instruction stringclasses 100
values | code stringlengths 78 193k | response stringlengths 259 170k | file stringlengths 59 203 |
|---|---|---|---|
Document this module using docstrings | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -29,6 +29,26 @@
class KimiK2TikTokenTokenizer(PreTrainedTokenizer):
+ """
+ Tokenizing and encoding/decoding text using the Tiktoken tokenizer. See megatron/tokenizer/tiktoken_tokenizer.py.
+ This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users shou... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/kimi_k2/tokenizer.py |
Add detailed documentation for each class | # coding=utf-8
# Copyright (c) 2026 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2026 The Moonshot AI Inc. team and 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 obtai... | --- +++ @@ -13,12 +13,25 @@ # 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.
+"""processor class for Kimi-K2.5."""
from ..image_processing_utils import BatchFeature
from ..processing_ut... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/kimi_k25/processor.py |
Replace inline comments with docstrings | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -43,6 +43,9 @@
class Gemma3TextScaledWordEmbedding(nn.Embedding):
+ """
+ This module overrides nn.Embeddings' forward by multiplying with embeddings scale.
+ """
def __init__(self, config):
num_embeddings = config.vocab_size
@@ -132,6 +135,17 @@ config: Optional[Gemma3Te... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/gemma3_text/modeling.py |
Help me document legacy Python code | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -22,16 +22,37 @@
class NewGELUActivation(nn.Layer):
-
- def forward(self, input: Tensor) -> Tensor:
+ """
+ Implementation of the GELU activation function currently in Google BERT repo (identical to OpenAI GPT). Also see
+ the Gaussian Error Linear Units paper: https://arxiv.org/abs/1606.0841... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/ernie4_5_moe_vl/model/dfnrope/activation.py |
Write docstrings including parameters and return values | # coding=utf-8
# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2022 The HuggingFace Inc. team.
#
# 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.a... | --- +++ @@ -38,6 +38,17 @@
class PaddleImageProcessingMixin:
+ """
+ A mixin that extends Hugging Face image processor classes to support
+ PaddlePaddle tensor return types.
+
+ This mixin dynamically wraps common preprocessing and transformation
+ methods, enabling them to return Paddle tensors when... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/image_processing_utils.py |
Write docstrings for data processing functions | # coding=utf-8
# Copyright (c) 2026 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2026 The Moonshot AI Inc. team and 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 obtai... | --- +++ @@ -32,6 +32,26 @@
class TikTokenTokenizer(PreTrainedTokenizer):
+ """
+ Tokenizing and encoding/decoding text using the Tiktoken tokenizer. See megatron/tokenizer/tiktoken_tokenizer.py.
+ This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users should ref... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/kimi_k25/tokenizer.py |
Write documentation strings for class attributes | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -12,6 +12,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""This module provides some utilities for training"""
import copy
import logging
@@ -41,6 +42,9 @@
class SmoothedValue:
+ """Track a series of values and provide access to smo... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/ernie4_5_moe_vl/model/utils/misc.py |
Auto-generate documentation strings for this file | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -67,6 +67,10 @@ return_tensors: Optional[Union[str, TensorType]] = None,
data_format: ChannelDimension = ChannelDimension.FIRST,
):
+ """
+ Checks validity of typically used arguments in an `ImageProcessorFast` `preprocess` method.
+ Raises `ValueError` if arguments incompatibility is caug... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/image_processing_utils_fast.py |
Fully document this Python code with docstrings | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -43,6 +43,17 @@ output_split_sizes=None,
input_split_sizes=None,
):
+ """
+ All-to-all communication in the group.
+
+ Args:
+ ctx (Any): Context object.
+ input (Tensor): Input tensor.
+ group (Group): The group object.
+
+ R... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/ernie4_5_moe_vl/model/dfnrope/modeling.py |
Add docstrings to improve collaboration | # coding=utf-8
# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2021 The HuggingFace Inc. team.
#
# 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.a... | --- +++ @@ -67,6 +67,9 @@
def to_numpy(obj):
+ """
+ Convert a Paddle tensor, Numpy array, python list or python dict to a Numpy array.
