen2_5_VLVisionConfig(PretrainedConfig): - model_type = "qwen2_5_vl" - base_config_key = "vision_config" +@strict +class Qwen2_5_VLTextConfig(TransformersQwen2_5_VLTextConfig): # noqa: N801 + """Native Qwen2.5-VL text config plus Wall-X's hard-routed MoE settings.""" - def __init__( - self, - depth=32, - hidden_size=3584, - hidden_act="silu", - intermediate_size=3420, - num_heads=16, - in_channels=3, - patch_size=14, - spatial_merge_size=2, - temporal_patch_size=2, - tokens_per_second=4, - window_size=112, - out_hidden_size=3584, - fullatt_block_indexes=[7, 15, 23, 31], - initializer_range=0.02, - **kwargs, - ): - super().__init__(**kwargs) + num_experts: int = 4 + experts: list[dict] | None = None + dof_config: dict | None = None + noise_scheduler: dict | None = None + dim_inputs: tuple[int, ...] | list[int] = (1536, 1536) + attention_moe: bool = False + mlp_moe: bool = False - self.depth = depth - self.hidden_size = hidden_size - self.hidden_act = hidden_act - self.intermediate_size = intermediate_size - self.num_heads = num_heads - self.in_channels = in_channels - self.patch_size = patch_size - self.spatial_merge_size = spatial_merge_size - self.temporal_patch_size = temporal_patch_size - self.tokens_per_second = tokens_per_second - self.window_size = window_size - self.fullatt_block_indexes = fullatt_block_indexes - self.out_hidden_size = out_hidden_size - self.initializer_range = initializer_range + def __post_init__(self, **kwargs): + self.dim_inputs = tuple(self.dim_inputs) + super().__post_init__(**kwargs) -class Qwen2_5_VLConfig(PretrainedConfig): - r""" - This is the configuration class to store the configuration of a [`Qwen2_5_VLModel`]. It is used to instantiate a - Qwen2-VL model according to the specified arguments, defining the model architecture. Instantiating a configuration - with the defaults will yield a similar configuration to that of - Qwen2-VL-7B-Instruct [Qwen/Qwen2-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct). +@strict +class Qwen2_5_VLConfig(TransformersQwen2_5_VLConfig): # noqa: N801 + """Native composite Qwen2.5-VL config with a Wall-X text sub-config. - Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the - documentation from [`PretrainedConfig`] for more information. + The native composite loader supports both current nested configs and the + flat layout used by existing ``wall-oss-flow`` checkpoints. + """ - - Args: - vocab_size (`int`, *optional*, defaults to 152064): - Vocabulary size of the Qwen2_5_VL model. Defines the number of different tokens that can be represented by the - `inputs_ids` passed when calling [`Qwen2_5_VLModel`] - hidden_size (`int`, *optional*, defaults to 8192): - Dimension of the hidden representations. - intermediate_size (`int`, *optional*, defaults to 29568): - Dimension of the MLP representations. - num_hidden_layers (`int`, *optional*, defaults to 80): - Number of hidden layers in the Transformer encoder. - num_attention_heads (`int`, *optional*, defaults to 64): - Number of attention heads for each attention layer in the Transformer encoder. - num_key_value_heads (`int`, *optional*, defaults to 8): - This is the number of key_value heads that should be used to implement Grouped Query Attention. If - `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if - `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When - converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed - by meanpooling all the original heads within that group. For more details checkout [this - paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to `32`. - hidden_act (`str` or `function`, *optional*, defaults to `"silu"`): - The non-linear activation function (function or string) in the decoder. - max_position_embeddings (`int`, *optional*, defaults to 32768): - The maximum sequence length that this model might ever be used with. - initializer_range (`float`, *optional*, defaults to 0.02): - The standard deviation of the truncated_normal_initializer for initializing all weight matrices. - rms_norm_eps (`float`, *optional*, defaults to 1e-05): - The epsilon used by the rms normalization layers. - use_cache (`bool`, *optional*, defaults to `True`): - Whether or not the model should return the last key/values attentions (not used by all models). Only - relevant if `config.is_decoder=True`. - tie_word_embeddings (`bool`, *optional*, defaults to `False`): - Whether the model's input and output word embeddings should be tied. - rope_theta (`float`, *optional*, defaults to 1000000.0): - The base period of the RoPE embeddings. - use_sliding_window (`bool`, *optional*, defaults to `False`): - Whether to use sliding window attention. - sliding_window (`int`, *optional*, defaults to 4096): - Sliding window attention (SWA) window size. If not specified, will default to `4096`. - max_window_layers (`int`, *optional*, defaults to 80): - The number of layers that use SWA (Sliding Window Attention). The bottom layers use SWA while the top use full attention. - attention_dropout (`float`, *optional*, defaults to 0.0): - The dropout ratio for the attention probabilities. - vision_config (`Dict`, *optional*): - The config for the visual encoder initialization. - rope_scaling (`Dict`, *optional*): - Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type - and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value - accordingly. - Expected contents: - `rope_type` (`str`): - The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope', - 'llama3'], with 'default' being the original RoPE implementation. - `factor` (`float`, *optional*): - Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In - most scaling types, a `factor` of x will enable the model to handle sequences of length x * - original maximum pre-trained length. - `original_max_position_embeddings` (`int`, *optional*): - Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during - pretraining. - `attention_factor` (`float`, *optional*): - Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention - computation. If unspecified, it defaults to value recommended by the implementation, using the - `factor` field to infer the suggested value. - `beta_fast` (`float`, *optional*): - Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear - ramp function. If unspecified, it defaults to 32. - `beta_slow` (`float`, *optional*): - Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear - ramp function. If unspecified, it defaults to 1. - `short_factor` (`List[float]`, *optional*): - Only used with 'longrope'. The scaling factor to be applied to short contexts (< - `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden - size divided by the number of attention heads divided by 2 - `long_factor` (`List[float]`, *optional*): - Only used with 'longrope'. The scaling factor to be applied to long contexts (< - `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden - size divided by the number of attention heads divided by 2 - `low_freq_factor` (`float`, *optional*): - Only used with 'llama3'. Scaling factor applied to low frequency components of the RoPE - `high_freq_factor` (`float`, *optional*): - Only used with 'llama3'. Scaling factor applied to high frequency components of the RoPE - - ```python - >>> from transformers import Qwen2_5_VLForConditionalGeneration, Qwen2_5_VLConfig - - >>> # Initializing a Qwen2_5_VL style configuration - >>> configuration = Qwen2_5_VLConfig() - - >>> # Initializing a model from the Qwen2-VL-7B style configuration - >>> model = Qwen2_5_VLForConditionalGeneration(configuration) - - >>> # Accessing the model configuration - >>> configuration = model.config - ```""" - - model_type = "qwen2_5_vl" - sub_configs = {"vision_config": Qwen2_5_VLVisionConfig} - keys_to_ignore_at_inference = ["past_key_values"] - # Default tensor parallel plan for base model `Qwen2_5_VL` - base_model_tp_plan = { - "layers.*.self_attn.q_proj": "colwise", - "layers.*.self_attn.k_proj": "colwise", - "layers.*.self_attn.v_proj": "colwise", - "layers.*.self_attn.o_proj": "rowwise", - "layers.*.mlp.gate_proj": "colwise", - "layers.*.mlp.up_proj": "colwise", - "layers.*.mlp.down_proj": "rowwise", - } - base_model_pp_plan = { - "embed_tokens": (["input_ids"], ["inputs_embeds"]), - "layers": (["hidden_states", "attention_mask"], ["hidden_states"]), - "norm": (["hidden_states"], ["hidden_states"]), + sub_configs = { + "vision_config": Qwen2_5_VLVisionConfig, + "text_config": Qwen2_5_VLTextConfig, } - def __init__( - self, - vocab_size=152064, - hidden_size=8192, - intermediate_size=29568, - num_hidden_layers=80, - num_attention_heads=64, - num_key_value_heads=8, - hidden_act="silu", - max_position_embeddings=32768, - initializer_range=0.02, - rms_norm_eps=1e-05, - use_cache=True, - tie_word_embeddings=False, - rope_theta=1000000.0, - use_sliding_window=False, - sliding_window=4096, - max_window_layers=80, - attention_dropout=0.0, - vision_config=None, - rope_scaling=None, - num_experts=4, - experts=None, - dof_config=None, - noise_scheduler=None, - dim_inputs=(1536, 1536), - attention_moe=False, - mlp_moe=False, - **kwargs, - ): - if isinstance(vision_config, dict): - self.vision_config = self.sub_configs["vision_config"](**vision_config) - elif vision_config is None: - self.vision_config = self.sub_configs["vision_config"]() + def __getattr__(self, name): + """Keep legacy direct access to fields now owned by ``text_config``. - self.vocab_size = vocab_size - self.max_position_embeddings = max_position_embeddings - self.hidden_size = hidden_size - self.intermediate_size = intermediate_size - self.num_hidden_layers = num_hidden_layers - self.num_attention_heads = num_attention_heads - self.use_sliding_window = use_sliding_window - self.sliding_window = sliding_window - self.max_window_layers = max_window_layers - self.layer_types = ["dense"] * num_hidden_layers - - # for backward compatibility - if num_key_value_heads is None: - num_key_value_heads = num_attention_heads - - self.num_key_value_heads = num_key_value_heads - self.hidden_act = hidden_act - self.initializer_range = initializer_range - self.rms_norm_eps = rms_norm_eps - self.use_cache = use_cache - self.rope_theta = rope_theta - self.attention_dropout = attention_dropout - self.rope_scaling = rope_scaling - - self.num_experts = num_experts - self.experts = experts - self.dof_config = dof_config - self.noise_scheduler = noise_scheduler - self.dim_inputs = tuple(dim_inputs) - self.attention_moe = attention_moe - self.mlp_moe = mlp_moe - - if self.rope_scaling is not None and "type" in self.rope_scaling: - if self.rope_scaling["type"] == "mrope": - self.rope_scaling["type"] = "default" - self.rope_scaling["rope_type"] = self.rope_scaling["type"] - rope_config_validation(self, ignore_keys={"mrope_section"}) - - super().__init__(tie_word_embeddings=tie_word_embeddings, **kwargs) - - @property - def text_config(self): - return self - - -__all__ = ["Qwen2_5_VLConfig"] + Wall-X historically used a flat config and accesses fields such as + ``hidden_size`` and ``num_experts`` directly. Forwarding unknown + attributes preserves that API without duplicating the native config. + """ + text_config = self.__dict__.get("text_config") + if name in _LEGACY_TEXT_ATTRIBUTES and text_config is not None and hasattr(text_config, name): + return getattr(text_config, name) + raise AttributeError(f"{type(self).__name__!s} has no attribute {name!r}") diff --git a/src/lerobot/policies/wall_x/qwen_model/qwen2_5_vl_moe.py b/src/lerobot/policies/wall_x/qwen_model/qwen2_5_vl_moe.py index fd60f976f..0aa1cb704 100644 --- a/src/lerobot/policies/wall_x/qwen_model/qwen2_5_vl_moe.py +++ b/src/lerobot/policies/wall_x/qwen_model/qwen2_5_vl_moe.py @@ -1,2219 +1,77 @@ -import math +#!/usr/bin/env python + +# Copyright 2025 HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Wall-X Mixture-of-Experts additions on top of the native transformers Qwen2.5-VL model. + +It is rebased on the native ``transformers.models.qwen2_5_vl`` classes and only keeps what Wall-X genuinely adds: + +- ``BlockSparseMLP`` / ``SparseMoeBlock``: hard-routed (token-type indexed) expert MLPs. +- ``Qwen2_5_VLDecoderLayer_with_MoE``: native decoder layer whose MLP is replaced by the sparse MoE + block and whose forward casts activations to the parameter dtypes (Wall-X keeps the layernorms in + float32 while the projections run in bfloat16, see ``to_bfloat16_for_selected_params``). +- ``Qwen2_5_VLMoEModel``: native text model with MoE decoder layers and a ``moe_token_types``-aware + causal-mask override (tokens of type 1 — the action tokens — attend to each other bidirectionally, + everything else stays causal). +- ``Qwen2_5_VLACausalLMOutputWithPast``: output dataclass with the extra Wall-X loss fields. +""" + +from __future__ import annotations + from dataclasses import dataclass -from typing import Any +from typing import TYPE_CHECKING import torch import torch.nn as nn -import torch.nn.functional as F -from torch.nn import CrossEntropyLoss -from transformers import AutoConfig -from transformers.activations import ACT2FN -from transformers.cache_utils import ( - Cache, - DynamicCache, - StaticCache, -) -from transformers.generation import GenerationMixin -from transformers.modeling_attn_mask_utils import AttentionMaskConverter -from transformers.modeling_outputs import BaseModelOutputWithPast, ModelOutput -from transformers.modeling_rope_utils import ROPE_INIT_FUNCTIONS -from transformers.modeling_utils import PreTrainedModel -from transformers.utils import ( - add_start_docstrings, - add_start_docstrings_to_model_forward, - is_flash_attn_2_available, - is_flash_attn_greater_or_equal, - is_torchdynamo_compiling, - logging, - replace_return_docstrings, -) -from .configuration_qwen2_5_vl import Qwen2_5_VLConfig, Qwen2_5_VLVisionConfig +from lerobot.utils.import_utils import _transformers_available - -# TODO(Steven): SlidingWindowCache was removed in transformers v5. Define a placeholder so isinstance checks -# always return False (which is the correct behavior when no sliding window cache is in use). -class _SlidingWindowCachePlaceholder: - pass - - -SlidingWindowCache = _SlidingWindowCachePlaceholder - -if is_flash_attn_2_available(): - from flash_attn import flash_attn_func, flash_attn_varlen_func - from flash_attn.layers.rotary import apply_rotary_emb +if TYPE_CHECKING or _transformers_available: + from transformers.activations import ACT2FN + from transformers.cache_utils import Cache, DynamicCache + from transformers.masking_utils import create_causal_mask, create_sliding_window_causal_mask + from transformers.modeling_outputs import BaseModelOutputWithPast, ModelOutput + from transformers.models.qwen2_5_vl.modeling_qwen2_5_vl import ( + Qwen2_5_VLDecoderLayer, + Qwen2_5_VLTextModel, + ) + from transformers.utils.generic import merge_with_config_defaults + from transformers.utils.output_capturing import capture_outputs else: - flash_attn_varlen_func = None - apply_rotary_emb = None - flash_attn_func = None + ACT2FN = None + Cache = None + DynamicCache = None + create_causal_mask = None + create_sliding_window_causal_mask = None + BaseModelOutputWithPast = object + ModelOutput = object + Qwen2_5_VLDecoderLayer = nn.Module + Qwen2_5_VLTextModel = nn.Module + def merge_with_config_defaults(func): + return func -if is_flash_attn_2_available(): - pass -else: - flash_attn_varlen_func = None + def capture_outputs(func): + return func -logger = logging.get_logger(__name__) - -_CONFIG_FOR_DOC = "Qwen2_5_VLConfig" - - -class Qwen2_5_VLMLP(nn.Module): - def __init__(self, config, bias: bool = False): - super().__init__() - self.hidden_size = config.hidden_size - self.intermediate_size = config.intermediate_size - self.gate_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=bias) - self.up_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=bias) - self.down_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=bias) - self.act_fn = ACT2FN[config.hidden_act] - - def forward(self, hidden_state): - return self.down_proj(self.act_fn(self.gate_proj(hidden_state)) * self.up_proj(hidden_state)) - - -class Qwen2_5_VisionPatchEmbed(nn.Module): - def __init__( - self, - patch_size: int = 14, - temporal_patch_size: int = 2, - in_channels: int = 3, - embed_dim: int = 1152, - ) -> None: - super().__init__() - self.patch_size = patch_size - self.temporal_patch_size = temporal_patch_size - self.in_channels = in_channels - self.embed_dim = embed_dim - - kernel_size = [temporal_patch_size, patch_size, patch_size] - self.proj = nn.Conv3d( - in_channels, - embed_dim, - kernel_size=kernel_size, - stride=kernel_size, - bias=False, - ) - - def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: - target_dtype = self.proj.weight.dtype - hidden_states = hidden_states.view( - -1, - self.in_channels, - self.temporal_patch_size, - self.patch_size, - self.patch_size, - ) - hidden_states = self.proj(hidden_states.to(dtype=target_dtype)).view(-1, self.embed_dim) - return hidden_states - - -class Qwen2_5_VisionRotaryEmbedding(nn.Module): - def __init__(self, dim: int, theta: float = 10000.0) -> None: - super().__init__() - inv_freq = 1.0 / (theta ** (torch.arange(0, dim, 2, dtype=torch.float) / dim)) - self.register_buffer("inv_freq", inv_freq, persistent=False) - - def forward(self, seqlen: int) -> torch.Tensor: - seq = torch.arange(seqlen, device=self.inv_freq.device, dtype=self.inv_freq.dtype) - freqs = torch.outer(seq, self.inv_freq) - return freqs - - -class Qwen2RMSNorm(nn.Module): - def __init__(self, hidden_size, eps=1e-6): - """ - Qwen2RMSNorm is equivalent to T5LayerNorm - """ - super().__init__() - self.weight = nn.Parameter(torch.ones(hidden_size)) - self.variance_epsilon = eps - - def forward(self, hidden_states): - input_dtype = hidden_states.dtype - hidden_states = hidden_states.to(torch.float32) - variance = hidden_states.pow(2).mean(-1, keepdim=True) - hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon) - return self.weight * hidden_states.to(input_dtype) - - def extra_repr(self): - return f"{tuple(self.weight.shape)}, eps={self.variance_epsilon}" - - -class Qwen2_5_VLPatchMerger(nn.Module): - def __init__(self, dim: int, context_dim: int, spatial_merge_size: int = 2) -> None: - super().__init__() - self.hidden_size = context_dim * (spatial_merge_size**2) - self.ln_q = Qwen2RMSNorm(context_dim, eps=1e-6) - self.mlp = nn.Sequential( - nn.Linear(self.hidden_size, self.hidden_size), - nn.GELU(), - nn.Linear(self.hidden_size, dim), - ) - - def forward(self, x: torch.Tensor) -> torch.Tensor: - x = self.mlp(self.ln_q(x).view(-1, self.hidden_size)) - return x - - -def apply_rotary_pos_emb_flashatt( - q: torch.Tensor, k: torch.Tensor, cos: torch.Tensor, sin: torch.Tensor -) -> tuple[torch.Tensor, torch.Tensor]: - cos = cos.chunk(2, dim=-1)[0].contiguous() - sin = sin.chunk(2, dim=-1)[0].contiguous() - q_embed = apply_rotary_emb(q.float(), cos.float(), sin.float()).type_as(q) - k_embed = apply_rotary_emb(k.float(), cos.float(), sin.float()).type_as(k) - return q_embed, k_embed - - -class Qwen2_5_VLVisionFlashAttention2(nn.Module): - def __init__(self, dim: int, num_heads: int = 16) -> None: - super().__init__() - self.num_heads = num_heads - self.qkv = nn.Linear(dim, dim * 3, bias=True) - self.proj = nn.Linear(dim, dim) - - def forward( - self, - hidden_states: torch.Tensor, - cu_seqlens: torch.Tensor, - max_seqlen: int | None = None, - rotary_pos_emb: torch.Tensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, - ) -> torch.Tensor: - seq_length = hidden_states.shape[0] - q, k, v = ( - self.qkv(hidden_states).reshape(seq_length, 3, self.num_heads, -1).permute(1, 0, 2, 3).unbind(0) - ) - if position_embeddings is None: - logger.warning_once( - "The attention layers in this model are transitioning from computing the RoPE embeddings internally " - "through `rotary_pos_emb` (2D tensor of RoPE theta values), to using externally computed " - "`position_embeddings` (Tuple of tensors, containing cos and sin). In v4.54 `rotary_pos_emb` will be " - "removed and `position_embeddings` will be mandatory." - ) - emb = torch.cat((rotary_pos_emb, rotary_pos_emb), dim=-1) - cos = emb.cos().float() - sin = emb.sin().float() - else: - cos, sin = position_embeddings - q, k = apply_rotary_pos_emb_flashatt(q.unsqueeze(0), k.unsqueeze(0), cos, sin) - q = q.squeeze(0) - k = k.squeeze(0) - - if max_seqlen is None: - max_seqlen = (cu_seqlens[1:] - cu_seqlens[:-1]).max().item() - attn_output = flash_attn_varlen_func(q, k, v, cu_seqlens, cu_seqlens, max_seqlen, max_seqlen).reshape( - seq_length, -1 - ) - attn_output = self.proj(attn_output) - return attn_output - - -def rotate_half(x): - """Rotates half the hidden dims of the input.""" - x1 = x[..., : x.shape[-1] // 2] - x2 = x[..., x.shape[-1] // 2 :] - return torch.cat((-x2, x1), dim=-1) - - -def apply_rotary_pos_emb_vision( - q: torch.Tensor, k: torch.Tensor, cos: torch.Tensor, sin: torch.Tensor -) -> tuple[torch.Tensor, torch.Tensor]: - orig_q_dtype = q.dtype - orig_k_dtype = k.dtype - q, k = q.float(), k.float() - cos, sin = cos.unsqueeze(-2), sin.unsqueeze(-2) - q_embed = (q * cos) + (rotate_half(q) * sin) - k_embed = (k * cos) + (rotate_half(k) * sin) - q_embed = q_embed.to(orig_q_dtype) - k_embed = k_embed.to(orig_k_dtype) - return q_embed, k_embed - - -class Qwen2_5_VLVisionAttention(nn.Module): - def __init__(self, dim: int, num_heads: int = 16) -> None: - super().__init__() - self.num_heads = num_heads - self.head_dim = dim // num_heads - self.qkv = nn.Linear(dim, dim * 3, bias=True) - self.proj = nn.Linear(dim, dim) - - def forward( - self, - hidden_states: torch.Tensor, - cu_seqlens: torch.Tensor, - max_seqlen: int | None = None, - rotary_pos_emb: torch.Tensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, - ) -> torch.Tensor: - seq_length = hidden_states.shape[0] - q, k, v = ( - self.qkv(hidden_states).reshape(seq_length, 3, self.num_heads, -1).permute(1, 0, 2, 3).unbind(0) - ) - if position_embeddings is None: - logger.warning_once( - "The attention layers in this model are transitioning from computing the RoPE embeddings internally " - "through `rotary_pos_emb` (2D tensor of RoPE theta values), to using externally computed " - "`position_embeddings` (Tuple of tensors, containing cos and sin). In v4.54 `rotary_pos_emb` will be " - "removed and `position_embeddings` will be mandatory." - ) - emb = torch.cat((rotary_pos_emb, rotary_pos_emb), dim=-1) - cos = emb.cos().float() - sin = emb.sin().float() - else: - cos, sin = position_embeddings - q, k = apply_rotary_pos_emb_vision(q, k, cos, sin) - - attention_mask = torch.full( - [1, seq_length, seq_length], - torch.finfo(q.dtype).min, - device=q.device, - dtype=q.dtype, - ) - for i in range(1, len(cu_seqlens)): - attention_mask[ - ..., - cu_seqlens[i - 1] : cu_seqlens[i], - cu_seqlens[i - 1] : cu_seqlens[i], - ] = 0 - - q = q.transpose(0, 1) - k = k.transpose(0, 1) - v = v.transpose(0, 1) - attn_weights = torch.matmul(q, k.transpose(1, 2)) / math.sqrt(self.head_dim) - attn_weights = attn_weights + attention_mask - attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(q.dtype) - attn_output = torch.matmul(attn_weights, v) - attn_output = attn_output.transpose(0, 1) - attn_output = attn_output.reshape(seq_length, -1) - attn_output = self.proj(attn_output) - return attn_output - - -class Qwen2_5_VLVisionSdpaAttention(nn.Module): - def __init__(self, dim: int, num_heads: int = 16) -> None: - super().__init__() - self.num_heads = num_heads - self.qkv = nn.Linear(dim, dim * 3, bias=True) - self.proj = nn.Linear(dim, dim) - - def forward( - self, - hidden_states: torch.Tensor, - cu_seqlens: torch.Tensor, - max_seqlen: int | None = None, - rotary_pos_emb: torch.Tensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, - ) -> torch.Tensor: - seq_length = hidden_states.shape[0] - q, k, v = ( - self.qkv(hidden_states).reshape(seq_length, 3, self.num_heads, -1).permute(1, 0, 2, 3).unbind(0) - ) - if position_embeddings is None: - logger.warning_once( - "The attention layers in this model are transitioning from computing the RoPE embeddings internally " - "through `rotary_pos_emb` (2D tensor of RoPE theta values), to using externally computed " - "`position_embeddings` (Tuple of tensors, containing cos and sin). In v4.54 `rotary_pos_emb` will be " - "removed and `position_embeddings` will be mandatory." - ) - emb = torch.cat((rotary_pos_emb, rotary_pos_emb), dim=-1) - cos = emb.cos().float() - sin = emb.sin().float() - else: - cos, sin = position_embeddings - q, k = apply_rotary_pos_emb_vision(q, k, cos, sin) - - attention_mask = torch.zeros([1, seq_length, seq_length], device=q.device, dtype=torch.bool) - for i in range(1, len(cu_seqlens)): - attention_mask[ - ..., - cu_seqlens[i - 1] : cu_seqlens[i], - cu_seqlens[i - 1] : cu_seqlens[i], - ] = True - q = q.transpose(0, 1) - k = k.transpose(0, 1) - v = v.transpose(0, 1) - attn_output = F.scaled_dot_product_attention(q, k, v, attention_mask, dropout_p=0.0) - attn_output = attn_output.transpose(0, 1) - attn_output = attn_output.reshape(seq_length, -1) - attn_output = self.proj(attn_output) - return attn_output - - -QWEN2_5_VL_VISION_ATTENTION_CLASSES = { - "eager": Qwen2_5_VLVisionAttention, - "flash_attention_2": Qwen2_5_VLVisionFlashAttention2, - "sdpa": Qwen2_5_VLVisionSdpaAttention, -} - - -class Qwen2_5_VLVisionBlock(nn.Module): - def __init__(self, config, attn_implementation: str = "sdpa") -> None: - super().__init__() - self.norm1 = Qwen2RMSNorm(config.hidden_size, eps=1e-6) - self.norm2 = Qwen2RMSNorm(config.hidden_size, eps=1e-6) - self.attn = QWEN2_5_VL_VISION_ATTENTION_CLASSES[attn_implementation]( - config.hidden_size, num_heads=config.num_heads - ) - self.mlp = Qwen2_5_VLMLP(config, bias=True) - - def forward( - self, - hidden_states: torch.Tensor, - cu_seqlens: torch.Tensor, - max_seqlen: int | None = None, - rotary_pos_emb: torch.Tensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, - ) -> torch.Tensor: - hidden_states = hidden_states + self.attn( - self.norm1(hidden_states), - cu_seqlens=cu_seqlens, - max_seqlen=max_seqlen, - rotary_pos_emb=rotary_pos_emb, - position_embeddings=position_embeddings, - ) - hidden_states = hidden_states + self.mlp(self.norm2(hidden_states)) - return hidden_states - - -Qwen2_5_VL_START_DOCSTRING = r""" - This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the - library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads - etc.) - - This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. - Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage - and behavior. - - Parameters: - config ([`Qwen2_5_VLConfig`]): - Model configuration class with all the parameters of the model. Initializing with a config file does not - load the weights associated with the model, only the configuration. Check out the - [`~PreTrainedModel.from_pretrained`] method to load the model weights. -""" - - -@add_start_docstrings( - "The bare Qwen2_5_VL Model outputting raw hidden-states without any specific head on top.", - Qwen2_5_VL_START_DOCSTRING, -) -class Qwen2_5_VLPreTrainedModel(PreTrainedModel): - config_class = Qwen2_5_VLConfig - base_model_prefix = "model" - supports_gradient_checkpointing = True - _no_split_modules = ["Qwen2_5_VLDecoderLayer", "Qwen2_5_VLVisionBlock"] - _skip_keys_device_placement = "past_key_values" - _supports_flash_attn_2 = True - _supports_sdpa = True - _supports_cache_class = True - _supports_static_cache = ( - False # TODO (joao): fix. torch.compile failing probably due to `cache_positions` - ) - - def _init_weights(self, module): - std = self.config.initializer_range - if isinstance(module, (nn.Linear, nn.Conv3d)): - module.weight.data.normal_(mean=0.0, std=std) - if module.bias is not None: - module.bias.data.zero_() - elif isinstance(module, nn.Embedding): - module.weight.data.normal_(mean=0.0, std=std) - if module.padding_idx is not None: - module.weight.data[module.padding_idx].zero_() - - -class Qwen2_5_VisionTransformerPretrainedModel(Qwen2_5_VLPreTrainedModel): - config_class = Qwen2_5_VLVisionConfig - _no_split_modules = ["Qwen2_5_VLVisionBlock"] - - def __init__(self, config, *inputs, **kwargs) -> None: - super().__init__(config, *inputs, **kwargs) - self.spatial_merge_size = config.spatial_merge_size - self.patch_size = config.patch_size - self.fullatt_block_indexes = config.fullatt_block_indexes - self.window_size = config.window_size - self.spatial_merge_unit = self.spatial_merge_size * self.spatial_merge_size - - self.patch_embed = Qwen2_5_VisionPatchEmbed( - patch_size=config.patch_size, - temporal_patch_size=config.temporal_patch_size, - in_channels=config.in_channels, - embed_dim=config.hidden_size, - ) - - head_dim = config.hidden_size // config.num_heads - self.rotary_pos_emb = Qwen2_5_VisionRotaryEmbedding(head_dim // 2) - - self.blocks = nn.ModuleList( - [Qwen2_5_VLVisionBlock(config, config._attn_implementation) for _ in range(config.depth)] - ) - self.merger = Qwen2_5_VLPatchMerger( - dim=config.out_hidden_size, - context_dim=config.hidden_size, - spatial_merge_size=config.spatial_merge_size, - ) - self.gradient_checkpointing = False - - def rot_pos_emb(self, grid_thw): - pos_ids = [] - for t, h, w in grid_thw: - hpos_ids = torch.arange(h).unsqueeze(1).expand(-1, w) - hpos_ids = hpos_ids.reshape( - h // self.spatial_merge_size, - self.spatial_merge_size, - w // self.spatial_merge_size, - self.spatial_merge_size, - ) - hpos_ids = hpos_ids.permute(0, 2, 1, 3) - hpos_ids = hpos_ids.flatten() - - wpos_ids = torch.arange(w).unsqueeze(0).expand(h, -1) - wpos_ids = wpos_ids.reshape( - h // self.spatial_merge_size, - self.spatial_merge_size, - w // self.spatial_merge_size, - self.spatial_merge_size, - ) - wpos_ids = wpos_ids.permute(0, 2, 1, 3) - wpos_ids = wpos_ids.flatten() - pos_ids.append(torch.stack([hpos_ids, wpos_ids], dim=-1).repeat(t, 1)) - pos_ids = torch.cat(pos_ids, dim=0) - max_grid_size = grid_thw[:, 1:].max() - rotary_pos_emb_full = self.rotary_pos_emb(max_grid_size) - rotary_pos_emb = rotary_pos_emb_full[pos_ids].flatten(1) - return rotary_pos_emb - - def get_window_index(self, grid_thw): - window_index: list = [] - cu_window_seqlens: list = [0] - window_index_id = 0 - vit_merger_window_size = self.window_size // self.spatial_merge_size // self.patch_size - - for grid_t, grid_h, grid_w in grid_thw: - llm_grid_h, llm_grid_w = ( - grid_h // self.spatial_merge_size, - grid_w // self.spatial_merge_size, - ) - index = torch.arange(grid_t * llm_grid_h * llm_grid_w).reshape(grid_t, llm_grid_h, llm_grid_w) - pad_h = vit_merger_window_size - llm_grid_h % vit_merger_window_size - pad_w = vit_merger_window_size - llm_grid_w % vit_merger_window_size - num_windows_h = (llm_grid_h + pad_h) // vit_merger_window_size - num_windows_w = (llm_grid_w + pad_w) // vit_merger_window_size - index_padded = F.pad(index, (0, pad_w, 0, pad_h), "constant", -100) - index_padded = index_padded.reshape( - grid_t, - num_windows_h, - vit_merger_window_size, - num_windows_w, - vit_merger_window_size, - ) - index_padded = index_padded.permute(0, 1, 3, 2, 4).reshape( - grid_t, - num_windows_h * num_windows_w, - vit_merger_window_size, - vit_merger_window_size, - ) - seqlens = (index_padded != -100).sum([2, 3]).reshape(-1) - index_padded = index_padded.reshape(-1) - index_new = index_padded[index_padded != -100] - window_index.append(index_new + window_index_id) - cu_seqlens_tmp = seqlens.cumsum(0) * self.spatial_merge_unit + cu_window_seqlens[-1] - cu_window_seqlens.extend(cu_seqlens_tmp.tolist()) - window_index_id += (grid_t * llm_grid_h * llm_grid_w).item() - window_index = torch.cat(window_index, dim=0) - - return window_index, cu_window_seqlens - - def forward(self, hidden_states: torch.Tensor, grid_thw: torch.Tensor) -> torch.Tensor: - """ - Args: - hidden_states (`torch.Tensor` of shape `(seq_len, hidden_size)`): - The final hidden states of the model. - grid_thw (`torch.Tensor` of shape `(num_images_or_videos, 3)`): - The temporal, height and width of feature shape of each image in LLM. - - Returns: - `torch.Tensor`: hidden_states. - """ - hidden_states = self.patch_embed(hidden_states) - rotary_pos_emb = self.rot_pos_emb(grid_thw) - window_index, cu_window_seqlens = self.get_window_index(grid_thw) - window_index = window_index.to(hidden_states.device) - cu_window_seqlens = torch.tensor( - cu_window_seqlens, - device=hidden_states.device, - dtype=grid_thw.dtype if torch.jit.is_tracing() else torch.int32, - ) - cu_window_seqlens = torch.unique_consecutive(cu_window_seqlens) - - seq_len, _ = hidden_states.size() - hidden_states = hidden_states.reshape(seq_len // self.spatial_merge_unit, self.spatial_merge_unit, -1) - hidden_states = hidden_states[window_index, :, :] - hidden_states = hidden_states.reshape(seq_len, -1) - rotary_pos_emb = rotary_pos_emb.reshape( - seq_len // self.spatial_merge_unit, self.spatial_merge_unit, -1 - ) - rotary_pos_emb = rotary_pos_emb[window_index, :, :] - rotary_pos_emb = rotary_pos_emb.reshape(seq_len, -1) - emb = torch.cat((rotary_pos_emb, rotary_pos_emb), dim=-1) - position_embeddings = (emb.cos(), emb.sin()) - - cu_seqlens = torch.repeat_interleave(grid_thw[:, 1] * grid_thw[:, 2], grid_thw[:, 0]).cumsum( - dim=0, - # Select dtype based on the following factors: - # - FA2 requires that cu_seqlens_q must have dtype int32 - # - torch.onnx.export requires that cu_seqlens_q must have same dtype as grid_thw - # See https://github.com/huggingface/transformers/pull/34852 for more information - dtype=grid_thw.dtype if torch.jit.is_tracing() else torch.int32, - ) - cu_seqlens = F.pad(cu_seqlens, (1, 0), value=0) - max_seqlen_full = (cu_seqlens[1:] - cu_seqlens[:-1]).max().item() - max_seqlen_window = (cu_window_seqlens[1:] - cu_window_seqlens[:-1]).max().item() - - for layer_num, blk in enumerate(self.blocks): - if layer_num in self.fullatt_block_indexes: - cu_seqlens_now = cu_seqlens - max_seqlen_now = max_seqlen_full - else: - cu_seqlens_now = cu_window_seqlens - max_seqlen_now = max_seqlen_window - if self.gradient_checkpointing and self.training: - hidden_states = self._gradient_checkpointing_func( - blk.__call__, - hidden_states, - cu_seqlens_now, - None, - position_embeddings, - ) - else: - hidden_states = blk( - hidden_states, - cu_seqlens=cu_seqlens_now, - max_seqlen=max_seqlen_now, - position_embeddings=position_embeddings, - ) - - hidden_states = self.merger(hidden_states) - reverse_indices = torch.argsort(window_index) - hidden_states = hidden_states[reverse_indices, :] - - return hidden_states - - -def _compute_default_rope_parameters_qwen2_5_vl(config, device=None): - """ - compute default rope parameters for Qwen2_5_VL - """ - base = config.text_config.rope_parameters["rope_theta"] - dim = config.hidden_size // config.num_attention_heads - inv_freq = 1.0 / ( - base ** (torch.arange(0, dim, 2, dtype=torch.int64).to(device=device, dtype=torch.float) / dim) - ) - return inv_freq, 1.0 - - -class Qwen2_5_VLRotaryEmbedding(nn.Module): - def __init__(self, config: Qwen2_5_VLConfig, device=None): - super().__init__() - # BC: "rope_type" was originally "type" - if hasattr(config, "rope_scaling") and config.rope_scaling is not None: - self.rope_type = config.rope_scaling.get("rope_type", config.rope_scaling.get("type")) - elif hasattr(config, "rope_parameters") and config.rope_parameters is not None: - self.rope_type = config.rope_parameters.get("rope_type", "default") - else: - self.rope_type = "default" - self.max_seq_len_cached = config.max_position_embeddings - self.original_max_seq_len = config.max_position_embeddings - - self.config = config - - if self.rope_type == "default": - self.rope_init_fn = _compute_default_rope_parameters_qwen2_5_vl - self.rope_kwargs = {} - else: - rope_type_key = "linear" if self.rope_type == "linear" else self.rope_type - self.rope_init_fn = ROPE_INIT_FUNCTIONS[rope_type_key] - self.rope_kwargs = {} - - inv_freq, self.attention_scaling = self.rope_init_fn(self.config, device) - self.register_buffer("inv_freq", inv_freq, persistent=False) - self.original_inv_freq = self.inv_freq - - def _dynamic_frequency_update(self, position_ids, device): - """ - dynamic RoPE layers should recompute `inv_freq` in the following situations: - 1 - growing beyond the cached sequence length (allow scaling) - 2 - the current sequence length is in the original scale (avoid losing precision with small sequences) - """ - seq_len = torch.max(position_ids) + 1 - if seq_len > self.max_seq_len_cached: # growth - inv_freq, self.attention_scaling = self.rope_init_fn( - self.config, device, seq_len=seq_len, **self.rope_kwargs - ) - self.register_buffer( - "inv_freq", inv_freq, persistent=False - ) # TODO joao: may break with compilation - self.max_seq_len_cached = seq_len - - if ( - seq_len < self.original_max_seq_len and self.max_seq_len_cached > self.original_max_seq_len - ): # reset - self.register_buffer("inv_freq", self.original_inv_freq, persistent=False) - self.max_seq_len_cached = self.original_max_seq_len - - @torch.no_grad() - def forward(self, x, position_ids): - if "dynamic" in self.rope_type: - self._dynamic_frequency_update(position_ids, device=x.device) - - # Core RoPE block. In contrast to other models, Qwen2_5_VL has different position ids for thw grids - # So we expand the inv_freq to shape (3, ...) - inv_freq_expanded = self.inv_freq[None, None, :, None].float().expand(3, position_ids.shape[1], -1, 1) - position_ids_expanded = position_ids[:, :, None, :].float() # shape (3, bs, 1, positions) - # Force float32 (see https://github.com/huggingface/transformers/pull/29285) - device_type = x.device.type - device_type = device_type if isinstance(device_type, str) and device_type != "mps" else "cpu" - with torch.autocast(device_type=device_type, enabled=False): - freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(2, 3) - emb = torch.cat((freqs, freqs), dim=-1) - cos = emb.cos() - sin = emb.sin() - - # Advanced RoPE types (e.g. yarn) apply a post-processing scaling factor, equivalent to scaling attention - cos = cos * self.attention_scaling - sin = sin * self.attention_scaling - - return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype) - - -class Qwen2MLP(nn.Module): - def __init__(self, config): - super().__init__() - self.config = config - self.hidden_size = config.hidden_size - self.intermediate_size = config.intermediate_size - self.gate_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False) - self.up_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False) - self.down_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=False) - self.act_fn = ACT2FN[config.hidden_act] - - def forward(self, x): - down_proj = self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x)) - return down_proj - - -def apply_multimodal_rotary_pos_emb(q, k, cos, sin, mrope_section, unsqueeze_dim=1): - """Applies Rotary Position Embedding with Multimodal Sections to the query and key tensors (https://qwenlm.github.io/blog/qwen2-vl/). - - Explanation: - Multimodal 3D rotary position embedding is an extension to 1D rotary position embedding. The input embedding - sequence contains vision (images / videos) embedding and text embedding or just contains text embedding. For - vision embedding part, we apply rotary position embedding on temporal, height and width dimension separately. - Here we split the channel dimension to 3 chunks for the temporal, height and width rotary position embedding. - For text embedding part, we just apply 1D rotary position embedding. The three rotary position index (temporal, - height and width) of text embedding is always the same, so the text embedding rotary position embedding has no - difference with modern LLMs. - - Args: - q (`torch.Tensor`): The query tensor. - k (`torch.Tensor`): The key tensor. - cos (`torch.Tensor`): The cosine part of the rotary embedding. - sin (`torch.Tensor`): The sine part of the rotary embedding. - position_ids (`torch.Tensor`): - The position indices of the tokens corresponding to the query and key tensors. For example, this can be - used to pass offsetted position ids when working with a KV-cache. - mrope_section(`List(int)`): - Multimodal rope section is for channel dimension of temporal, height and width in rope calculation. - unsqueeze_dim (`int`, *optional*, defaults to 1): - The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and - sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note - that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and - k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes - cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have - the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2. - Returns: - `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding. - """ - mrope_section = mrope_section * 2 - cos = torch.cat([m[i % 3] for i, m in enumerate(cos.split(mrope_section, dim=-1))], dim=-1).unsqueeze( - unsqueeze_dim - ) - sin = torch.cat([m[i % 3] for i, m in enumerate(sin.split(mrope_section, dim=-1))], dim=-1).unsqueeze( - unsqueeze_dim - ) - - q_embed = (q * cos) + (rotate_half(q) * sin) - k_embed = (k * cos) + (rotate_half(k) * sin) - return q_embed, k_embed - - -def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor: - """ - This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, - num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim) - """ - batch, num_key_value_heads, slen, head_dim = hidden_states.shape - if n_rep == 1: - return hidden_states - hidden_states = hidden_states[:, :, None, :, :].expand(batch, num_key_value_heads, n_rep, slen, head_dim) - return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim) - - -class Qwen2_5_VLAttention(nn.Module): - """ - Multi-headed attention from 'Attention Is All You Need' paper. Modified to use sliding window attention: Longformer - and "Generating Long Sequences with Sparse Transformers". - """ - - def __init__(self, config: Qwen2_5_VLConfig, layer_idx: int | None = None): - super().