+ """
if isinstance(obj, (dict, UserDict)):
return {k: to_numpy(v) for k, v in obj.items()}
elif isinstance(obj, (list, tuple)):
@@ -125,6 +128,16 @@
... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/image_utils.py |
Add missing documentation to my Python functions | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -38,6 +38,7 @@
def rotate_half(x: paddle.Tensor) -> paddle.Tensor:
+ """Rotates half the hidden dims of the input."""
x1 = x[..., : x.shape[-1] // 2]
@@ -47,6 +48,15 @@
def apply_rotary_pos_emb(q, k, cos, sin, position_ids=None, unsqueeze_dim=1):
+ """
+ Applies rotary positional em... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/llama/modeling.py |
Please document this code using docstrings | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -62,6 +62,19 @@
class ModelProviderMixin(abc.ABC, Generic[ModelT]):
+ """A mixin that implements the ModelProvider pattern for PaddleFormers.
+
+ The ModelProvider pattern solves ecosystem fragmentation by providing a standardized
+ way to instantiate models. This mixin provides a consistent `pr... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/model_provider.py |
Add concise docstrings to each method | # Copyright (c) 2026 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2026 The Moonshot AI 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:/... | --- +++ @@ -13,6 +13,9 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""
+Encode structured tool declaration to typescript style string.
+"""
import dataclasses
import json
from collections.abc import Sequence
@@ -25,18 +28,21 @@
class _SchemaRegist... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/kimi_k25/tool_declaration_ts.py |
Annotate my code with docstrings | # Copyright (c) 2021 PaddlePaddle Authors. 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 appli... | --- +++ @@ -50,6 +50,10 @@
def mha_ofa_forward(self, query, key, value, attn_mask=None, cache=None):
+ """
+ monkey patch for MultiHeadAttention forward to accept head_mask
+ attn_mask[0] = attn_mask, attn_mask[1] = head_mask
+ """
key = query if key is None else key
value = query if value is ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/ofa_utils.py |
Write docstrings describing functionality | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. 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 re... | --- +++ @@ -12,6 +12,9 @@ # 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.
+"""
+API for image and video processing, serving as a backend for PaddlePaddle processors.
+"""
from typing im... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/paddle_vision_utils.py |
Add docstrings that explain logic | # Copyright 2025 the HuggingFace Team & PaddlePaddle Authors. 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
#
# Unles... | --- +++ @@ -12,6 +12,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""Image processor class for GLM-4.6V (PaddlePaddle version)."""
import math
from typing import Dict, List, Optional, Union
@@ -41,12 +42,25 @@
def is_scaled_image(image: np.ndarr... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm_ocr/image_processor.py |
Generate docstrings for script automation | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -23,6 +23,20 @@ startend_row_indices: paddle.Tensor,
window_size: int = 5,
):
+ """
+ Restrict start-end row indices to a sliding window range.
+
+ Args:
+ startend_row_indices (`paddle.Tensor`, optional):
+ Tensor of shape (batch_size, num_heads, seq_length, bound_num) c... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/masking_utils.py |
Please document this code using docstrings | # coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this fi... | --- +++ @@ -82,6 +82,12 @@
@dataclass(frozen=False, eq=True)
class AddedToken:
+ """
+ AddedToken represents a token to be added to a Tokenizer An AddedToken can have special options defining the
+ way it should behave.
+ The `normalized` will default to `not special` if it is not... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/legacy/tokenizer_utils_base.py |
Generate docstrings with parameter types | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -55,12 +55,24 @@
class GPTModel(FleetGPTModel, PretrainedModel):
+ """
+ GPTModel class that inherits from FleetGPTModel.
+ This class requires paddlefleet to be installed.
+ """
# GPTModel = FleetGPTModel
def local_layer_spec(config: "GPTModelProvider") -> LayerSpec:
+ """Create... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/gpt_provider.py |
Can you add docstrings to this Python file? | # coding=utf-8
# Copyright (c) 2026 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2026 The Moonshot AI Inc. team and 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 obtai... | --- +++ @@ -13,6 +13,7 @@ # 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.