__init__() - self.config = config - self.layer_idx = layer_idx - if layer_idx is None: - logger.warning_once( - f"Instantiating {self.__class__.__name__} without passing `layer_idx` is not recommended and will " - "to errors during the forward call, if caching is used. Please make sure to provide a `layer_idx` " - "when creating this class." - ) - - self.hidden_size = config.hidden_size - self.num_heads = config.num_attention_heads - self.head_dim = self.hidden_size // self.num_heads - self.num_key_value_heads = config.num_key_value_heads - self.num_key_value_groups = self.num_heads // self.num_key_value_heads - self.is_causal = True - self.attention_dropout = config.attention_dropout - self.rope_scaling = config.rope_scaling - - if (self.head_dim * self.num_heads) != self.hidden_size: - raise ValueError( - f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}" - f" and `num_heads`: {self.num_heads})." - ) - self.q_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=True) - self.k_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=True) - self.v_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=True) - self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=False) - - self.rotary_emb = Qwen2_5_VLRotaryEmbedding(config=config) - - def forward( - self, - hidden_states: torch.Tensor, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_value: Cache | None = None, - output_attentions: bool = False, - use_cache: bool = False, - cache_position: torch.LongTensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] - | None = None, # necessary, but kept here for BC - ) -> tuple[torch.Tensor, torch.Tensor | None, tuple[torch.Tensor] | None]: - bsz, q_len, _ = hidden_states.size() - - query_states = self.q_proj(hidden_states) - key_states = self.k_proj(hidden_states) - value_states = self.v_proj(hidden_states) - - query_states = query_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - key_states = key_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - value_states = value_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - - cos, sin = position_embeddings - query_states, key_states = apply_multimodal_rotary_pos_emb( - query_states, key_states, cos, sin, self.rope_scaling["mrope_section"] - ) - - if past_key_value is not None: - cache_kwargs = { - "sin": sin, - "cos": cos, - "cache_position": cache_position, - } # Specific to RoPE models - key_states, value_states = past_key_value.update( - key_states, value_states, self.layer_idx, cache_kwargs - ) - - # repeat k/v heads if n_kv_heads < n_heads - key_states = repeat_kv(key_states, self.num_key_value_groups) - value_states = repeat_kv(value_states, self.num_key_value_groups) - - attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim) - - if attention_mask is not None: # no matter the length, we just slice it - causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] - attn_weights = attn_weights + causal_mask - - # Fix precision issues in Qwen2-VL float16 inference - # Replace inf values with zeros in attention weights to prevent NaN propagation - if query_states.dtype == torch.float16: - attn_weights = torch.where( - torch.isinf(attn_weights), torch.zeros_like(attn_weights), attn_weights - ) - - # upcast attention to fp32 - attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query_states.dtype) - attn_weights = nn.functional.dropout(attn_weights, p=self.attention_dropout, training=self.training) - attn_output = torch.matmul(attn_weights, value_states) - - if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim): - raise ValueError( - f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is" - f" {attn_output.size()}" - ) - - attn_output = attn_output.transpose(1, 2).contiguous() - attn_output = attn_output.reshape(bsz, q_len, -1) - - attn_output = self.o_proj(attn_output) - - if not output_attentions: - attn_weights = None - - return attn_output, attn_weights, past_key_value - - -class Qwen2_5_VLFlashAttention2(Qwen2_5_VLAttention): - """ - Qwen2_5_VL flash attention module, following Qwen2_5_VL attention module. This module inherits from `Qwen2_5_VLAttention` - as the weights of the module stays untouched. The only required change would be on the forward pass - where it needs to correctly call the public API of flash attention and deal with padding tokens - in case the input contains any of them. Additionally, for sliding window attention, we apply SWA only to the bottom - config.max_window_layers layers. - """ - - def __init__(self, *args, **kwargs): - super().__init__(*args, **kwargs) - - # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1. - # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignment, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0. - # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left). - self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal("2.1.0") - - def forward( - self, - hidden_states: torch.Tensor, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_value: Cache | None = None, - output_attentions: bool = False, - use_cache: bool = False, - cache_position: torch.LongTensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] - | None = None, # necessary, but kept here for BC - ): - bsz, q_len, _ = hidden_states.size() - query_states = self.q_proj(hidden_states) - key_states = self.k_proj(hidden_states) - value_states = self.v_proj(hidden_states) - - query_states = query_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - key_states = key_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - value_states = value_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - - # Because the input can be padded, the absolute sequence length depends on the max position id. - cos, sin = position_embeddings - query_states, key_states = apply_multimodal_rotary_pos_emb( - query_states, key_states, cos, sin, self.rope_scaling["mrope_section"] - ) - if past_key_value is not None: - cache_kwargs = { - "sin": sin, - "cos": cos, - "cache_position": cache_position, - } # Specific to RoPE models - key_states, value_states = past_key_value.update( - key_states, value_states, self.layer_idx, cache_kwargs - ) - - # repeat k/v heads if n_kv_heads < n_heads - # key_states = repeat_kv(key_states, self.num_key_value_groups) - # value_states = repeat_kv(value_states, self.num_key_value_groups) - dropout_rate = 0.0 if not self.training else self.attention_dropout - - # In PEFT, usually we cast the layer norms in float32 for training stability reasons - # therefore the input hidden states gets silently casted in float32. Hence, we need - # cast them back in float16 just to be sure everything works as expected. - input_dtype = query_states.dtype - if input_dtype == torch.float32: - if torch.is_autocast_enabled(): - target_dtype = torch.get_autocast_dtype(query_states.device.type) - # Handle the case where the model is quantized - elif hasattr(self.config, "_pre_quantization_dtype"): - target_dtype = self.config._pre_quantization_dtype - else: - target_dtype = self.q_proj.weight.dtype - - logger.warning_once( - f"The input hidden states seems to be silently casted in float32, this might be related to" - f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in" - f" {target_dtype}." - ) - - query_states = query_states.to(target_dtype) - key_states = key_states.to(target_dtype) - value_states = value_states.to(target_dtype) - - # Reashape to the expected shape for Flash Attention - query_states = query_states.transpose(1, 2) - key_states = key_states.transpose(1, 2) - value_states = value_states.transpose(1, 2) - - attn_output = flash_attn_func( - query_states, - key_states, - value_states, - dropout_rate, - softmax_scale=None, - causal=self.is_causal, - ) - - attn_output = attn_output.reshape(bsz, q_len, self.hidden_size).contiguous() - attn_output = self.o_proj(attn_output) - - if not output_attentions: - attn_weights = None - - return attn_output, attn_weights, past_key_value - - -class Qwen2_5_VLSdpaAttention(Qwen2_5_VLAttention): - """ - Qwen2 attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from - `Qwen2Attention` as the weights of the module stays untouched. The only changes are on the forward pass to adapt to - SDPA API. - """ - - # Adapted from Qwen2Attention.forward - def forward( - self, - hidden_states: torch.Tensor, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_value: Cache | None = None, - output_attentions: bool = False, - use_cache: bool = False, - cache_position: torch.LongTensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] - | None = None, # necessary, but kept here for BC - ) -> tuple[torch.Tensor, torch.Tensor | None, tuple[torch.Tensor] | None]: - if output_attentions: - # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented. - logger.warning_once( - "Qwen2_5_VLModel is using Qwen2_5_VLSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, " - 'but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.' - ) - return super().forward( - hidden_states=hidden_states, - attention_mask=attention_mask, - position_ids=position_ids, - past_key_value=past_key_value, - output_attentions=output_attentions, - use_cache=use_cache, - cache_position=cache_position, - position_embeddings=position_embeddings, - ) - - bsz, q_len, _ = hidden_states.size() - - query_states = self.q_proj(hidden_states) - key_states = self.k_proj(hidden_states) - value_states = self.v_proj(hidden_states) - - query_states = query_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - key_states = key_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - value_states = value_states.view(bsz, q_len, -1, self.head_dim).transpose(1, 2) - - cos, sin = position_embeddings - query_states, key_states = apply_multimodal_rotary_pos_emb( - query_states, key_states, cos, sin, self.rope_scaling["mrope_section"] - ) - - if past_key_value is not None: - cache_kwargs = { - "sin": sin, - "cos": cos, - "cache_position": cache_position, - } # Specific to RoPE models - key_states, value_states = past_key_value.update( - key_states, value_states, self.layer_idx, cache_kwargs - ) - - key_states = repeat_kv(key_states, self.num_key_value_groups) - value_states = repeat_kv(value_states, self.num_key_value_groups) - - causal_mask = attention_mask - if attention_mask is not None: # no matter the length, we just slice it - causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] - - # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask, - # Reference: https://github.com/pytorch/pytorch/issues/112577. - if query_states.device.type == "cuda" and attention_mask is not None: - query_states = query_states.contiguous() - key_states = key_states.contiguous() - value_states = value_states.contiguous() - - # We dispatch to SDPA's Flash Attention or Efficient kernels via this `is_causal` if statement instead of an inline conditional assignment - # in SDPA to support both torch.compile's dynamic shapes and full graph options. An inline conditional prevents dynamic shapes from compiling. - # The q_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case q_len == 1. - is_causal = True if causal_mask is None and q_len > 1 else False - - attn_output = torch.nn.functional.scaled_dot_product_attention( - query_states, - key_states, - value_states, - attn_mask=causal_mask, - dropout_p=self.attention_dropout if self.training else 0.0, - is_causal=is_causal, - ) - - attn_output = attn_output.transpose(1, 2).contiguous() - attn_output = attn_output.view(bsz, q_len, self.hidden_size) - - attn_output = self.o_proj(attn_output) - - return attn_output, None, past_key_value - - -QWEN2_5_VL_ATTENTION_CLASSES = { - "eager": Qwen2_5_VLAttention, - "flash_attention_2": Qwen2_5_VLFlashAttention2, - "sdpa": Qwen2_5_VLSdpaAttention, -} - - -class Qwen2_5_VLDecoderLayer(nn.Module): - def __init__(self, config: Qwen2_5_VLConfig, layer_idx: int): - super().__init__() - self.hidden_size = config.hidden_size - - if config.use_sliding_window and config._attn_implementation != "flash_attention_2": - logger.warning_once( - f"Sliding Window Attention is enabled but not implemented for `{config._attn_implementation}`; " - "unexpected results may be encountered." - ) - self.self_attn = QWEN2_5_VL_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx) - - self.mlp = Qwen2MLP(config) - self.input_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - self.post_attention_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - - def forward( - self, - hidden_states: torch.Tensor, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_value: tuple[torch.Tensor] | None = None, - output_attentions: bool | None = False, - use_cache: bool | None = False, - cache_position: torch.LongTensor | None = None, - position_embeddings: tuple[torch.Tensor, torch.Tensor] - | None = None, # necessary, but kept here for BC - **kwargs, - ) -> tuple[torch.FloatTensor, tuple[torch.FloatTensor, torch.FloatTensor] | None]: - """ - Args: - hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` - attention_mask (`torch.FloatTensor`, *optional*): attention mask of size - `(batch, sequence_length)` where padding elements are indicated by 0. - output_attentions (`bool`, *optional*): - Whether or not to return the attentions tensors of all attention layers. See `attentions` under - returned tensors for more detail. - use_cache (`bool`, *optional*): - If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding - (see `past_key_values`). - past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states - cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*): - Indices depicting the position of the input sequence tokens in the sequence. - position_embeddings (`Tuple[torch.FloatTensor, torch.FloatTensor]`, *optional*): - Tuple containing the cosine and sine positional embeddings of shape `(batch_size, seq_len, head_dim)`, - with `head_dim` being the embedding dimension of each attention head. - kwargs (`dict`, *optional*): - Arbitrary kwargs to be ignored, used for FSDP and other methods that injects code - into the model - """ - - residual = hidden_states - - hidden_states = self.input_layernorm(hidden_states) - - # Self Attention - hidden_states, self_attn_weights, present_key_value = self.self_attn( - hidden_states=hidden_states, - attention_mask=attention_mask, - position_ids=position_ids, - past_key_value=past_key_value, - output_attentions=output_attentions, - use_cache=use_cache, - cache_position=cache_position, - position_embeddings=position_embeddings, - ) - hidden_states = residual + hidden_states - - # Fully Connected - residual = hidden_states - hidden_states = self.post_attention_layernorm(hidden_states) - hidden_states = self.mlp(hidden_states) - hidden_states = residual + hidden_states - - outputs = (hidden_states,) - - if output_attentions: - outputs += (self_attn_weights,) - - if use_cache: - outputs += (present_key_value,) - - return outputs - - -@add_start_docstrings( - "The bare Qwen2_5_VL Model outputting raw hidden-states without any specific head on top.", - Qwen2_5_VL_START_DOCSTRING, -) -class Qwen2_5_VLModel(Qwen2_5_VLPreTrainedModel): - def __init__(self, config: Qwen2_5_VLConfig): - super().__init__(config) - self.padding_idx = config.pad_token_id - self.vocab_size = config.vocab_size - - self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx) - self.layers = nn.ModuleList( - [Qwen2_5_VLDecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)] - ) - self._attn_implementation = config._attn_implementation - self.norm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - self.rotary_emb = Qwen2_5_VLRotaryEmbedding(config=config) - - self.gradient_checkpointing = False - # Initialize weights and apply final processing - self.post_init() - - def get_input_embeddings(self): - return self.embed_tokens - - def set_input_embeddings(self, value): - self.embed_tokens = value - - def forward( - self, - input_ids: torch.LongTensor = None, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_values: list[torch.FloatTensor] | None = None, - inputs_embeds: torch.FloatTensor | None = None, - use_cache: bool | None = None, - output_attentions: bool | None = None, - output_hidden_states: bool | None = None, - return_dict: bool | None = None, - cache_position: torch.LongTensor | None = None, - ) -> tuple | BaseModelOutputWithPast: - output_attentions = ( - output_attentions if output_attentions is not None else self.config.output_attentions - ) - output_hidden_states = ( - output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states - ) - use_cache = use_cache if use_cache is not None else self.config.use_cache - - return_dict = return_dict if return_dict is not None else self.config.use_return_dict - - if (input_ids is None) ^ (inputs_embeds is not None): - raise ValueError("You must specify exactly one of input_ids or inputs_embeds") - - if self.gradient_checkpointing and self.training: - if use_cache: - logger.warning_once( - "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..." - ) - use_cache = False - - # torch.jit.trace() doesn't support cache objects in the output - if use_cache and past_key_values is None and not torch.jit.is_tracing(): - past_key_values = DynamicCache() - - if inputs_embeds is None: - inputs_embeds = self.embed_tokens(input_ids) - - if cache_position is None: - past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 - cache_position = torch.arange( - past_seen_tokens, - past_seen_tokens + inputs_embeds.shape[1], - device=inputs_embeds.device, - ) - - # the hard coded `3` is for temporal, height and width. - if position_ids is None: - position_ids = cache_position.view(1, 1, -1).expand(3, inputs_embeds.shape[0], -1) - elif position_ids.dim() == 2: - position_ids = position_ids[None, ...].expand(3, position_ids.shape[0], -1) - - causal_mask = self._update_causal_mask( - attention_mask, - inputs_embeds, - cache_position, - past_key_values, - output_attentions, - ) - - hidden_states = inputs_embeds - - # create position embeddings to be shared across the decoder layers - position_embeddings = self.rotary_emb(hidden_states, position_ids) - - # decoder layers - all_hidden_states = () if output_hidden_states else None - all_self_attns = () if output_attentions else None - next_decoder_cache = None - - for decoder_layer in self.layers: - if output_hidden_states: - all_hidden_states += (hidden_states,) - - if self.gradient_checkpointing and self.training: - layer_outputs = self._gradient_checkpointing_func( - decoder_layer.__call__, - hidden_states, - causal_mask, - position_ids, - past_key_values, - output_attentions, - use_cache, - cache_position, - position_embeddings, - ) - else: - layer_outputs = decoder_layer( - hidden_states, - attention_mask=causal_mask, - position_ids=position_ids, - past_key_value=past_key_values, - output_attentions=output_attentions, - use_cache=use_cache, - cache_position=cache_position, - position_embeddings=position_embeddings, - ) - - hidden_states = layer_outputs[0] - - if use_cache: - next_decoder_cache = layer_outputs[2 if output_attentions else 1] - - if output_attentions: - all_self_attns += (layer_outputs[1],) - - hidden_states = self.norm(hidden_states) - - # add hidden states from the last decoder layer - if output_hidden_states: - all_hidden_states += (hidden_states,) - - next_cache = next_decoder_cache if use_cache else None - - if not return_dict: - return tuple( - v for v in [hidden_states, next_cache, all_hidden_states, all_self_attns] if v is not None - ) - return BaseModelOutputWithPast( - last_hidden_state=hidden_states, - past_key_values=next_cache, - hidden_states=all_hidden_states, - attentions=all_self_attns, - ) - - def _update_causal_mask( - self, - attention_mask: torch.Tensor, - input_tensor: torch.Tensor, - cache_position: torch.Tensor, - past_key_values: Cache, - output_attentions: bool, - ): - if self.config._attn_implementation == "flash_attention_2": - if attention_mask is not None and past_key_values is not None: - is_padding_right = attention_mask[:, -1].sum().item() != input_tensor.size()[0] - if is_padding_right: - raise ValueError( - "You are attempting to perform batched generation with padding_side='right'" - " this may lead to unexpected behaviour for Flash Attention version of Qwen2_5_VL. Make sure to " - " call `tokenizer.padding_side = 'left'` before tokenizing the input. " - ) - if attention_mask is not None and 0.0 in attention_mask: - return attention_mask - return None - - # For SDPA, when possible, we will rely on its `is_causal` argument instead of its `attn_mask` argument, in - # order to dispatch on Flash Attention 2. This feature is not compatible with static cache, as SDPA will fail - # to infer the attention mask. - past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 - using_static_cache = isinstance(past_key_values, StaticCache) - using_sliding_window_cache = isinstance(past_key_values, SlidingWindowCache) - - # When output attentions is True, sdpa implementation's forward method calls the eager implementation's forward - if ( - self.config._attn_implementation == "sdpa" - and not (using_static_cache or using_sliding_window_cache) - and not output_attentions - ): - if AttentionMaskConverter._ignore_causal_mask_sdpa( - attention_mask, - inputs_embeds=input_tensor, - past_key_values_length=past_seen_tokens, - sliding_window=self.config.sliding_window, - is_training=self.training, - ): - return None - - dtype, device = input_tensor.dtype, input_tensor.device - min_dtype = torch.finfo(dtype).min - sequence_length = input_tensor.shape[1] - # SlidingWindowCache or StaticCache - if using_sliding_window_cache or using_static_cache: - target_length = past_key_values.get_max_cache_shape() - # DynamicCache or no cache - else: - target_length = ( - attention_mask.shape[-1] - if isinstance(attention_mask, torch.Tensor) - else past_seen_tokens + sequence_length + 1 - ) - - # In case the provided `attention` mask is 2D, we generate a causal mask here (4D). - causal_mask = self._prepare_4d_causal_attention_mask_with_cache_position( - attention_mask, - sequence_length=sequence_length, - target_length=target_length, - dtype=dtype, - device=device, - cache_position=cache_position, - batch_size=input_tensor.shape[0], - config=self.config, - past_key_values=past_key_values, - ) - - if ( - self.config._attn_implementation == "sdpa" - and attention_mask is not None - and attention_mask.device.type in ["cuda", "xpu"] - and not output_attentions - ): - # Attend to all tokens in fully masked rows in the causal_mask, for example the relevant first rows when - # using left padding. This is required by F.scaled_dot_product_attention memory-efficient attention path. - # Details: https://github.com/pytorch/pytorch/issues/110213 - causal_mask = AttentionMaskConverter._unmask_unattended(causal_mask, min_dtype) - - return causal_mask - - @staticmethod - def _prepare_4d_causal_attention_mask_with_cache_position( - attention_mask: torch.Tensor, - sequence_length: int, - target_length: int, - dtype: torch.dtype, - device: torch.device, - cache_position: torch.Tensor, - batch_size: int, - config: Qwen2_5_VLConfig, - past_key_values: Cache, - ): - """ - Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape - `(batch_size, key_value_length)`, or if the input `attention_mask` is already 4D, do nothing. - - Args: - attention_mask (`torch.Tensor`): - A 2D attention mask of shape `(batch_size, key_value_length)` or a 4D attention mask of shape `(batch_size, 1, query_length, key_value_length)`. - sequence_length (`int`): - The sequence length being processed. - target_length (`int`): - The target length: when generating with static cache, the mask should be as long as the static cache, to account for the 0 padding, the part of the cache that is not filled yet. - dtype (`torch.dtype`): - The dtype to use for the 4D attention mask. - device (`torch.device`): - The device to place the 4D attention mask on. - cache_position (`torch.Tensor`): - Indices depicting the position of the input sequence tokens in the sequence. - batch_size (`torch.Tensor`): - Batch size. - config (`Qwen2_5_VLConfig`): - The model's configuration class - past_key_values (`Cache`): - The cache class that is being used currently to generate - """ - if attention_mask is not None and attention_mask.dim() == 4: - # In this case we assume that the mask comes already in inverted form and requires no inversion or slicing. - causal_mask = attention_mask - else: - min_dtype = torch.finfo(dtype).min - causal_mask = torch.full( - (sequence_length, target_length), - fill_value=min_dtype, - dtype=dtype, - device=device, - ) - diagonal_attend_mask = torch.arange(target_length, device=device) > cache_position.reshape(-1, 1) - if config.sliding_window is not None: - # if we have sliding window, we should not attend to tokens beyond sliding window length, so we mask them out also - # the check is needed to verify is current checkpoint was trained with sliding window or not - if not isinstance(past_key_values, SlidingWindowCache) or sequence_length > target_length: - sliding_attend_mask = torch.arange(target_length, device=device) <= ( - cache_position.reshape(-1, 1) - config.sliding_window - ) - diagonal_attend_mask.bitwise_or_(sliding_attend_mask) - causal_mask *= diagonal_attend_mask - causal_mask = causal_mask[None, None, :, :].expand(batch_size, 1, -1, -1) - if attention_mask is not None: - causal_mask = causal_mask.clone() # copy to contiguous memory for in-place edit - if attention_mask.shape[-1] > target_length: - attention_mask = attention_mask[:, :target_length] - mask_length = attention_mask.shape[-1] - padding_mask = causal_mask[:, :, :, :mask_length] + attention_mask[:, None, None, :].to( - causal_mask.device - ) - padding_mask = padding_mask == 0 - causal_mask[:, :, :, :mask_length] = causal_mask[:, :, :, :mask_length].masked_fill( - padding_mask, min_dtype - ) - return causal_mask +from .configuration_qwen2_5_vl import Qwen2_5_VLConfig, Qwen2_5_VLTextConfig @dataclass -class Qwen2_5_VLCausalLMOutputWithPast(ModelOutput): - """ - Base class for Qwen2_5_VL causal language model (or autoregressive) outputs. - - Args: - loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided): - Language modeling loss (for next-token prediction). - logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`): - Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax). - past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`): - Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape - `(batch_size, num_heads, sequence_length, embed_size_per_head)`) - - Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see - `past_key_values` input) to speed up sequential decoding. - hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`): - Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + - one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. - - Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`): - Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, - sequence_length)`. - - Attentions weights after the attention softmax, used to compute the weighted average in the self-attention - heads. - rope_deltas (`torch.LongTensor` of shape `(batch_size, )`, *optional*): - The rope index difference between sequence length and multimodal rope. - """ - - loss: torch.FloatTensor | None = None - logits: torch.FloatTensor = None - past_key_values: list[torch.FloatTensor] | None = None - hidden_states: tuple[torch.FloatTensor] | None = None - attentions: tuple[torch.FloatTensor] | None = None - rope_deltas: torch.LongTensor | None = None - - -QWEN2_5_VL_INPUTS_DOCSTRING = r""" - Args: - input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): - Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide - it. - - Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and - [`PreTrainedTokenizer.__call__`] for details. - - [What are input IDs?](../glossary#input-ids) - attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*): - Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - - - 1 for tokens that are **not masked**, - - 0 for tokens that are **masked**. - - [What are attention masks?](../glossary#attention-mask) - - Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and - [`PreTrainedTokenizer.__call__`] for details. - - If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see - `past_key_values`). - - If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`] - and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more - information on the default strategy. - - - 1 indicates the head is **not masked**, - - 0 indicates the head is **masked**. - position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): - Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, - config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`): - Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape - `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape - `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`. - - Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention - blocks) that can be used (see `past_key_values` input) to speed up sequential decoding. - - If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that - don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all - `decoder_input_ids` of shape `(batch_size, sequence_length)`. - inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*): - Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This - is useful if you want more control over how to convert `input_ids` indices into associated vectors than the - model's internal embedding lookup matrix. - use_cache (`bool`, *optional*): - If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see - `past_key_values`). - output_attentions (`bool`, *optional*): - Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned - tensors for more detail. - output_hidden_states (`bool`, *optional*): - Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for - more detail. - return_dict (`bool`, *optional*): - Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. - pixel_values (`torch.FloatTensor` of shape `(seq_length, num_channels * image_size * image_size)): - The tensors corresponding to the input images. Pixel values can be obtained using - [`AutoImageProcessor`]. See [`Qwen2_5_VLImageProcessor.__call__`] for details. [`Qwen2_5_VLProcessor`] uses - [`Qwen2_5_VLImageProcessor`] for processing images. - pixel_values_videos (`torch.FloatTensor` of shape `(seq_length, num_channels * temporal_size * image_size * image_size)): - The tensors corresponding to the input videos. Pixel values can be obtained using - [`AutoImageProcessor`]. See [`Qwen2_5_VLImageProcessor.__call__`] for details. [`Qwen2_5_VLProcessor`] uses - [`Qwen2_5_VLImageProcessor`] for processing videos. - image_grid_thw (`torch.LongTensor` of shape `(num_images, 3)`, *optional*): - The temporal, height and width of feature shape of each image in LLM. - video_grid_thw (`torch.LongTensor` of shape `(num_videos, 3)`, *optional*): - The temporal, height and width of feature shape of each video in LLM. - rope_deltas (`torch.LongTensor` of shape `(batch_size, )`, *optional*): - The rope index difference between sequence length and multimodal rope. -""" - - -class Qwen2_5_VLForConditionalGeneration(Qwen2_5_VLPreTrainedModel, GenerationMixin): - _tied_weights_keys = {"lm_head.weight": "model.embed_tokens.weight"} - config_class = Qwen2_5_VLConfig - _no_split_modules = ["Qwen2_5_VLDecoderLayer", "Qwen2_5_VLVisionBlock"] - - def __init__(self, config): - super().__init__(config) - self.visual = Qwen2_5_VisionTransformerPretrainedModel._from_config(config.vision_config) - self.model = Qwen2_5_VLModel(config) - self.vocab_size = config.vocab_size - self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False) - self.rope_deltas = None # cache rope_deltas here - - # Initialize weights and apply final processing - self.post_init() - - def get_input_embeddings(self): - return self.model.embed_tokens - - def set_input_embeddings(self, value): - self.model.embed_tokens = value - - def get_output_embeddings(self): - return self.lm_head - - def set_output_embeddings(self, new_embeddings): - self.lm_head = new_embeddings - - def set_decoder(self, decoder): - self.model = decoder - - def get_decoder(self): - return self.model - - def get_rope_index( - self, - input_ids: torch.LongTensor | None = None, - image_grid_thw: torch.LongTensor | None = None, - video_grid_thw: torch.LongTensor | None = None, - second_per_grid_ts: torch.Tensor | None = None, - attention_mask: torch.Tensor | None = None, - ) -> tuple[torch.Tensor, torch.Tensor]: - """ - Calculate the 3D rope index based on image and video's temporal, height and width in LLM. - - Explanation: - Each embedding sequence contains vision embedding and text embedding or just contains text embedding. - - For pure text embedding sequence, the rotary position embedding has no difference with modern LLMs. - Examples: - input_ids: [T T T T T], here T is for text. - temporal position_ids: [0, 1, 2, 3, 4] - height position_ids: [0, 1, 2, 3, 4] - width position_ids: [0, 1, 2, 3, 4] - - For vision and text embedding sequence, we calculate 3D rotary position embedding for vision part - and 1D rotary position embedding for text part. - Examples: - Temporal (Time): 3 patches, representing different segments of the video in time. - Height: 2 patches, dividing each frame vertically. - Width: 2 patches, dividing each frame horizontally. - We also have some important parameters: - fps (Frames Per Second): The video's frame rate, set to 1. This means one frame is processed each second. - tokens_per_second: This is a crucial parameter. It dictates how many "time-steps" or "temporal tokens" are conceptually packed into a one-second interval of the video. In this case, we have 25 tokens per second. So each second of the video will be represented with 25 separate time points. It essentially defines the temporal granularity. - temporal_patch_size: The number of frames that compose one temporal patch. Here, it's 2 frames. - interval: The step size for the temporal position IDs, calculated as tokens_per_second * temporal_patch_size / fps. In this case, 25 * 2 / 1 = 50. This means that each temporal patch will be have a difference of 50 in the temporal position IDs. - input_ids: [V V V V V V V V V V V V T T T T T], here V is for vision. - vision temporal position_ids: [0, 0, 0, 0, 50, 50, 50, 50, 100, 100, 100, 100] - vision height position_ids: [0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1] - vision width position_ids: [0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1] - text temporal position_ids: [101, 102, 103, 104, 105] - text height position_ids: [101, 102, 103, 104, 105] - text width position_ids: [101, 102, 103, 104, 105] - Here we calculate the text start position_ids as the max vision position_ids plus 1. - - Args: - input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): - Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide - it. - image_grid_thw (`torch.LongTensor` of shape `(num_images, 3)`, *optional*): - The temporal, height and width of feature shape of each image in LLM. - video_grid_thw (`torch.LongTensor` of shape `(num_videos, 3)`, *optional*): - The temporal, height and width of feature shape of each video in LLM. - second_per_grid_ts (`torch.Tensor` of shape `(num_videos)`, *optional*): - The time interval (in seconds) for each grid along the temporal dimension in the 3D position IDs. - attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*): - Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - - - 1 for tokens that are **not masked**, - - 0 for tokens that are **masked**. - - Returns: - position_ids (`torch.LongTensor` of shape `(3, batch_size, sequence_length)`) - mrope_position_deltas (`torch.Tensor` of shape `(batch_size)`) - """ - spatial_merge_size = self.config.vision_config.spatial_merge_size - image_token_id = self.config.image_token_id - video_token_id = self.config.video_token_id - vision_start_token_id = self.config.vision_start_token_id - mrope_position_deltas = [] - if input_ids is not None and (image_grid_thw is not None or video_grid_thw is not None): - total_input_ids = input_ids - if attention_mask is None: - attention_mask = torch.ones_like(total_input_ids) - position_ids = torch.ones( - 3, - input_ids.shape[0], - input_ids.shape[1], - dtype=input_ids.dtype, - device=input_ids.device, - ) - image_index, video_index = 0, 0 - attention_mask = attention_mask.to(total_input_ids.device) - for i, input_ids in enumerate(total_input_ids): - input_ids = input_ids[attention_mask[i] == 1] - image_nums, video_nums = 0, 0 - vision_start_indices = torch.argwhere(input_ids == vision_start_token_id).squeeze(1) - vision_tokens = input_ids[vision_start_indices + 1] - image_nums = (vision_tokens == image_token_id).sum() - video_nums = (vision_tokens == video_token_id).sum() - input_tokens = input_ids.tolist() - llm_pos_ids_list: list = [] - st = 0 - remain_images, remain_videos = image_nums, video_nums - for _ in range(image_nums + video_nums): - if image_token_id in input_tokens and remain_images > 0: - ed_image = input_tokens.index(image_token_id, st) - else: - ed_image = len(input_tokens) + 1 - if video_token_id in input_tokens and remain_videos > 0: - ed_video = input_tokens.index(video_token_id, st) - else: - ed_video = len(input_tokens) + 1 - if ed_image < ed_video: - t, h, w = ( - image_grid_thw[image_index][0], - image_grid_thw[image_index][1], - image_grid_thw[image_index][2], - ) - second_per_grid_t = 0 - image_index += 1 - remain_images -= 1 - ed = ed_image - - else: - t, h, w = ( - video_grid_thw[video_index][0], - video_grid_thw[video_index][1], - video_grid_thw[video_index][2], - ) - if second_per_grid_ts is not None: - second_per_grid_t = second_per_grid_ts[video_index] - else: - second_per_grid_t = 1.0 - video_index += 1 - remain_videos -= 1 - ed = ed_video - llm_grid_t, llm_grid_h, llm_grid_w = ( - t.item(), - h.item() // spatial_merge_size, - w.item() // spatial_merge_size, - ) - text_len = ed - st - - st_idx = llm_pos_ids_list[-1].max() + 1 if len(llm_pos_ids_list) > 0 else 0 - llm_pos_ids_list.append(torch.arange(text_len).view(1, -1).expand(3, -1) + st_idx) - - range_tensor = torch.arange(llm_grid_t).view(-1, 1) - expanded_range = range_tensor.expand(-1, llm_grid_h * llm_grid_w) - - time_tensor = ( - expanded_range * second_per_grid_t * self.config.vision_config.tokens_per_second - ) - - time_tensor_long = time_tensor.long() - t_index = time_tensor_long.flatten() - - h_index = ( - torch.arange(llm_grid_h).view(1, -1, 1).expand(llm_grid_t, -1, llm_grid_w).flatten() - ) - w_index = ( - torch.arange(llm_grid_w).view(1, 1, -1).expand(llm_grid_t, llm_grid_h, -1).flatten() - ) - llm_pos_ids_list.append(torch.stack([t_index, h_index, w_index]) + text_len + st_idx) - st = ed + llm_grid_t * llm_grid_h * llm_grid_w - - if st < len(input_tokens): - st_idx = llm_pos_ids_list[-1].max() + 1 if len(llm_pos_ids_list) > 0 else 0 - text_len = len(input_tokens) - st - llm_pos_ids_list.append(torch.arange(text_len).view(1, -1).expand(3, -1) + st_idx) - - llm_positions = torch.cat(llm_pos_ids_list, dim=1).reshape(3, -1) - position_ids[..., i, attention_mask[i] == 1] = llm_positions.to(position_ids.device) - mrope_position_deltas.append(llm_positions.max() + 1 - len(total_input_ids[i])) - mrope_position_deltas = torch.tensor(mrope_position_deltas, device=input_ids.device).unsqueeze(1) - return position_ids, mrope_position_deltas - else: - if attention_mask is not None: - position_ids = attention_mask.long().cumsum(-1) - 1 - position_ids.masked_fill_(attention_mask == 0, 1) - position_ids = position_ids.unsqueeze(0).expand(3, -1, -1).to(attention_mask.device) - max_position_ids = position_ids.max(0, keepdim=False)[0].max(-1, keepdim=True)[0] - mrope_position_deltas = max_position_ids + 1 - attention_mask.shape[-1] - else: - position_ids = ( - torch.arange(input_ids.shape[1], device=input_ids.device) - .view(1, 1, -1) - .expand(3, input_ids.shape[0], -1) - ) - mrope_position_deltas = torch.zeros( - [input_ids.shape[0], 1], - device=input_ids.device, - dtype=input_ids.dtype, - ) - - return position_ids, mrope_position_deltas - - @add_start_docstrings_to_model_forward(QWEN2_5_VL_INPUTS_DOCSTRING) - @replace_return_docstrings(output_type=Qwen2_5_VLCausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC) - def forward( - self, - input_ids: torch.LongTensor = None, - attention_mask: torch.Tensor | None = None, - position_ids: torch.LongTensor | None = None, - past_key_values: list[torch.FloatTensor] | None = None, - inputs_embeds: torch.FloatTensor | None = None, - labels: torch.LongTensor | None = None, - use_cache: bool | None = None, - output_attentions: bool | None = None, - output_hidden_states: bool | None = None, - return_dict: bool | None = None, - pixel_values: torch.Tensor | None = None, - pixel_values_videos: torch.FloatTensor | None = None, - image_grid_thw: torch.LongTensor | None = None, - video_grid_thw: torch.LongTensor | None = None, - rope_deltas: torch.LongTensor | None = None, - cache_position: torch.LongTensor | None = None, - second_per_grid_ts: torch.Tensor | None = None, - ) -> tuple | Qwen2_5_VLCausalLMOutputWithPast: - r""" - Args: - labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): - Labels for computing the masked language modeling loss. Indices should either be in `[0, ..., - config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored - (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`. - - Returns: - - Example: - - ```python - >>> from PIL import Image - >>> import requests - >>> from transformers import AutoProcessor, Qwen2_5_VLForConditionalGeneration - - >>> model = Qwen2_5_VLForConditionalGeneration.