+"""Image processor class for Kimi-K2.5."""
import json
import math
@@ -186,6 +187,16 @@ medias: list[... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/kimi_k25/vision_processor.py |
Insert docstrings into my code | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -23,6 +23,10 @@
def standardize_rope_params(config, rope_theta: float | dict[str, float] | None = None):
+ """
+ Helper to standardize the config's rope params field by ensuring the params are defined for each
+ later type. For old model the fn will duplicate a single rope param in each layer ty... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/modeling_rope_utils.py |
Write docstrings for algorithm functions | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -53,6 +53,7 @@
@dataclass
class GLMMoEModelProvider(GPTModelProvider):
+ """Base provider for GLM MoE Models."""
moe_router_load_balancing_type: str = "seq_aux_loss"
@@ -120,12 +121,14 @@
def rotate_half(x):
+ """Rotates half the hidden dims of the input."""
x1 = x[..., : x.shape[-... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm4_moe/modeling.py |
Create docstrings for all classes and functions | # Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2020 The HuggingFace 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.... | --- +++ @@ -29,6 +29,12 @@
def tuple_output(outputs: Tuple[Tensor], loss: Optional[Tensor] = None):
+ """re-construct the outputs with one method which contains the simple logic
+
+ Args:
+ outputs (Tuple[Tensor]): the source of the outputs
+ loss (Optional[Tensor], optional): the loss of the mo... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/model_outputs.py |
Write docstrings for this repository | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
# Copyright (c) 2025 DeepSeek
#
# 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 t... | --- +++ @@ -43,6 +43,19 @@ num_out_tokens: Optional[int] = None,
drop_and_pad: bool = False,
):
+ """Permute the tokens and probs based on the mask.
+ Tokens with the same designated expert will be grouped together.
+ The shape of mask is [tokens, num_experts], it indicates which experts were selecte... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/moe_utils.py |
Add return value explanations in docstrings | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -52,6 +52,14 @@ chat_template: Optional[str] = None,
**kwargs,
):
+ """
+ PaddleFormers Glm46VProcessor fallback apply_chat_template.
+
+ This method intentionally bypasses ProcessorMixin.apply_chat_template to avoid
+ processing_utils kwargs-grouping issues ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm_ocr/processor.py |
Write docstrings for this repository | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) Microsoft Corporation.
# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved.
# Copyright (C) 2024 THL A29 Limited, a Tencent company. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License")... | --- +++ @@ -29,6 +29,19 @@
def dispatching(x, dispatch_mask, scatter_index, num_experts, capacity):
+ """
+ Rearranges the input tensor `x` based on gate results, truncates it according to the specified capacity, and performs padding.
+
+ Args:
+ x (Tensor)[Seq, Dim]: The input tensor.
+ disp... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/moe_layer.py |
Write reusable docstrings | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -51,6 +51,10 @@
def repeat_kv(hidden_states: paddle.Tensor, n_rep: int) -> paddle.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_di... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm_ocr/modeling.py |
Add standardized docstrings across the file | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -12,12 +12,55 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+""" GlmOCR model configuration."""
from ..configuration_utils import PretrainedConfig
from ..modeling_rope_utils import rope_config_validation, standardize_rope_params
class Gl... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm_ocr/configuration.py |
Write clean docstrings for readability | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) Microsoft Corporation.
# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved.
# Copyright (C) 2024 THL A29 Limited, a Tencent company. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License")... | --- +++ @@ -77,6 +77,16 @@ gates: paddle.Tensor,
capacity_factor: float,
) -> paddle.Tensor:
+ """Calculate the capacity for each expert based on the gates and capacity factor.
+
+ Args:
+ gates (paddle.Tensor): A tensor of shape [num_tokens, num_experts] representing the ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/moe_gate_auto.py |
Add documentation for all methods | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Qwen2_5_VL model configuration"""
from ..configuration_utils import PretrainedConfig, layer_type_validation... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_5_vl/configuration.py |
Create docstrings for each class method | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) Microsoft Corporation.
# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved.
# Copyright (C) 2024 THL A29 Limited, a Tencent company. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License")... | --- +++ @@ -77,6 +77,16 @@ gates: paddle.Tensor,
capacity_factor: float,
) -> paddle.Tensor:
+ """Calculate the capacity for each expert based on the gates and capacity factor.