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct") - >>> processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct") - - >>> messages = [ - { - "role": "user", - "content": [ - {"type": "image"}, - {"type": "text", "text": "What is shown in this image?"}, - ], - }, - ] - >>> url = "https://www.ilankelman.org/stopsigns/australia.jpg" - >>> image = Image.open(requests.get(url, stream=True).raw) - - >>> text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) - >>> inputs = processor(text=[text], images=[image], vision_infos=[vision_infos]) - - >>> # Generate - >>> generate_ids = model.generate(inputs.input_ids, max_length=30) - >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0] - "The image shows a street scene with a red stop sign in the foreground. In the background, there is a large red gate with Chinese characters ..." - ```""" - - output_attentions = ( - output_attentions if output_attentions is not None else self.config.output_attentions - ) - output_hidden_states = ( - output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states - ) - return_dict = return_dict if return_dict is not None else self.config.use_return_dict - - if inputs_embeds is None: - inputs_embeds = self.model.embed_tokens(input_ids) - if pixel_values is not None: - pixel_values = pixel_values.type(self.visual.dtype) - image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw) - n_image_tokens = (input_ids == self.config.image_token_id).sum().item() - n_image_features = image_embeds.shape[0] - if n_image_tokens != n_image_features: - raise ValueError( - f"Image features and image tokens do not match: tokens: {n_image_tokens}, features {n_image_features}" - ) - - mask = input_ids == self.config.image_token_id - mask_unsqueezed = mask.unsqueeze(-1) - mask_expanded = mask_unsqueezed.expand_as(inputs_embeds) - image_mask = mask_expanded.to(inputs_embeds.device) - - image_embeds = image_embeds.to(inputs_embeds.device, inputs_embeds.dtype) - inputs_embeds = inputs_embeds.masked_scatter(image_mask, image_embeds) - - if pixel_values_videos is not None: - pixel_values_videos = pixel_values_videos.type(self.visual.dtype) - video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw) - n_video_tokens = (input_ids == self.config.video_token_id).sum().item() - n_video_features = video_embeds.shape[0] - if n_video_tokens != n_video_features: - raise ValueError( - f"Video features and video tokens do not match: tokens: {n_video_tokens}, features {n_video_features}" - ) - - mask = input_ids == self.config.video_token_id - mask_unsqueezed = mask.unsqueeze(-1) - mask_expanded = mask_unsqueezed.expand_as(inputs_embeds) - video_mask = mask_expanded.to(inputs_embeds.device) - - video_embeds = video_embeds.to(inputs_embeds.device, inputs_embeds.dtype) - inputs_embeds = inputs_embeds.masked_scatter(video_mask, video_embeds) - - if attention_mask is not None: - attention_mask = attention_mask.to(inputs_embeds.device) - - # if we get 4D attention mask we cannot calculate rope deltas anymore. TODO @raushan fixme - if position_ids is None and (attention_mask is None or attention_mask.ndim == 2): - # calculate RoPE index once per generation in the pre-fill stage only - if ( - (cache_position is not None and cache_position[0] == 0) - or self.rope_deltas is None - or (past_key_values is None or past_key_values.get_seq_length() == 0) - ): - position_ids, rope_deltas = self.get_rope_index( - input_ids, - image_grid_thw, - video_grid_thw, - second_per_grid_ts, - attention_mask, - ) - self.rope_deltas = rope_deltas - # then use the prev pre-calculated rope-deltas to get the correct position ids - else: - batch_size, seq_length, _ = inputs_embeds.shape - delta = ( - (cache_position[0] + self.rope_deltas).to(inputs_embeds.device) - if cache_position is not None - else 0 - ) - position_ids = torch.arange(seq_length, device=inputs_embeds.device) - position_ids = position_ids.view(1, -1).expand(batch_size, -1) - if cache_position is not None: # otherwise `deltas` is an int `0` - delta = delta.repeat_interleave(batch_size // delta.shape[0], dim=0) - position_ids = position_ids.add(delta) - position_ids = position_ids.unsqueeze(0).expand(3, -1, -1) - - outputs = self.model( - input_ids=None, - position_ids=position_ids, - attention_mask=attention_mask, - past_key_values=past_key_values, - inputs_embeds=inputs_embeds, - use_cache=use_cache, - output_attentions=output_attentions, - output_hidden_states=output_hidden_states, - return_dict=return_dict, - cache_position=cache_position, - ) - - hidden_states = outputs[0] - logits = self.lm_head(hidden_states) - - loss = None - if labels is not None: - # Upcast to float if we need to compute the loss to avoid potential precision issues - logits = logits.float() - # Shift so that tokens < n predict n - shift_logits = logits[..., :-1, :].contiguous() - shift_labels = labels[..., 1:].contiguous() - # Flatten the tokens - loss_fct = CrossEntropyLoss() - shift_logits = shift_logits.view(-1, self.config.vocab_size) - shift_labels = shift_labels.view(-1) - # Enable model parallelism - shift_labels = shift_labels.to(shift_logits.device) - loss = loss_fct(shift_logits, shift_labels) - - if not return_dict: - output = (logits,) + outputs[1:] - return (loss,) + output if loss is not None else output - - return Qwen2_5_VLCausalLMOutputWithPast( - loss=loss, - logits=logits, - past_key_values=outputs.past_key_values, - hidden_states=outputs.hidden_states, - attentions=outputs.attentions, - rope_deltas=self.rope_deltas, - ) - - def prepare_inputs_for_generation( - self, - input_ids, - past_key_values=None, - attention_mask=None, - inputs_embeds=None, - cache_position=None, - position_ids=None, - use_cache=True, - pixel_values=None, - pixel_values_videos=None, - image_grid_thw=None, - video_grid_thw=None, - second_per_grid_ts=None, - **kwargs, - ): - # Overwritten -- in specific circumstances we don't want to forward image inputs to the model - - # If we have cache: let's slice `input_ids` through `cache_position`, to keep only the unprocessed tokens - # Exception 1: when passing input_embeds, input_ids may be missing entries - # Exception 2: some generation methods do special slicing of input_ids, so we don't need to do it here - # Exception 3: with synced GPUs cache_position may go out of bounds, but we only want dummy token in that case. - # (we can't check exception 3 while compiling) - # Exception 4: If input_embeds are passed then slice it through `cache_position`, to keep only the unprocessed tokens and - # generate the first token for each sequence. Later use the generated Input ids for continuation. - if past_key_values is not None: - if inputs_embeds is not None and input_ids.shape[1] == 0: # Exception 4 - inputs_embeds = inputs_embeds[:, -cache_position.shape[0] :] - elif inputs_embeds is not None or ( # Exception 1 - is_torchdynamo_compiling() or cache_position[-1] >= input_ids.shape[1] - ): # Exception 3 - input_ids = input_ids[:, -cache_position.shape[0] :] - elif ( - input_ids.shape[1] != cache_position.shape[0] - ): # Default case (the "else", a no op, is Exception 2) - input_ids = input_ids[:, cache_position] - - if cache_position[0] != 0: - pixel_values = None - pixel_values_videos = None - - # if `inputs_embeds` are passed, we only want to use them in the 1st generation step - if inputs_embeds is not None and len(cache_position) == inputs_embeds.shape[1]: - model_inputs = {"inputs_embeds": inputs_embeds, "input_ids": None} - else: - model_inputs = {"input_ids": input_ids, "inputs_embeds": None} - - if isinstance(past_key_values, StaticCache) and attention_mask.ndim == 2: - if model_inputs["inputs_embeds"] is not None: - batch_size, sequence_length, _ = inputs_embeds.shape - device = inputs_embeds.device - else: - batch_size, sequence_length = input_ids.shape - device = input_ids.device - - attention_mask = self.model._prepare_4d_causal_attention_mask_with_cache_position( - attention_mask, - sequence_length=sequence_length, - target_length=past_key_values.get_max_cache_shape(), - dtype=self.lm_head.weight.dtype, - device=device, - cache_position=cache_position, - batch_size=batch_size, - config=self.config, - past_key_values=past_key_values, - ) - - model_inputs.update( - { - "position_ids": position_ids, - "past_key_values": past_key_values, - "use_cache": use_cache, - "attention_mask": attention_mask, - "pixel_values": pixel_values, - "pixel_values_videos": pixel_values_videos, - "image_grid_thw": image_grid_thw, - "video_grid_thw": video_grid_thw, - "cache_position": cache_position, - "second_per_grid_ts": second_per_grid_ts, - } - ) - return model_inputs - - def _get_image_nums_and_video_nums( - self, - input_ids: torch.LongTensor | None, - ) -> tuple[torch.Tensor, torch.Tensor]: - """ - Get the number of images and videos for each sample to calculate the separation length of the sample tensor. - These parameters are not passed through the processor to avoid unpredictable impacts from interface modifications. - - Args: - input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): - Indices of input sequence tokens in the vocabulary. - - Returns: - image_nums (`torch.LongTensor` of shape `(batch_size, num_images_sample)`) - video_nums (`torch.LongTensor` of shape `(batch_size, num_videos_sample)`) - """ - image_token_id = self.config.image_token_id - video_token_id = self.config.video_token_id - vision_start_token_id = self.config.vision_start_token_id - - vision_start_mask = input_ids == vision_start_token_id - vision_first_mask = torch.roll(vision_start_mask, shifts=1, dims=1) - image_mask = input_ids == image_token_id - video_mask = input_ids == video_token_id - image_nums = torch.sum(vision_first_mask & image_mask, dim=1) - video_nums = torch.sum(vision_first_mask & video_mask, dim=1) - - return image_nums, video_nums - - def _expand_inputs_for_generation( - self, - expand_size: int = 1, - is_encoder_decoder: bool = False, - input_ids: torch.LongTensor | None = None, - **model_kwargs, - ) -> tuple[torch.LongTensor, dict[str, Any]]: - # Overwritten -- Support for expanding tensors without a batch size dimension - # e.g., pixel_values, image_grid_thw, pixel_values_videos, video_grid_thw, second_per_grid_t - # pixel_values.shape[0] is sum(seqlen_images for samples) - # image_grid_thw.shape[0] is sum(num_images for samples) - - if expand_size == 1: - return input_ids, model_kwargs - - visual_keys = [ - "pixel_values", - "image_grid_thw", - "pixel_values_videos", - "video_grid_thw", - "second_per_grid_ts", - ] - - def _expand_dict_for_generation_visual(dict_to_expand): - image_grid_thw = model_kwargs.get("image_grid_thw", None) - video_grid_thw = model_kwargs.get("video_grid_thw", None) - image_nums, video_nums = self._get_image_nums_and_video_nums(input_ids) - - def _repeat_interleave_samples(x, lengths, repeat_times): - samples = torch.split(x, lengths) - repeat_args = [repeat_times] + [1] * (x.dim() - 1) - result = torch.cat([sample.repeat(*repeat_args) for sample in samples], dim=0) - return result - - for key in dict_to_expand: - if key == "pixel_values": - # split images into samples - samples = torch.split(image_grid_thw, list(image_nums)) - # compute the sequence length of images for each sample - lengths = [torch.prod(sample, dim=1).sum() for sample in samples] - dict_to_expand[key] = _repeat_interleave_samples( - dict_to_expand[key], lengths=lengths, repeat_times=expand_size - ) - elif key == "image_grid_thw": - # get the num of images for each sample - lengths = list(image_nums) - dict_to_expand[key] = _repeat_interleave_samples( - dict_to_expand[key], lengths=lengths, repeat_times=expand_size - ) - elif key == "pixel_values_videos": - samples = torch.split(video_grid_thw, list(video_nums)) - lengths = [torch.prod(sample, dim=1).sum() for sample in samples] - dict_to_expand[key] = _repeat_interleave_samples( - dict_to_expand[key], lengths=lengths, repeat_times=expand_size - ) - elif key == "video_grid_thw": - lengths = list(video_nums) - dict_to_expand[key] = _repeat_interleave_samples( - dict_to_expand[key], lengths=lengths, repeat_times=expand_size - ) - elif key == "second_per_grid_ts": - if not isinstance(dict_to_expand[key], list): - raise TypeError( - f"Expected value for key '{key}' to be a list, but got {type(dict_to_expand[key])} instead." - ) - tensor = torch.tensor(dict_to_expand[key]) - lengths = list(video_nums) - tensor = _repeat_interleave_samples(tensor, lengths=lengths, repeat_times=expand_size) - dict_to_expand[key] = tensor.tolist() - return dict_to_expand - - def _expand_dict_for_generation(dict_to_expand): - for key in dict_to_expand: - if ( - key != "cache_position" - and dict_to_expand[key] is not None - and isinstance(dict_to_expand[key], torch.Tensor) - and key not in visual_keys - ): - dict_to_expand[key] = dict_to_expand[key].repeat_interleave(expand_size, dim=0) - return dict_to_expand - - # input_ids is required for expanding visual inputs - # If input_ids is unavailable, visual inputs will not be used; therefore, there is no need to expand visual inputs. - if input_ids is not None and input_ids.numel() != 0: - model_kwargs = _expand_dict_for_generation_visual(model_kwargs) - - if input_ids is not None: - input_ids = input_ids.repeat_interleave(expand_size, dim=0) - - model_kwargs = _expand_dict_for_generation(model_kwargs) - - if is_encoder_decoder: - if model_kwargs.get("encoder_outputs") is None: - raise ValueError( - "If `is_encoder_decoder` is True, make sure that `encoder_outputs` is defined." - ) - model_kwargs["encoder_outputs"] = _expand_dict_for_generation(model_kwargs["encoder_outputs"]) - - return input_ids, model_kwargs - - -@dataclass -class Qwen2_5_VLACausalLMOutputWithPast(ModelOutput): +class Qwen2_5_VLACausalLMOutputWithPast(ModelOutput): # noqa: N801 loss: torch.FloatTensor | None = None flow_loss: torch.FloatTensor | None = None cross_entropy_loss: torch.FloatTensor | None = None @@ -2284,83 +142,48 @@ class SparseMoeBlock(nn.Module): return output -QWEN2_5_VL_ATTENTION_CLASSES = { - "eager": Qwen2_5_VLAttention, - "flash_attention_2": Qwen2_5_VLFlashAttention2, - "sdpa": Qwen2_5_VLSdpaAttention, -} +class Qwen2_5_VLDecoderLayer_with_MoE(Qwen2_5_VLDecoderLayer): # noqa: N801 + """Native Qwen2.5-VL decoder layer with an optional hard-routed sparse-MoE MLP. + Differences from the native layer forward: + - routes the post-attention hidden states through ``SparseMoeBlock`` keyed on ``token_types`` + when ``config.mlp_moe`` is set; + - casts activations to the parameter dtype before each block, since Wall-X runs with float32 + layernorms and bfloat16 projections in the same module. + """ -class Qwen2_5_VLDecoderLayer_with_MoE(nn.Module): def __init__(self, config: Qwen2_5_VLConfig, layer_idx: int, num_experts: int): - super().__init__() - self.hidden_size = config.hidden_size - - if config.use_sliding_window and config._attn_implementation != "flash_attention_2": - logger.warning_once( - f"Sliding Window Attention is enabled but not implemented for `{config._attn_implementation}`; " - "unexpected results may be encountered." - ) - - self.self_attn = QWEN2_5_VL_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx) - - self.input_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - self.post_attention_layernorm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - + super().__init__(config, layer_idx) if config.mlp_moe: - self.moe = SparseMoeBlock(config, num_experts=num_experts) + del self.mlp self.mlp = None - else: - self.mlp = Qwen2_5_VLMLP(config) + self.moe = SparseMoeBlock(config, num_experts=num_experts) def forward( self, hidden_states: torch.Tensor, attention_mask: torch.Tensor | None = None, position_ids: torch.LongTensor | None = None, - past_key_value: tuple[torch.Tensor] | None = None, - token_types=None, - output_attentions: bool | None = False, + past_key_values: Cache | None = None, + token_types: torch.LongTensor | None = None, use_cache: bool | None = False, - cache_position: torch.LongTensor | None = None, position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, **kwargs, - ) -> tuple[torch.FloatTensor, tuple[torch.FloatTensor, torch.FloatTensor] | None]: - """ - Args: - hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` - attention_mask (`torch.FloatTensor`, *optional*): attention mask of size - `(batch, sequence_length)` where padding elements are indicated by 0. - output_attentions (`bool`, *optional*): - Whether or not to return the attentions tensors of all attention layers. See `attentions` under - returned tensors for more detail. - use_cache (`bool`, *optional*): - If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding - (see `past_key_values`). - past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states - cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*): - Indices depicting the position of the input sequence tokens in the sequence. - position_embeddings (`Tuple[torch.FloatTensor, torch.FloatTensor]`, *optional*): - Tuple containing the cosine and sine positional embeddings of shape `(batch_size, seq_len, head_dim)`, - with `head_dim` being the embedding dimension of each attention head. - kwargs (`dict`, *optional*): - Arbitrary kwargs to be ignored, used for FSDP and other methods that injects code - into the model - """ + ) -> torch.Tensor: residual = hidden_states hidden_states = hidden_states.to(self.input_layernorm.weight.dtype) hidden_states = self.input_layernorm(hidden_states) hidden_states = hidden_states.to(self.self_attn.q_proj.weight.dtype) + # Self Attention - hidden_states, self_attn_weights, present_key_value = self.self_attn( + hidden_states, _ = self.self_attn( hidden_states=hidden_states, attention_mask=attention_mask, position_ids=position_ids, - past_key_value=past_key_value, - output_attentions=output_attentions, + past_key_values=past_key_values, use_cache=use_cache, - cache_position=cache_position, position_embeddings=position_embeddings, + **kwargs, ) hidden_states = residual + hidden_states @@ -2376,442 +199,193 @@ class Qwen2_5_VLDecoderLayer_with_MoE(nn.Module): hidden_states = self.mlp(hidden_states) hidden_states = residual + hidden_states - - outputs = (hidden_states,) - - if output_attentions: - outputs += (self_attn_weights,) - if use_cache: - outputs += (present_key_value,) - return outputs + return hidden_states -class Qwen2_5_VLMoEModel(Qwen2_5_VLPreTrainedModel): - """Qwen2.5-VL model with Mixture of Experts (MoE) architecture. +class Qwen2_5_VLMoEModel(Qwen2_5_VLTextModel): # noqa: N801 + """Qwen2.5-VL text model with Mixture of Experts (MoE) decoder layers. - This model extends the base Qwen2.5-VL model by incorporating MoE layers - for improved scalability and specialization across different token types. + Extends the native ``Qwen2_5_VLTextModel`` with per-token-type expert routing and a causal-mask + override that gives action-token blocks (``moe_token_types == 1``) bidirectional attention among + themselves while keeping causal attention everywhere else. """ - @classmethod - def from_pretrained( - cls, - pretrained_model_name_or_path: str, - num_experts: int | None = None, - *args, - **kwargs, - ): - """Load a pretrained model with optional MoE configuration. + config_class = Qwen2_5_VLTextConfig + _no_split_modules = ["Qwen2_5_VLDecoderLayer_with_MoE"] - Args: - pretrained_model_name_or_path: Path or name of the pretrained model - num_experts: Number of experts for MoE layers (if not in config) - *args: Additional arguments passed to parent class - **kwargs: Additional keyword arguments passed to parent class - - Returns: - Initialized model instance with MoE configuration - """ - config = kwargs.get("config") - if config is None: - config = AutoConfig.from_pretrained(pretrained_model_name_or_path) - - # Override number of experts if specified - if num_experts is not None: - config.num_experts = num_experts - - kwargs["config"] = config - return super().from_pretrained(pretrained_model_name_or_path, *args, **kwargs) - - def __init__(self, config: Qwen2_5_VLConfig): - """Initialize the Qwen2.5-VL MoE model. - - Args: - config: Model configuration containing architecture parameters - """ - super().__init__(config) - - # Basic model parameters - self.padding_idx = config.pad_token_id - self.vocab_size = config.vocab_size - - # Model components - self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx) - - # Decoder layers with MoE support + def __init__(self, config: Qwen2_5_VLConfig | Qwen2_5_VLTextConfig): + text_config = config.text_config if isinstance(config, Qwen2_5_VLConfig) else config + self._require_eager_attention(text_config._attn_implementation) + # Transformers selects SDPA automatically when no implementation is + # requested. Wall-X's action-token islands require an explicit 4D + # mask, so opt into eager before PreTrainedModel performs that choice. + text_config._attn_implementation = "eager" + super().__init__(text_config) + # Free the parent-allocated dense layers before replacing them (pi_gemma.py precedent). + del self.layers self.layers = nn.ModuleList( [ - Qwen2_5_VLDecoderLayer_with_MoE(config, layer_idx, config.num_experts) - for layer_idx in range(config.num_hidden_layers) + Qwen2_5_VLDecoderLayer_with_MoE(text_config, layer_idx, text_config.num_experts) + for layer_idx in range(text_config.num_hidden_layers) ] ) - - # Model configuration - self._attn_implementation = config._attn_implementation - self.norm = Qwen2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) - self.rotary_emb = Qwen2_5_VLRotaryEmbedding(config=config) - self.gradient_checkpointing = False - # Initialize weights and apply final processing self.post_init() - def get_input_embeddings(self) -> nn.Embedding: - """Get the input embedding layer. + @staticmethod + def _require_eager_attention(attn_implementation: str | None) -> None: + if attn_implementation not in {None, "eager"}: + raise ValueError( + "Wall-X currently supports only attn_implementation='eager'. " + "Its bidirectional action-token islands cannot be represented " + f"correctly by {attn_implementation!r}." + ) - Returns: - The token embedding layer - """ + def get_input_embeddings(self) -> nn.Embedding: return self.embed_tokens def set_input_embeddings(self, value: nn.Embedding) -> None: - """Set the input embedding layer. - - Args: - value: New embedding layer to use - """ self.embed_tokens = value + @merge_with_config_defaults + # ``capture_outputs`` reads output_hidden_states/output_attentions from kwargs and populates + # BaseModelOutputWithPast via hooks on the decoder layers and attention modules. + @capture_outputs def forward( self, input_ids: torch.LongTensor | None = None, attention_mask: torch.Tensor | None = None, position_ids: torch.LongTensor | None = None, - past_key_values: list[torch.FloatTensor] | None = None, + past_key_values: Cache | None = None, inputs_embeds: torch.FloatTensor | None = None, moe_token_types: torch.LongTensor | None = None, use_cache: bool | None = None, - output_attentions: bool | None = None, - output_hidden_states: bool | None = None, - return_dict: bool | None = None, cache_position: torch.LongTensor | None = None, **kwargs, - ) -> tuple | BaseModelOutputWithPast: - # Set default output options - output_attentions = ( - output_attentions if output_attentions is not None else self.config.output_attentions - ) - output_hidden_states = ( - output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states - ) - use_cache = use_cache if use_cache is not None else self.config.use_cache - return_dict = return_dict if return_dict is not None else self.config.use_return_dict + ) -> BaseModelOutputWithPast: + self._require_eager_attention(self.config._attn_implementation) - # Validate inputs if (input_ids is None) ^ (inputs_embeds is not None): raise ValueError("You must specify exactly one of input_ids or inputs_embeds") if moe_token_types is None: raise ValueError("moe_token_types must be provided for MoE routing") - # Handle gradient checkpointing compatibility - if self.gradient_checkpointing and self.training: - if use_cache: - logger.warning_once( - "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..." - ) - use_cache = False - - # Initialize cache if needed if use_cache and past_key_values is None and not torch.jit.is_tracing(): - past_key_values = DynamicCache() + past_key_values = DynamicCache(config=self.config) - # Get input embeddings if inputs_embeds is None: inputs_embeds = self.embed_tokens(input_ids) - - # Set up cache position - if cache_position is None: - past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 - cache_position = torch.arange( - past_seen_tokens, - past_seen_tokens + inputs_embeds.shape[1], - device=inputs_embeds.device, + if moe_token_types.shape[-1] < inputs_embeds.shape[1]: + raise ValueError( + "moe_token_types must cover every input token; got " + f"{moe_token_types.shape[-1]} types for {inputs_embeds.shape[1]} tokens" ) + routing_token_types = moe_token_types[:, -inputs_embeds.shape[1] :] - # Set up position IDs (hardcoded 3 dimensions for temporal, height, width) + past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 if position_ids is None: - position_ids = cache_position.view(1, 1, -1).expand(3, inputs_embeds.shape[0], -1) - elif position_ids.dim() == 2: + position_ids = ( + torch.arange(inputs_embeds.shape[1], device=inputs_embeds.device) + past_seen_tokens + ) + position_ids = position_ids.view(1, 1, -1).expand(3, inputs_embeds.shape[0], -1) + elif position_ids.ndim == 2: position_ids = position_ids[None, ...].expand(3, position_ids.shape[0], -1) - # Create causal attention mask - causal_mask = self._update_causal_mask( - attention_mask, - inputs_embeds, - cache_position, - past_key_values, - output_attentions, - moe_token_types, - ) - - hidden_states = inputs_embeds - - # Create position embeddings to be shared across decoder layers - position_embeddings = self.rotary_emb(hidden_states, position_ids) - - # Initialize output collections - all_hidden_states = () if output_hidden_states else None - all_self_attns = () if output_attentions else None - next_decoder_cache = None - - # Process through decoder layers - for decoder_layer in self.layers: - if output_hidden_states: - all_hidden_states += (hidden_states,) - - if self.gradient_checkpointing and self.training: - # Use gradient checkpointing during training - layer_outputs = self._gradient_checkpointing_func( - decoder_layer.__call__, - hidden_states, - causal_mask, - position_ids, - past_key_values, - moe_token_types, - output_attentions, - use_cache, - cache_position, - position_embeddings, - ) - else: - # Regular forward pass - layer_outputs = decoder_layer( - hidden_states, - attention_mask=causal_mask, - position_ids=position_ids, - past_key_value=past_key_values, - token_types=moe_token_types, - output_attentions=output_attentions, - use_cache=use_cache, - cache_position=cache_position, - position_embeddings=position_embeddings, - ) - - hidden_states = layer_outputs[0] - - # Update cache if using it - if use_cache: - next_decoder_cache = layer_outputs[2 if output_attentions else 1] - - # Collect attention weights if requested - if output_attentions: - all_self_attns += (layer_outputs[1],) - - # Apply final layer normalization - hidden_states = self.norm(hidden_states) - - # Add final hidden states if collecting all states - if output_hidden_states: - all_hidden_states += (hidden_states,) - - next_cache = next_decoder_cache if use_cache else None - - # Return outputs in requested format - if not return_dict: - return tuple( - v for v in [hidden_states, next_cache, all_hidden_states, all_self_attns] if v is not None - ) - - return BaseModelOutputWithPast( - last_hidden_state=hidden_states, - past_key_values=next_cache, - hidden_states=all_hidden_states, - attentions=all_self_attns, - ) - - def _update_causal_mask( - self, - attention_mask: torch.Tensor, - input_tensor: torch.Tensor, - cache_position: torch.Tensor, - past_key_values: Cache, - output_attentions: bool, - moe_token_types: torch.LongTensor | None = None, - ): - """Update causal attention mask with support for bidirectional attention for specific token types. - - This method creates and modifies attention masks to support different attention patterns: - - Standard causal (unidirectional) attention for most tokens - - Bidirectional attention for specific token types (e.g., MoE routing tokens) - - Args: - attention_mask: Input attention mask to avoid attending to padding tokens - input_tensor: Input embeddings tensor for shape and device information - cache_position: Position indices for caching mechanisms - past_key_values: Cached key-value pairs from previous forward passes - output_attentions: Whether attention weights will be returned - moe_token_types: Optional tensor indicating token types for MoE routing - (type 1 tokens will use bidirectional attention) - - Returns: - Updated causal attention mask, or None if using Flash Attention 2 - """ - # Flash Attention 2 handles masking internally - if self.config._attn_implementation == "flash_attention_2": - return None - - # Calculate sequence lengths for cache management - past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 - using_static_cache = isinstance(past_key_values, StaticCache) - using_sliding_window_cache = isinstance(past_key_values, SlidingWindowCache) - - # For SDPA (Scaled Dot Product Attention), use `is_causal` argument when possible - # instead of explicit attention mask to enable Flash Attention 2 dispatch - # Note: This optimization is not compatible with static cache - if ( - self.config._attn_implementation == "sdpa" - and not (using_static_cache or using_sliding_window_cache) - and not output_attentions - ): - # Check if we can ignore the causal mask and rely on SDPA's internal handling - if AttentionMaskConverter._ignore_causal_mask_sdpa( - attention_mask, - inputs_embeds=input_tensor, - past_key_values_length=past_seen_tokens, - sliding_window=self.config.sliding_window, - is_training=self.training, - ): - return None - - # Extract tensor properties for mask creation - dtype, device = input_tensor.dtype, input_tensor.device - min_dtype = torch.finfo(dtype).min - sequence_length = input_tensor.shape[1] - - # Determine target length based on cache type - if using_sliding_window_cache or using_static_cache: - # Use maximum cache shape for sliding window or static caches - target_length = past_key_values.get_max_cache_shape() + # Native Qwen uses a fourth, text-only position-id row to describe + # packed sequences. Keep the three multimodal rows for RoPE and pass + # the text row to both masking and decoder attention. + if position_ids.ndim == 3 and position_ids.shape[0] == 4: + text_position_ids = position_ids[0] + position_ids = position_ids[1:] else: - # For dynamic cache or no cache, calculate based on attention mask or sequence length - target_length = ( - attention_mask.shape[-1] - if isinstance(attention_mask, torch.Tensor) - else past_seen_tokens + sequence_length + 1 - ) + text_position_ids = None - # Generate 4D causal attention mask from 2D input mask if provided - causal_mask = self._prepare_4d_causal_attention_mask_with_cache_position( - attention_mask, - sequence_length=sequence_length, - target_length=target_length, - dtype=dtype, - device=device, - cache_position=cache_position, - batch_size=input_tensor.shape[0], - config=self.config, + full_token_types = self._prepare_action_token_types( + moe_token_types, + inputs_embeds=inputs_embeds, + attention_mask=attention_mask, past_key_values=past_key_values, ) + action_island_mask = self._action_island_mask(full_token_types) + mask_kwargs = { + "config": self.config, + "inputs_embeds": inputs_embeds, + "attention_mask": attention_mask, + "cache_position": cache_position, + "past_key_values": past_key_values, + "position_ids": text_position_ids, + "or_mask_function": action_island_mask, + } + causal_mask_mapping = {"full_attention": create_causal_mask(**mask_kwargs)} + if self.has_sliding_layers: + causal_mask_mapping["sliding_attention"] = create_sliding_window_causal_mask(**mask_kwargs) - # Modify mask to support bidirectional attention for specific token types - if moe_token_types is not None: - # Identify positions of type 1 tokens (MoE routing tokens) - type1_tokens = (moe_token_types == 1).unsqueeze(1).unsqueeze(2) # Shape: [B, 1, 1, S] + hidden_states = inputs_embeds + position_embeddings = self.rotary_emb(hidden_states, position_ids) - # Create bidirectional attention region for type 1 tokens - # This allows type 1 tokens to attend to each other bidirectionally - type1_mask = torch.zeros_like(causal_mask) # Shape: [B, num_heads, S, S] - type1_region = type1_tokens & type1_tokens.transpose(-1, -2) # Shape: [B, 1, S, S] - type1_mask = type1_mask.masked_fill(type1_region, 1.0).to(torch.bool) - - # Apply bidirectional attention: zero out causal constraints in type 1 regions - causal_mask = torch.where( - type1_mask, # Where type 1 tokens interact with each other - torch.zeros_like(causal_mask), # Remove causal masking (allow bidirectional) - causal_mask, # Keep original causal masking for other regions + for i, decoder_layer in enumerate(self.layers): + hidden_states = decoder_layer( + hidden_states, + attention_mask=causal_mask_mapping[self.config.layer_types[i]], + position_ids=text_position_ids, + past_key_values=past_key_values, + token_types=routing_token_types, + use_cache=use_cache, + position_embeddings=position_embeddings, + **kwargs, ) - # Handle special case for SDPA with CUDA/XPU devices - if ( - self.config._attn_implementation == "sdpa" - and attention_mask is not None - and attention_mask.device.type in ["cuda", "xpu"] - and not output_attentions - ): - # Ensure attention to all tokens in fully masked rows for memory-efficient attention - # This is required for F.scaled_dot_product_attention's memory-efficient path - # when using left padding. See: https://github.com/pytorch/pytorch/issues/110213 - causal_mask = AttentionMaskConverter._unmask_unattended(causal_mask, min_dtype) - - return causal_mask + hidden_states = self.norm(hidden_states) + return BaseModelOutputWithPast( + last_hidden_state=hidden_states, + past_key_values=past_key_values if use_cache else None, + ) @staticmethod - def _prepare_4d_causal_attention_mask_with_cache_position( - attention_mask: torch.Tensor, - sequence_length: int, - target_length: int, - dtype: torch.dtype, - device: torch.device, - cache_position: torch.Tensor, - batch_size: int, - config: Qwen2_5_VLConfig, - past_key_values: Cache, - ): + def _action_island_mask(token_types: torch.Tensor): + action_tokens = token_types == 1 + + def action_island(batch_idx, head_idx, query_idx, key_idx): + del head_idx + return action_tokens[batch_idx, query_idx] & action_tokens[batch_idx, key_idx] + + return action_island + + @staticmethod + def _prepare_action_token_types( + token_types: torch.Tensor, + *, + inputs_embeds: torch.Tensor, + attention_mask: torch.Tensor | None, + past_key_values: Cache | None, + ) -> torch.Tensor: + """Align current-step token types with absolute mask indices. + + Generation passes token types only for the new query tokens, while the + native mask callbacks receive absolute query/key indices. Cached tokens + default to type 0; callers may instead pass full-history token types. """ - Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape - `(batch_size, key_value_length)`, or if the input `attention_mask` is already 4D, do nothing. + query_length = inputs_embeds.shape[1] + past_length = past_key_values.get_seq_length() if past_key_values is not None else 0 + if isinstance(past_length, torch.Tensor): + past_length = int(past_length.item()) - Args: - attention_mask (`torch.Tensor`): - A 2D attention mask of shape `(batch_size, key_value_length)` or a 4D attention mask of shape `(batch_size, 1, query_length, key_value_length)`. - sequence_length (`int`): - The sequence length being processed. - target_length (`int`): - The target length: when generating with static cache, the mask should be as long as the static cache, to account for the 0 padding, the part of the cache that is not filled yet. - dtype (`torch.dtype`): - The dtype to use for the 4D attention mask. - device (`torch.device`): - The device to place the 4D attention mask on. - cache_position (`torch.Tensor`): - Indices depicting the position of the input sequence tokens in the sequence. - batch_size (`torch.Tensor`): - Batch size. - config (`Qwen2_5_VLConfig`): - The model's configuration class - past_key_values (`Cache`): - The cache class that is being used currently to generate - """ - if attention_mask is not None and attention_mask.dim() == 4: - # In this case we assume that the mask comes already in inverted form and requires no inversion or slicing. - causal_mask = attention_mask - else: - min_dtype = torch.finfo(dtype).min - causal_mask = torch.full( - (sequence_length, target_length), - fill_value=min_dtype, - dtype=dtype, - device=device, - ) - diagonal_attend_mask = torch.arange(target_length, device=device) > cache_position.reshape(-1, 1) - if config.sliding_window is not None: - # if we have sliding window, we should not attend to tokens beyond sliding window length, so we mask them out also - # the check is needed to verify is current checkpoint was trained with sliding window or not - if not isinstance(past_key_values, SlidingWindowCache) or sequence_length > target_length: - sliding_attend_mask = torch.arange(target_length, device=device) <= ( - cache_position.reshape(-1, 1) - config.sliding_window - ) - diagonal_attend_mask.bitwise_or_(sliding_attend_mask) - causal_mask *= diagonal_attend_mask - causal_mask = causal_mask[None, None, :, :].expand(batch_size, 1, -1, -1) - if attention_mask is not None: - causal_mask = causal_mask.clone() # copy to contiguous memory for in-place edit - if attention_mask.shape[-1] > target_length: - attention_mask = attention_mask[:, :target_length] - mask_length = attention_mask.shape[-1] - padding_mask = causal_mask[:, :, :, :mask_length] + attention_mask[:, None, None, :].to( - causal_mask.device - ) - padding_mask = padding_mask == 0 - causal_mask[:, :, :, :mask_length] = causal_mask[:, :, :, :mask_length].masked_fill( - padding_mask, min_dtype - ) - return causal_mask + token_types = token_types.to(device=inputs_embeds.device) + if token_types.shape[-1] == query_length and past_length: + token_types = torch.nn.functional.pad(token_types, (past_length, 0)) + required_length = past_length + query_length + if attention_mask is not None and attention_mask.ndim == 2: + required_length = max(required_length, attention_mask.shape[-1]) + if past_key_values is not None: + kv_length, kv_offset = past_key_values.get_mask_sizes(query_length, 0) + if isinstance(kv_length, torch.Tensor): + kv_length = int(kv_length.item()) + if isinstance(kv_offset, torch.Tensor): + kv_offset = int(kv_offset.item()) + required_length = max(required_length, kv_length + kv_offset) -__all__ = [ - "Qwen2_5_VLForConditionalGeneration", - "Qwen2_5_VLModel", - "Qwen2_5_VLPreTrainedModel", - "Qwen2_5_VLDecoderLayer_with_MoE", - "Qwen2_5_VLMoEModel", -] + if token_types.shape[-1] < required_length: + token_types = torch.nn.functional.pad(token_types, (0, required_length - token_types.shape[-1])) + return token_types diff --git a/src/lerobot/policies/wall_x/qwen_model/vision_attention.py b/src/lerobot/policies/wall_x/qwen_model/vision_attention.py new file mode 100644 index 000000000..f8c37aa6c --- /dev/null +++ b/src/lerobot/policies/wall_x/qwen_model/vision_attention.py @@ -0,0 +1,208 @@ +#!