+
+ Args:
+ gates (paddle.Tensor): A tensor of shape [num_tokens, num_experts] representing the ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/moe_gate.py |
Help me comply with documentation standards | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Paddle Qwen2_5_VL model."""
from __future__ import annotations
from dataclasses import dataclass
@@ -112,1... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_5_vl/modeling.py |
Generate docstrings for this script | # Copyright (c) 2020 PaddlePaddle Authors. 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 appli... | --- +++ @@ -151,6 +151,16 @@
def prune_linear_layer(layer: nn.Linear, index: paddle.Tensor, dim: int = 0) -> nn.Linear:
+ """
+ Prune a linear layer to keep only entries in index.
+ Used to remove heads.
+ Args:
+ layer (`paddle.nn.Linear`): The layer to prune.
+ index (`paddle.Tensor`): T... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/model_utils.py |
Document helper functions with docstrings | # coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this fi... | --- +++ @@ -78,6 +78,13 @@
def convert_to_unicode(text):
+ """
+ Converts `text` to Unicode (if it's not already), assuming utf-8 input.
+ Args:
+ text (str|bytes): Text to be converted to unicode.
+ Returns:
+ str: converted text.
+ """
if isinstance(text, str):
return te... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/legacy/tokenizer_utils.py |
Add docstrings to meet PEP guidelines | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The ZhipuAI Inc. team and 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 Lic... | --- +++ @@ -20,6 +20,45 @@
class RopeParameters(TypedDict, total=False):
+ """
+ Args:
+ rope_theta (`float`):
+ The base period of the RoPE embeddings.
+ rope_type (`str`, *optional*, defaults to "default"):
+ The sub-variant of RoPE to use. Can be one of ['default', 'line... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/glm4v_moe/configuration.py |
Create docstrings for all classes and functions | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) Microsoft Corporation.
# Copyright (c) Facebook, Inc. and its affiliates. All rights reserved.
# Copyright (C) 2024 THL A29 Limited, a Tencent company. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License")... | --- +++ @@ -29,6 +29,19 @@
def dispatching(x, dispatch_mask, scatter_index, num_experts, capacity):
+ """
+ Rearranges the input tensor `x` based on gate results, truncates it according to the specified capacity, and performs padding.
+
+ Args:
+ x (Tensor)[Seq, Dim]: The input tensor.
+ disp... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/moe_layer_auto.py |
Add standardized docstrings across the file | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2024 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from i... | --- +++ @@ -18,6 +18,7 @@ # 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.
+"""processor class for Qwen2-VL."""
from typing import Optional, Union
@@ -40,6 +41,20 @@
class Qwen2VLP... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_vl/processor.py |
Generate docstrings with parameter types | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Qwen3VL model configuration"""
from ..configuration_utils import PretrainedConfig, layer_type_validation
f... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl/configuration.py |
Generate NumPy-style docstrings | # # Copyright 2024 Microsoft and 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 ... | --- +++ @@ -11,6 +11,7 @@ # 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.
+"""Paddle Phi3 model."""
from typing import Optional, Tuple, Union
@@ -40,12 +41,14 @@
def rotate_half(x... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/phi3/modeling.py |
Add docstrings that explain inputs and outputs | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original f... | --- +++ @@ -18,6 +18,7 @@ # 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.
+"""processor class for Qwen2.5-VL."""
from typing import Optional, Union
@@ -40,6 +41,20 @@
class Qwen2_... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_5_vl/processor.py |
Auto-generate documentation strings for this file | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2024 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from i... | --- +++ @@ -18,6 +18,7 @@ # 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.
+"""Fast image processor class for Qwen2-VL."""
from typing import Optional, Union
import paddle
@@ -45,6 +46,... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_vl/image_processor_fast.py |
Add docstrings with type hints explained | # Copyright (c) 2024 PaddlePaddle Authors. 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 appli... | --- +++ @@ -12,6 +12,9 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""
+This file is used for replacing Paddle's native Linear implementations with vendors' customized implementations
+"""
import paddle.distributed.fleet.meta_parallel as mpu
from pad... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/linear_utils.py |
Generate docstrings with examples | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and 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... | --- +++ @@ -13,6 +13,7 @@ # 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.