/usr/bin/env python + +# Copyright 2025 HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Wall-X vision attention backends. + +Qwen2.5-VL's native non-Flash vision path splits a packed image sequence into +Python-level chunks before calling attention. Wall-X batches many camera frames, +so that path launches thousands of tiny attention operations per training step. +This module keeps the native SDPA path as a portable fallback and adds a packed +``torch.nn.attention.varlen`` path that consumes Qwen's existing ``cu_seqlens`` +metadata directly. +""" + +from __future__ import annotations + +import inspect +import logging +from functools import lru_cache +from typing import TYPE_CHECKING, Literal + +import torch +import torch.nn as nn + +from lerobot.utils.import_utils import _transformers_available + +if TYPE_CHECKING or _transformers_available: + from transformers.models.qwen2_5_vl.modeling_qwen2_5_vl import ( + Qwen2_5_VLVisionAttention, + apply_rotary_pos_emb_vision, + ) +else: + Qwen2_5_VLVisionAttention = nn.Module + apply_rotary_pos_emb_vision = None + +try: + from torch.nn.attention.varlen import varlen_attn as _varlen_attn +except ImportError: # torch<2.10 + _varlen_attn = None + +_VARLEN_USES_WINDOW_SIZE = ( + _varlen_attn is not None and "window_size" in inspect.signature(_varlen_attn).parameters +) + + +VisionAttentionBackend = Literal["auto", "sdpa", "varlen"] + +logger = logging.getLogger(__name__) + + +@lru_cache +def _log_resolved_backend(requested: str, resolved: str) -> None: + logger.info("Wall-X vision attention backend: %s (requested: %s)", resolved, requested) + + +def _varlen_unavailable_reason( + hidden_states: torch.Tensor, + position_embeddings: tuple[torch.Tensor, torch.Tensor] | None, +) -> str | None: + if _varlen_attn is None: + return "torch.nn.attention.varlen is unavailable (PyTorch 2.10 or newer is required)" + if position_embeddings is None: + return "precomputed vision position embeddings were not provided" + if hidden_states.device.type != "cuda" or torch.version.cuda is None: + return "packed varlen attention requires an NVIDIA CUDA device" + if hidden_states.dtype not in {torch.float16, torch.bfloat16}: + return f"packed varlen attention requires float16 or bfloat16 inputs, got {hidden_states.dtype}" + major, _minor = torch.cuda.get_device_capability(hidden_states.device) + if major < 8: + return "packed varlen attention requires an NVIDIA Ampere GPU or newer" + return None + + +def _supports_varlen_attention( + hidden_states: torch.Tensor, + position_embeddings: tuple[torch.Tensor, torch.Tensor] | None, +) -> bool: + return _varlen_unavailable_reason(hidden_states, position_embeddings) is None + + +class WallXVisionAttention(Qwen2_5_VLVisionAttention): + """Qwen2.5-VL vision attention with packed varlen and native SDPA fallback.""" + + def __init__(self, config, backend: VisionAttentionBackend): + super().__init__(config) + self.wallx_backend = backend + self._resolved_backend_key = None + self._resolved_backend = None + + def _resolve_backend( + self, + hidden_states: torch.Tensor, + position_embeddings: tuple[torch.Tensor, torch.Tensor] | None, + ) -> str: + key = ( + hidden_states.device.type, + hidden_states.device.index, + hidden_states.dtype, + position_embeddings is not None, + ) + if self._resolved_backend_key == key: + return self._resolved_backend + + use_varlen = self.wallx_backend != "sdpa" and _supports_varlen_attention( + hidden_states, position_embeddings + ) + if self.wallx_backend == "varlen" and not use_varlen: + reason = _varlen_unavailable_reason(hidden_states, position_embeddings) + raise RuntimeError(f"Wall-X vision_attn_implementation='varlen' cannot be used: {reason}") + + resolved_backend = "varlen" if use_varlen else "sdpa" + self._resolved_backend_key = key + self._resolved_backend = resolved_backend + _log_resolved_backend(self.wallx_backend, resolved_backend) + return resolved_backend + + def forward( + self, + hidden_states: torch.Tensor, + cu_seqlens: torch.Tensor, + rotary_pos_emb: torch.Tensor | None = None, + position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None, + **kwargs, + ) -> torch.Tensor: + del rotary_pos_emb + + if self._resolve_backend(hidden_states, position_embeddings) == "sdpa": + return super().forward( + hidden_states=hidden_states, + cu_seqlens=cu_seqlens, + position_embeddings=position_embeddings, + **kwargs, + ) + + seq_length = hidden_states.shape[0] + query_states, key_states, value_states = ( + self.qkv(hidden_states).reshape(seq_length, 3, self.num_heads, -1).permute(1, 0, 2, 3).unbind(0) + ) + + cos, sin = position_embeddings + query_states, key_states = apply_rotary_pos_emb_vision( + query_states, + key_states, + cos, + sin, + ) + + if cu_seqlens.dtype != torch.int32: + cu_seqlens = cu_seqlens.to(dtype=torch.int32) + max_seqlen = int((cu_seqlens[1:] - cu_seqlens[:-1]).max().item()) + varlen_kwargs = {"scale": self.scaling} + if _VARLEN_USES_WINDOW_SIZE: + varlen_kwargs["window_size"] = (-1, -1) + else: # Stable PyTorch 2.10 API; pre-release variants used window_size. + varlen_kwargs["is_causal"] = False + attn_output = _varlen_attn( + query_states, + key_states, + value_states, + cu_seqlens, + cu_seqlens, + max_seqlen, + max_seqlen, + **varlen_kwargs, + ) + attn_output = attn_output.reshape(seq_length, -1).contiguous() + return self.proj(attn_output) + + +def configure_wall_x_vision_attention( + vision_model: nn.Module, + backend: VisionAttentionBackend, +) -> None: + """Install Wall-X's scoped packed attention without changing checkpoint keys.""" + if backend == "sdpa": + _log_resolved_backend(backend, "sdpa") + return + if backend == "varlen" and _varlen_attn is None: + raise RuntimeError( + "Wall-X vision_attn_implementation='varlen' requires torch.nn.attention.varlen " + "from PyTorch 2.10 or newer" + ) + if backend == "auto" and _varlen_attn is None: + _log_resolved_backend(backend, "sdpa") + return + + for block in vision_model.blocks: + previous_attention = block.attn + replacement = WallXVisionAttention(previous_attention.config, backend=backend) + replacement.to( + device=previous_attention.qkv.weight.device, + dtype=previous_attention.qkv.weight.dtype, + ) + replacement.load_state_dict(previous_attention.state_dict(), strict=True) + replacement.train(previous_attention.training) + block.attn = replacement diff --git a/src/lerobot/policies/wall_x/utils.py b/src/lerobot/policies/wall_x/utils.py index d38a2d509..495b85cd3 100644 --- a/src/lerobot/policies/wall_x/utils.py +++ b/src/lerobot/policies/wall_x/utils.py @@ -116,6 +116,7 @@ def preprocesser_call( images: list | Any | None = None, text: str | list[str] | None = None, videos: list | Any | None = None, + device: torch.device | str | None = None, padding: bool | str = False, truncation: bool | None = None, max_length: int | None = None, @@ -134,6 +135,7 @@ def preprocesser_call( images: Input images (PIL, numpy arrays, or torch tensors) text: Text or list of texts to tokenize videos: Input videos (numpy arrays or torch tensors) + device: Device on which image/video preprocessing should run padding: Whether to pad sequences to same length truncation: Whether to truncate sequences longer than max_length max_length: Maximum length for truncation/padding @@ -151,7 +153,11 @@ def preprocesser_call( """ # Process image inputs if images is not None and len(images) > 0: - image_inputs = processor.image_processor(images=images, return_tensors=return_tensors) + image_inputs = processor.image_processor( + images=images, + return_tensors=return_tensors, + device=device, + ) image_grid_thw = image_inputs["image_grid_thw"] else: image_inputs = {} @@ -159,7 +165,11 @@ def preprocesser_call( # Process video inputs if videos is not None: - videos_inputs = processor.image_processor(videos=videos, return_tensors=return_tensors) + videos_inputs = processor.image_processor( + videos=videos, + return_tensors=return_tensors, + device=device, + ) video_grid_thw = videos_inputs["video_grid_thw"] else: videos_inputs = {} @@ -413,10 +423,7 @@ def get_task_instruction( } ) - if priority_order is not None: - priority_order = OrderedDict(priority_order) - else: - priority_order = default_priority_order + priority_order = OrderedDict(priority_order) if priority_order is not None else default_priority_order got_instruction = False task_instruction = "" @@ -424,9 +431,8 @@ def get_task_instruction( # Sample instruction components based on priority probabilities for key, prob in priority_order.items(): if key in frame_instruction_info and frame_instruction_info[key] != "": - if got_instruction: - if random.random() >= prob: - continue + if got_instruction and random.random() >= prob: + continue task_instruction += f"\n{frame_instruction_info[key]}" got_instruction = True @@ -538,10 +544,7 @@ def img_key_mapping(img_keys: list[str]) -> list[str]: if key in CAMERA_NAME_MAPPING: key = CAMERA_NAME_MAPPING[key] else: - if "view" in key: - key = key.replace("_", " ") - else: - key = key + " view" + key = key.replace("_", " ") if "view" in key else key + " view" processed_img_keys.append(key) return processed_img_keys diff --git a/src/lerobot/policies/xvla/processor_xvla.py b/src/lerobot/policies/xvla/processor_xvla.py index 0336ec722..02c54f9c9 100644 --- a/src/lerobot/policies/xvla/processor_xvla.py +++ b/src/lerobot/policies/xvla/processor_xvla.py @@ -22,19 +22,14 @@ import torch from lerobot.configs import PipelineFeatureType, PolicyFeature from lerobot.processor import ( - AddBatchDimensionProcessorStep, - DeviceProcessorStep, - NormalizerProcessorStep, ObservationProcessorStep, PolicyAction, PolicyProcessorPipeline, ProcessorStep, ProcessorStepRegistry, - RenameObservationsProcessorStep, TokenizerProcessorStep, - UnnormalizerProcessorStep, - policy_action_to_transition, - transition_to_policy_action, + make_default_policy_processor_steps, + make_policy_processor_pipelines, ) from lerobot.types import EnvTransition, TransitionKey from lerobot.utils.constants import ( @@ -42,8 +37,6 @@ from lerobot.utils.constants import ( OBS_IMAGES, OBS_PREFIX, OBS_STATE, - POLICY_POSTPROCESSOR_DEFAULT_NAME, - POLICY_PREPROCESSOR_DEFAULT_NAME, ) from .configuration_xvla import XVLAConfig @@ -61,10 +54,11 @@ def make_xvla_pre_post_processors( Build the LeRobot processor pipelines for XVLA. """ - features = {**config.input_features, **config.output_features} + steps = make_default_policy_processor_steps(config, dataset_stats) + input_steps = [ - RenameObservationsProcessorStep(rename_map={}), - AddBatchDimensionProcessorStep(), + steps.rename_observations, + steps.add_batch_dim, TokenizerProcessorStep( tokenizer_name=config.tokenizer_name, max_length=config.tokenizer_max_length, @@ -74,32 +68,15 @@ def make_xvla_pre_post_processors( XVLAImageToFloatProcessorStep(), XVLAImageNetNormalizeProcessorStep(), XVLAAddDomainIdProcessorStep(), - DeviceProcessorStep(device=config.device), - NormalizerProcessorStep( - features=features, norm_map=config.normalization_mapping, stats=dataset_stats - ), + steps.to_device, + steps.normalize, ] output_steps = [ - UnnormalizerProcessorStep( - features=config.output_features, - norm_map=config.normalization_mapping, - stats=dataset_stats, - ), - DeviceProcessorStep(device="cpu"), + steps.unnormalize, + steps.to_cpu, ] - return ( - PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( - steps=input_steps, - name=POLICY_PREPROCESSOR_DEFAULT_NAME, - ), - PolicyProcessorPipeline[PolicyAction, PolicyAction]( - steps=output_steps, - name=POLICY_POSTPROCESSOR_DEFAULT_NAME, - to_transition=policy_action_to_transition, - to_output=transition_to_policy_action, - ), - ) + return make_policy_processor_pipelines(input_steps=input_steps, output_steps=output_steps) # Custom XVLA processor steps diff --git a/src/lerobot/processor/__init__.py b/src/lerobot/processor/__init__.py index fe35af4b4..79bb44c85 100644 --- a/src/lerobot/processor/__init__.py +++ b/src/lerobot/processor/__init__.py @@ -42,10 +42,14 @@ from .delta_action_processor import MapDeltaActionToRobotActionStep, MapTensorTo from .device_processor import DeviceProcessorStep from .env_processor import IsaaclabArenaProcessorStep, LiberoProcessorStep from .factory import ( + DefaultPolicyProcessorSteps, + make_default_policy_processor_steps, + make_default_pre_post_processors, make_default_processors, make_default_robot_action_processor, make_default_robot_observation_processor, make_default_teleop_action_processor, + make_policy_processor_pipelines, ) from .gym_action_processor import ( Numpy2TorchActionProcessorStep, @@ -129,10 +133,14 @@ __all__ = [ "ImageCropResizeProcessorStep", "InfoProcessorStep", "InterventionActionProcessorStep", + "DefaultPolicyProcessorSteps", + "make_default_policy_processor_steps", + "make_default_pre_post_processors", "make_default_processors", "make_default_teleop_action_processor", "make_default_robot_action_processor", "make_default_robot_observation_processor", + "make_policy_processor_pipelines", "AbsoluteActionsProcessorStep", "RelativeActionsProcessorStep", "MapDeltaActionToRobotActionStep", diff --git a/src/lerobot/processor/factory.py b/src/lerobot/processor/factory.py index 5028122f1..ec929f365 100644 --- a/src/lerobot/processor/factory.py +++ b/src/lerobot/processor/factory.py @@ -14,15 +14,33 @@ # See the License for the specific language governing permissions and # limitations under the License. -from lerobot.types import RobotAction, RobotObservation +from dataclasses import dataclass +from typing import Any +import torch + +from lerobot.configs.policies import PreTrainedConfig +from lerobot.types import PolicyAction, RobotAction, RobotObservation +from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME + +from .batch_processor import AddBatchDimensionProcessorStep from .converters import ( observation_to_transition, + policy_action_to_transition, robot_action_observation_to_transition, transition_to_observation, + transition_to_policy_action, transition_to_robot_action, ) -from .pipeline import IdentityProcessorStep, RobotProcessorPipeline +from .device_processor import DeviceProcessorStep +from .normalize_processor import NormalizerProcessorStep, UnnormalizerProcessorStep +from .pipeline import ( + IdentityProcessorStep, + PolicyProcessorPipeline, + ProcessorStep, + RobotProcessorPipeline, +) +from .rename_processor import RenameObservationsProcessorStep def make_default_teleop_action_processor() -> RobotProcessorPipeline[ @@ -61,3 +79,97 @@ def make_default_processors(): robot_action_processor = make_default_robot_action_processor() robot_observation_processor = make_default_robot_observation_processor() return (teleop_action_processor, robot_action_processor, robot_observation_processor) + + +@dataclass +class DefaultPolicyProcessorSteps: + """The canonical processor steps shared by most policies' pre/post pipelines. + + Policies compose these in their own order (step ORDER is a Hub-serialized contract + and intentionally stays explicit per policy) and interleave their custom steps. + """ + + rename_observations: RenameObservationsProcessorStep + add_batch_dim: AddBatchDimensionProcessorStep + to_device: DeviceProcessorStep + normalize: NormalizerProcessorStep + unnormalize: UnnormalizerProcessorStep + to_cpu: DeviceProcessorStep + + +def make_default_policy_processor_steps( + config: PreTrainedConfig, + dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None, + *, + normalizer_device: torch.device | str | None = None, +) -> DefaultPolicyProcessorSteps: + """Construct the canonical policy processor steps from a policy config. + + Args: + config: A `PreTrainedConfig` providing `device`, `input_features`, + `output_features` and `normalization_mapping`. + dataset_stats: Dataset statistics used for (un)normalization. + normalizer_device: Device passed to `NormalizerProcessorStep` (some policies pin + their normalization stats to the policy device; most leave it unset). + """ + return DefaultPolicyProcessorSteps( + rename_observations=RenameObservationsProcessorStep(rename_map={}), + add_batch_dim=AddBatchDimensionProcessorStep(), + to_device=DeviceProcessorStep(device=config.device), + normalize=NormalizerProcessorStep( + features={**config.input_features, **config.output_features}, + norm_map=config.normalization_mapping, + stats=dataset_stats, + device=normalizer_device, + ), + unnormalize=UnnormalizerProcessorStep( + features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats + ), + to_cpu=DeviceProcessorStep(device="cpu"), + ) + + +def make_policy_processor_pipelines( + input_steps: list[ProcessorStep], + output_steps: list[ProcessorStep], +) -> tuple[ + PolicyProcessorPipeline[dict[str, Any], dict[str, Any]], + PolicyProcessorPipeline[PolicyAction, PolicyAction], +]: + """Wrap pre/post step lists into the canonical policy pipeline pair. + + Uses the standard pipeline names (which determine the serialized JSON filenames on + the Hub) and the standard policy-action converters on the postprocessor. + """ + return ( + PolicyProcessorPipeline[dict[str, Any], dict[str, Any]]( + steps=input_steps, + name=POLICY_PREPROCESSOR_DEFAULT_NAME, + ), + PolicyProcessorPipeline[PolicyAction, PolicyAction]( + steps=output_steps, + name=POLICY_POSTPROCESSOR_DEFAULT_NAME, + to_transition=policy_action_to_transition, + to_output=transition_to_policy_action, + ), + ) + + +def make_default_pre_post_processors( + config: PreTrainedConfig, + dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None, + *, + normalizer_device: torch.device | str | None = None, +) -> tuple[ + PolicyProcessorPipeline[dict[str, Any], dict[str, Any]], + PolicyProcessorPipeline[PolicyAction, PolicyAction], +]: + """The pure-scaffold policy pipeline pair: Rename -> Batch -> Device -> Normalize, + and Unnormalize -> Device(cpu). Policies with custom steps or a different step order + compose `make_default_policy_processor_steps` themselves instead. + """ + s = make_default_policy_processor_steps(config, dataset_stats, normalizer_device=normalizer_device) + return make_policy_processor_pipelines( + input_steps=[s.rename_observations, s.add_batch_dim, s.to_device, s.normalize], + output_steps=[s.unnormalize, s.to_cpu], + ) diff --git a/src/lerobot/rewards/pretrained.py b/src/lerobot/rewards/pretrained.py index ed4205bcc..185e66a3b 100644 --- a/src/lerobot/rewards/pretrained.py +++ b/src/lerobot/rewards/pretrained.py @@ -28,6 +28,7 @@ from safetensors.torch import load_model as load_model_as_safetensor, save_model from torch import Tensor, nn from lerobot.configs.rewards import RewardModelConfig +from lerobot.utils.device_utils import resolve_safetensors_device from lerobot.utils.hub import HubMixin if TYPE_CHECKING: @@ -128,7 +129,7 @@ class PreTrainedRewardModel(nn.Module, HubMixin, abc.ABC): @classmethod def _load_as_safetensor(cls, model: T, model_file: str, map_location: str, strict: bool) -> T: missing_keys, unexpected_keys = load_model_as_safetensor( - model, model_file, strict=strict, device=map_location + model, model_file, strict=strict, device=resolve_safetensors_device(map_location) ) if missing_keys: logging.warning(f"Missing key(s) when loading model: {missing_keys}") diff --git a/src/lerobot/scripts/lerobot_annotate.py b/src/lerobot/scripts/lerobot_annotate.py index 579175052..41fede908 100644 --- a/src/lerobot/scripts/lerobot_annotate.py +++ b/src/lerobot/scripts/lerobot_annotate.py @@ -28,7 +28,12 @@ For distributed runs, see ``examples/annotations/run_hf_job.py``. """ import logging +from contextlib import suppress from pathlib import Path +from typing import TYPE_CHECKING + +from huggingface_hub import HfApi, snapshot_download +from huggingface_hub.errors import RevisionNotFoundError from lerobot.annotations.steerable_pipeline.config import AnnotationPipelineConfig from lerobot.annotations.steerable_pipeline.executor import Executor @@ -42,6 +47,12 @@ from lerobot.annotations.steerable_pipeline.validator import StagingValidator from lerobot.annotations.steerable_pipeline.vlm_client import make_vlm_client from lerobot.annotations.steerable_pipeline.writer import LanguageColumnsWriter from lerobot.configs import parser +from lerobot.utils.import_utils import _datasets_available, require_package + +if TYPE_CHECKING or _datasets_available: + from lerobot.datasets.dataset_metadata import CODEBASE_VERSION + from lerobot.datasets.io_utils import load_info + from lerobot.datasets.utils import create_lerobot_dataset_card logger = logging.getLogger(__name__) @@ -50,8 +61,6 @@ def _resolve_root(cfg: AnnotationPipelineConfig) -> Path: if cfg.root is not None: return Path(cfg.root) if cfg.repo_id is not None: - from huggingface_hub import snapshot_download - return Path(snapshot_download(repo_id=cfg.repo_id, repo_type="dataset")) raise ValueError("Either --root or --repo_id must be provided.") @@ -125,10 +134,7 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None: Pushes to ``cfg.new_repo_id`` when set, otherwise back to ``cfg.repo_id``. """ - from huggingface_hub import HfApi # noqa: PLC0415 - - from lerobot.datasets.io_utils import load_info # noqa: PLC0415 - from lerobot.datasets.utils import create_lerobot_dataset_card # noqa: PLC0415 + require_package("datasets", "dataset") repo_id = cfg.new_repo_id or cfg.repo_id commit_message = cfg.push_commit_message or "Add steerable annotations (lerobot-annotate)" @@ -163,8 +169,6 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None: # ``RevisionNotFoundError``. Read the version straight from the # dataset's own ``meta/info.json`` so we tag whatever the writer # actually wrote (no accidental drift if the codebase floor moves). - from lerobot.datasets.dataset_metadata import CODEBASE_VERSION # noqa: PLC0415 - version_tag = ( dataset_info.codebase_version if dataset_info.codebase_version.startswith("v") else CODEBASE_VERSION ) @@ -178,10 +182,6 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None: tag_kwargs["revision"] = revision try: - from contextlib import suppress # noqa: PLC0415 - - from huggingface_hub.errors import RevisionNotFoundError # noqa: PLC0415 - with suppress(RevisionNotFoundError): api.delete_tag(repo_id, tag=version_tag, repo_type="dataset") api.create_tag(**tag_kwargs) diff --git a/src/lerobot/utils/device_utils.py b/src/lerobot/utils/device_utils.py index 37981f07f..3f9b58773 100644 --- a/src/lerobot/utils/device_utils.py +++ b/src/lerobot/utils/device_utils.py @@ -59,6 +59,20 @@ def get_safe_torch_device(try_device: str, log: bool = False) -> torch.device: return device +def resolve_safetensors_device(map_location: str | torch.device) -> str: + """Resolve a device string for a safetensors load, working around a device-mapping quirk. + + safetensors' load maps the bare string "cuda" to cuda:0 regardless of the current device + (unlike torch's .to("cuda"), which honors torch.cuda.current_device()). Under multi-GPU + accelerate/FSDP every rank would then load its weights onto GPU 0, OOMing it before sharding. + Resolve "cuda" to the concrete current-device index so each rank loads onto its own GPU. + """ + map_location = str(map_location) + if map_location == "cuda" and torch.cuda.is_available(): + return f"cuda:{torch.cuda.current_device()}" + return map_location + + def get_safe_dtype(dtype: torch.dtype, device: str | torch.device): """ mps is currently not compatible with float64 diff --git a/tests/policies/common/test_flow_matching.py b/tests/policies/common/test_flow_matching.py new file mode 100644 index 000000000..065883ebc --- /dev/null +++ b/tests/policies/common/test_flow_matching.py @@ -0,0 +1,193 @@ +#!/usr/bin/env python + +# Copyright 2026 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Behavior-pinning tests for the shared flow-matching sampling primitives. + +``euler_integrate`` is compared against a verbatim copy of the historical pi0/pi05/ +smolvla sampling loop (including its RTC hook semantics): any divergence from that +reference is a behavior change for released checkpoints. +""" + +import torch + +from lerobot.policies.common.flow_matching import ( + euler_integrate, + sample_beta, + sample_noise, + sample_time_beta, +) + + +def test_sample_beta_range_dtype_and_reproducibility(): + torch.manual_seed(0) + s1 = sample_beta(1.5, 1.0, 4096, "cpu") + torch.manual_seed(0) + s2 = sample_beta(1.5, 1.0, 4096, "cpu") + assert torch.equal(s1, s2) + assert s1.shape == (4096,) and s1.dtype == torch.float32 + assert s1.min() >= 0.0 and s1.max() <= 1.0 + # Beta(1.5, 1.0) mean is 1.5/2.5 = 0.6. + assert abs(s1.mean().item() - 0.6) < 0.02 + + +def test_sample_time_beta_openpi_convention(): + torch.manual_seed(1) + time = sample_time_beta(4096, "cpu", alpha=1.5, beta=1.0, scale=0.999, offset=0.001) + assert time.dtype == torch.float32 + assert time.min() >= 0.001 and time.max() <= 1.0 + # Exact composition: Beta sample * scale + offset, same RNG stream. + torch.manual_seed(1) + expected = sample_beta(1.5, 1.0, 4096, "cpu") * 0.999 + 0.001 + torch.testing.assert_close(time, expected, rtol=0, atol=0) + + +def test_sample_noise_seeded(): + torch.manual_seed(2) + n1 = sample_noise((2, 8, 4), "cpu") + torch.manual_seed(2) + n2 = sample_noise((2, 8, 4), "cpu") + assert torch.equal(n1, n2) + assert n1.dtype == torch.float32 and n1.shape == (2, 8, 4) + + +def test_euler_integrate_constant_velocity_is_exact(): + # With v_t == c constant, x_0 = x_1 + sum(dt * c) = x_1 - c exactly (num_steps * dt = -1). + noise = torch.randn(3, 5, 2) + c = torch.randn(3, 5, 2) + out = euler_integrate(lambda x_t, time: c, noise, num_steps=10) + torch.testing.assert_close(out, noise - c, rtol=0, atol=1e-6) + + +def _reference_pi0_loop(denoise_fn, noise, num_steps, rtc_enabled, rtc_processor, kw): + """Verbatim structure of the historical pi0/pi05/smolvla sample_actions loop.""" + bsize = noise.shape[0] + device = noise.device + dt = -1.0 / num_steps + x_t = noise + for step in range(num_steps): + time = 1.0 + step * dt + time_tensor = torch.tensor(time, dtype=torch.float32, device=device).expand(bsize) + + def denoise_step_partial_call(input_x_t, current_timestep=time_tensor): + return denoise_fn(input_x_t, current_timestep) + + if rtc_enabled: + v_t = rtc_processor.denoise_step( + x_t=x_t, + prev_chunk_left_over=kw.get("prev_chunk_left_over"), + inference_delay=kw.get("inference_delay"), + time=time, + original_denoise_step_partial=denoise_step_partial_call, + execution_horizon=kw.get("execution_horizon"), + ) + else: + v_t = denoise_step_partial_call(x_t) + x_t = x_t + dt * v_t + if rtc_processor is not None and rtc_processor.is_debug_enabled(): + rtc_processor.track(time=time, x_t=x_t, v_t=v_t) + return x_t + + +class _StubRTCProcessor: + def __init__(self, debug_enabled: bool): + self._debug = debug_enabled + self.tracked = [] + self.guidance_calls = [] + + def is_debug_enabled(self): + return self._debug + + def denoise_step( + self, + x_t, + prev_chunk_left_over, + inference_delay, + time, + original_denoise_step_partial, + execution_horizon, + ): + self.guidance_calls.append( + { + "time": time, + "inference_delay": inference_delay, + "execution_horizon": execution_horizon, + "x_t": x_t.clone(), + } + ) + return original_denoise_step_partial(x_t) * 0.5 + + def track(self, time, x_t, v_t): + self.tracked.append({"time": time, "x_t": x_t.clone(), "v_t": v_t.clone()}) + + +def _make_denoise_fn(): + weight = torch.randn(4, 4) * 0.1 + + def denoise_fn(x_t, time_tensor): + return x_t @ weight + time_tensor[:, None, None] + + return denoise_fn + + +def test_euler_integrate_matches_historical_loop(): + torch.manual_seed(3) + denoise_fn = _make_denoise_fn() + noise = torch.randn(2, 6, 4) + ref = _reference_pi0_loop(denoise_fn, noise, 10, rtc_enabled=False, rtc_processor=None, kw={}) + out = euler_integrate(denoise_fn, noise, 10) + assert torch.equal(out, ref) + + +def test_euler_integrate_rtc_guidance_and_kwarg_forwarding(): + torch.manual_seed(4) + denoise_fn = _make_denoise_fn() + noise = torch.randn(2, 6, 4) + leftover = torch.randn(2, 6, 4) + kw = {"inference_delay": 3, "prev_chunk_left_over": leftover, "execution_horizon": 25} + + ref_proc, new_proc = _StubRTCProcessor(False), _StubRTCProcessor(False) + ref = _reference_pi0_loop(denoise_fn, noise, 6, rtc_enabled=True, rtc_processor=ref_proc, kw=kw) + out = euler_integrate( + denoise_fn, + noise, + 6, + rtc_processor=new_proc, + rtc_enabled=True, + inference_delay=3, + prev_chunk_left_over=leftover, + execution_horizon=25, + ) + assert torch.equal(out, ref) + assert len(new_proc.guidance_calls) == 6 + for ref_call, new_call in zip(ref_proc.guidance_calls, new_proc.guidance_calls, strict=True): + assert ref_call["time"] == new_call["time"] + assert new_call["inference_delay"] == 3 and new_call["execution_horizon"] == 25 + # Guidance sees the PRE-update x_t. + assert torch.equal(ref_call["x_t"], new_call["x_t"]) + + +def test_euler_integrate_debug_tracking_fires_even_when_rtc_disabled(): + # Historical behavior: track() fires whenever the processor exists and has debugging + # enabled, independent of whether RTC guidance is active. + torch.manual_seed(5) + denoise_fn = _make_denoise_fn() + noise = torch.randn(2, 6, 4) + proc = _StubRTCProcessor(True) + out = euler_integrate(denoise_fn, noise, 4, rtc_processor=proc, rtc_enabled=False) + assert len(proc.guidance_calls) == 0 + assert len(proc.tracked) == 4 + # track() receives the POST-update x_t; the last one is the returned sample. + assert torch.equal(proc.tracked[-1]["x_t"], out) diff --git a/tests/policies/common/test_vla_utils.py b/tests/policies/common/test_vla_utils.py new file mode 100644 index 000000000..4d57b15ee --- /dev/null +++ b/tests/policies/common/test_vla_utils.py @@ -0,0 +1,195 @@ +#!/usr/bin/env python + +# Copyright 2026 The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Behavior-pinning tests for the shared VLA helpers. + +These helpers are the canonical versions of functions that used to be copy-pasted across +the openpi-derived policies (pi0, pi05, pi0_fast, smolvla, eo1, xvla). The expected +values below encode the historical per-policy behavior exactly; a failure here means a +behavior change that would silently affect released checkpoints. +""" + +import math + +import pytest +import torch + +from lerobot.policies.common.vla_utils import ( + create_sinusoidal_pos_embedding, + make_att_2d_masks, + pad_vector, + prepare_attention_masks_4d, + resize_with_pad, + resize_with_pad_torch, +) +from lerobot.utils.constants import OPENPI_ATTENTION_MASK_VALUE + + +def test_create_sinusoidal_pos_embedding_matches_openpi_formula(): + time = torch.tensor([0.0, 0.25, 1.0]) + dim, min_period, max_period = 8, 4e-3, 4.0 + emb = create_sinusoidal_pos_embedding(time, dim, min_period, max_period, device=torch.device("cpu")) + + assert emb.shape == (3, dim) + # Independent recomputation of the openpi formula in float64. + fraction = torch.linspace(0.0, 1.0, dim // 2, dtype=torch.float64) + period = min_period * (max_period / min_period) ** fraction + scaling = 1.0 / period * 2 * math.pi + sin_input = scaling[None, :] * time.to(torch.float64)[:, None] + expected = torch.cat([torch.sin(sin_input), torch.cos(sin_input)], dim=1) + torch.testing.assert_close(emb, expected, rtol=1e-9, atol=1e-9) + + +def test_create_sinusoidal_pos_embedding_validation(): + with pytest.raises(ValueError, match="divisible by 2"): + create_sinusoidal_pos_embedding(torch.zeros(2), 7, 4e-3, 4.0, device=torch.device("cpu")) + with pytest.raises(ValueError, match="batch_size"): + create_sinusoidal_pos_embedding(torch.zeros(2, 2), 8, 4e-3, 4.0, device=torch.device("cpu")) + + +def test_make_att_2d_masks_docstring_cases(): + # Pure causal attention: [[1 1 1]] + pad = torch.ones(1, 3, dtype=torch.bool) + att = torch.tensor([[1, 1, 1]], dtype=torch.int32) + expected = torch.tensor([[[1, 0, 0], [1, 1, 0], [1, 1, 1]]], dtype=torch.bool) + assert torch.equal(make_att_2d_masks(pad, att), expected) + + # Prefix-LM: [[0 0 1 1]] -> first two tokens attend bidirectionally, rest causal. + att = torch.tensor([[0, 0, 1, 1]], dtype=torch.int32) + pad = torch.ones(1, 4, dtype=torch.bool) + expected = torch.tensor([[[1, 1, 0, 0], [1, 1, 0, 0], [1, 1, 1, 0], [1, 1, 1, 1]]], dtype=torch.bool) + assert torch.equal(make_att_2d_masks(pad, att), expected) + + # Padding removes rows and columns. + pad = torch.tensor([[True, True, False]]) + att = torch.tensor([[0, 1, 1]], dtype=torch.int32) + out = make_att_2d_masks(pad, att) + assert not out[0, :, 2].any() and not out[0, 2, :].any() + + +def test_make_att_2d_masks_validation(): + with pytest.raises(ValueError): + make_att_2d_masks(torch.ones(3, dtype=torch.bool), torch.ones(1, 3, dtype=torch.int32)) + with pytest.raises(ValueError): + make_att_2d_masks(torch.ones(1, 3, dtype=torch.bool), torch.ones(3, dtype=torch.int32)) + + +def test_prepare_attention_masks_4d(): + masks = torch.tensor([[[True, False], [False, True]]]) + out = prepare_attention_masks_4d(masks) + assert out.shape == (1, 1, 2, 2) + expected = torch.tensor([[[[0.0, OPENPI_ATTENTION_MASK_VALUE], [OPENPI_ATTENTION_MASK_VALUE, 0.0]]]]) + assert torch.equal(out, expected) + + out_bf16 = prepare_attention_masks_4d(masks, dtype=torch.bfloat16) + assert out_bf16.dtype == torch.bfloat16 + assert torch.equal(out_bf16, expected.to(torch.bfloat16)) + + +def test_pad_vector_openpi_semantics(): + v = torch.arange(6.0).reshape(2, 3) + padded = pad_vector(v, 5) + assert padded.shape == (2, 5) + assert torch.equal(padded[:, :3], v) and not padded[:, 3:].any() + # Already large enough (>=): returned unchanged, same object. + assert pad_vector(v, 3) is v + assert pad_vector(v, 2) is v + # 3D input. + v3 = torch.ones(2, 4, 3) + assert pad_vector(v3, 7).shape == (2, 4, 7) + + +def test_pad_vector_truncate_semantics(): + v = torch.arange(6.0).reshape(2, 3) + out = pad_vector(v, 2, truncate=True) + assert out.shape == (2, 2) and torch.equal(out, v[:, :2]) + out = pad_vector(v, 5, truncate=True) + assert out.shape == (2, 5) and torch.equal(out[:, :3], v) and not out[:, 3:].any() + assert pad_vector(v, 0, truncate=True).shape == (2, 0) + assert pad_vector(v, 3, truncate=True) is v + + +@pytest.mark.parametrize("channels_last", [True, False]) +def test_resize_with_pad_torch_centered(channels_last): + img = torch.rand(2, 3, 30, 60) if not channels_last else torch.rand(2, 30, 60, 3) + out = resize_with_pad_torch(img, 64, 64) + if channels_last: + assert out.shape == (2, 64, 64, 3) + # Aspect ratio preserved: 30x60 -> 32x64, padded 16 top and 16 bottom (centered). + assert not out[:, :16].any() and not out[:, -16:].any() + assert out[:, 16:48].abs().sum() > 0 + else: + assert out.shape == (2, 3, 64, 64) + assert not out[:, :, :16].any() and not out[:, :, -16:].any() + + +def test_resize_with_pad_torch_uint8_roundtrip(): + img = (torch.rand(1, 3, 20, 20) * 255).to(torch.uint8) + out = resize_with_pad_torch(img, 40, 40) + assert out.dtype == torch.uint8 and out.shape == (1, 3, 40, 40) + with pytest.raises(ValueError, match="Unsupported image dtype"): + resize_with_pad_torch(torch.rand(1, 3, 8, 8, dtype=torch.float64), 16, 16) + + +def test_resize_with_pad_top_left(): + img = torch.rand(2, 3, 30, 60) + out = resize_with_pad(img, 64, 64, pad_value=-1.0) + assert out.shape == (2, 3, 64, 64) + # 30x60 -> 32x64; this variant pads on the TOP only (32 rows of pad_value). + assert torch.equal(out[:, :, :32], torch.full((2, 3, 32, 64), -1.0)) + assert out[:, :, 32:].min() >= 0 + # No-op fast path returns the same object. + assert resize_with_pad(img, 30, 60, pad_value=0.0) is img + with pytest.raises(ValueError, match="expected"): + resize_with_pad(torch.rand(3, 8, 8), 16, 16, pad_value=0.0) + + +def test_clone_past_key_values(): + pytest.importorskip("transformers") + from transformers import DynamicCache + + from lerobot.policies.common.vla_utils import clone_past_key_values + + cache = DynamicCache() + keys, values = torch.rand(1, 2, 4, 8), torch.rand(1, 2, 4, 8) + cache.update(keys, values, 0) + cloned = clone_past_key_values(cache) + (ck, cv, _), (ok, ov, _) = next(iter(cloned)), next(iter(cache)) + assert torch.equal(ck, ok) and torch.equal(cv, ov) + # Deep copy: mutating the clone must not touch the original. + ck.zero_() + assert not torch.equal(ck, ok) + + +def test_clone_past_key_values_is_fullgraph_compilable(): + pytest.importorskip("transformers") + from transformers import DynamicCache + + from lerobot.policies.common.vla_utils import clone_past_key_values + + cache = DynamicCache() + keys, values = torch.rand(1, 2, 4, 8), torch.rand(1, 2, 4, 8) + cache.update(keys, values, 0) + + compiled_clone = torch.compile(clone_past_key_values, backend="eager", fullgraph=True) + cloned = compiled_clone(cache) + + (cloned_keys, cloned_values, _), (original_keys, original_values, _) = ( + next(iter(cloned)), + next(iter(cache)), + ) + assert torch.equal(cloned_keys, original_keys) + assert torch.equal(cloned_values, original_values) diff --git a/tests/policies/wall_x/test_wallx.py b/tests/policies/wall_x/test_wallx.py index 85656eca2..ebff6b88d 100644 --- a/tests/policies/wall_x/test_wallx.py +++ b/tests/policies/wall_x/test_wallx.py @@ -25,13 +25,57 @@ pytest.importorskip("transformers") pytest.importorskip("torchdiffeq") from lerobot.policies.factory import make_policy_config # noqa: E402 -from lerobot.policies.wall_x import WallXConfig # noqa: E402 +from lerobot.policies.wall_x import ( + WallXConfig, # noqa: E402 +) from lerobot.policies.wall_x.modeling_wall_x import WallXPolicy # noqa: E402 from lerobot.policies.wall_x.processor_wall_x import make_wall_x_pre_post_processors # noqa: E402 +from lerobot.policies.wall_x.qwen_model import Qwen2_5_VLMoEModel, Qwen2_5_VLTextConfig # noqa: E402 from lerobot.utils.random_utils import set_seed # noqa: E402 from tests.utils import require_cuda, require_hf_token # noqa: E402 +def test_moe_model_captures_requested_hidden_states_and_attentions(): + hidden_size = 16 + expert_config = { + "hidden_size": hidden_size, + "intermediate_size": 32, + "hidden_act": "silu", + } + config = Qwen2_5_VLTextConfig( + vocab_size=32, + hidden_size=hidden_size, + intermediate_size=32, + num_hidden_layers=2, + num_attention_heads=4, + num_key_value_heads=4, + max_position_embeddings=32, + layer_types=["full_attention", "full_attention"], + rope_parameters={ + "rope_type": "default", + "rope_theta": 1_000_000.0, + "mrope_section": [1, 1, 0], + }, + num_experts=2, + experts=[expert_config, expert_config], + dim_inputs=(hidden_size, hidden_size), + mlp_moe=True, + ) + config._attn_implementation = "eager" + model = Qwen2_5_VLMoEModel(config) + input_ids = torch.tensor([[1, 2, 3]]) + + output = model( + input_ids=input_ids, + moe_token_types=torch.zeros_like(input_ids), + output_hidden_states=True, + output_attentions=True, + ) + + assert len(output.hidden_states) == config.num_hidden_layers + 1 + assert len(output.attentions) == config.num_hidden_layers + + @require_cuda @require_hf_token def test_policy_instantiation(): diff --git a/tests/scripts/test_lerobot_annotate.py b/tests/scripts/test_lerobot_annotate.py index 03f1ee104..b1db997e3 100644 --- a/tests/scripts/test_lerobot_annotate.py +++ b/tests/scripts/test_lerobot_annotate.py @@ -18,6 +18,8 @@ import json from types import SimpleNamespace import pytest +import requests +from huggingface_hub.errors import RevisionNotFoundError # ``lerobot.scripts.lerobot_annotate`` (and the ``_push_to_hub`` path it # exercises) imports ``lerobot.datasets``, which only ships under the @@ -26,7 +28,7 @@ pytest.importorskip("datasets", reason="datasets is required (install lerobot[da def test_push_to_hub_tags_uploaded_dataset_revision(tmp_path, monkeypatch): - from lerobot.scripts.lerobot_annotate import _push_to_hub + from lerobot.scripts import lerobot_annotate root = tmp_path / "dataset" (root / "meta").mkdir(parents=True) @@ -45,9 +47,6 @@ def test_push_to_hub_tags_uploaded_dataset_revision(tmp_path, monkeypatch): return SimpleNamespace(oid="abc123") def delete_tag(self, repo_id, **kwargs): - import requests - from huggingface_hub.errors import RevisionNotFoundError - calls["delete_tag"] = {"repo_id": repo_id, **kwargs} # Simulate the common case: no stale tag to delete. raise RevisionNotFoundError("no such tag", response=requests.Response()) @@ -55,7 +54,7 @@ def test_push_to_hub_tags_uploaded_dataset_revision(tmp_path, monkeypatch): def create_tag(self, **kwargs): calls["create_tag"] = kwargs - monkeypatch.setattr("huggingface_hub.HfApi", FakeHfApi) + monkeypatch.setattr(lerobot_annotate, "HfApi", FakeHfApi) def fake_card_push(self, **kwargs): calls["card_push"] = {"content": str(self), **kwargs} @@ -69,7 +68,7 @@ def test_push_to_hub_tags_uploaded_dataset_revision(tmp_path, monkeypatch): push_commit_message=None, ) - _push_to_hub(root, cfg) + lerobot_annotate._push_to_hub(root, cfg) assert calls["create_repo"] == { "repo_id": "annotated/dataset",