+"""processor class for Qwen3-VL."""
from typing import Optional, Union
@@ -36,6 +37,20 @@
class Qwen3VLP... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl/processor.py |
Create simple docstrings for beginners | # Copyright (c) 2024, NVIDIA CORPORATION. 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 appli... | --- +++ @@ -26,6 +26,9 @@
def get_hcg():
+ """
+ get the hybrid comm group from fleet
+ """
return fleet.get_hybrid_communicate_group()
@@ -63,9 +66,11 @@
class AllGatherVarlenOp(PyLayer):
+ """the shape of allgather can be not same for each rank"""
@staticmethod
def forward(ctx... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/sequence_parallel_utils.py |
Help me write clear docstrings | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original f... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Paddle Qwen3 model."""
from __future__ import annotations
import json
@@ -61,6 +62,7 @@
@dataclass
clas... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3/modeling.py |
Document this module using docstrings | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2024 The Qwen team, Alibaba Group and 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... | --- +++ @@ -12,6 +12,7 @@ # 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.
+"""Paddle Qwen2Moe model."""
from __future__ import annotations
import copy
@@ -55,6 +56,7 @@
@dataclass
c... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_moe/modeling.py |
Create Google-style docstrings for my code | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -12,6 +12,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""Image processor class for PaddleOCR-VL."""
import math
from typing import Dict, List, Optional, Union
@@ -41,6 +42,9 @@
def is_scaled_image(image: np.ndarray) -> bool:
+ """... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/paddleocr_vl/image_processor.py |
Create documentation for each function signature | # Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2024 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original f... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Paddle Qwen2 model."""
from __future__ import annotations
import json
@@ -61,6 +62,7 @@
@dataclass
clas... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2/modeling.py |
Write docstrings for this repository | # coding=utf-8
# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2022 The HuggingFace Inc. team.
#
# 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.a... | --- +++ @@ -13,6 +13,9 @@ # 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.
+"""
+ Processing saving/loading class for common processors.
+"""
import bisect
import copy
@@ -63,6 +66,10 @... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/processing_utils.py |
Generate docstrings with examples | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Qwen3VLMoemodel configuration"""
from ..configuration_utils import PretrainedConfig, layer_type_validation
... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl_moe/configuration.py |
Add docstrings for production code | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Paddle Qwen3_VL_Moe model."""
from __future__ import annotations
import types
@@ -186,12 +187,14 @@
de... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl_moe/modeling.py |
Create docstrings for all classes and functions | # coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this fi... | --- +++ @@ -69,6 +69,10 @@
class TensorType(ExplicitEnum):
+ """
+ Possible values for the `return_tensors` argument in [`PreTrainedTokenizerBase.__call__`]. Useful for
+ tab-completion in an IDE.
+ """
PADDLE = "pd"
NUMPY = "np"
@@ -80,6 +84,21 @@ self._wrap_return_tensor_methods()
... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/tokenizer_utils.py |
Add well-formatted docstrings | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -11,6 +11,7 @@ # 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.
+"""Paddle Qwen3-Next model."""
from typing import Any, List, Optional
@@ -74,6 +75,18 @@
class Qwen3Next... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_next/modeling.py |
Provide docstrings following PEP 257 | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
# Copyright (c) 2025 DeepSeek
#
# 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 t... | --- +++ @@ -25,33 +25,69 @@
class _DispatchManager(ABC):
+ """
+ A manager class to handle dispatch and combine processes for MoE models.
+
+ DispatcherManager handles token dispatching according to the routing_map of format
+ [num_local_tokens, world_size, num_instances]. The routing_map is a 3D tensor... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/token_dispatcher.py |
Add docstrings to existing functions | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2024 The Qwen team, Alibaba Group. 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... | --- +++ @@ -14,6 +14,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
# This code is adapted from https://github.com/QwenLM/Qwen3-VL/blob/main/qwen-vl-utils/src/qwen_vl_utils/vision_process.py
+"""Processing functions for image and video inputs for Qwen2/2.5-... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_vl/vision_process.py |
Add docstrings to my Python code | # Copyright 2024 Microsoft and 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 re... | --- +++ @@ -12,11 +12,18 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+""" Phi-3 model configuration."""
from ..configuration_utils import PretrainedConfig, layer_type_validation
from ..modeling_rope_utils import rope_config_validation, standardize_rope_... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/phi3/configuration.py |
Add docstrings explaining edge cases | # Copyright (c) 2022 PaddlePaddle Authors. 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 app... | --- +++ @@ -12,6 +12,10 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""
+This code is based on https://github.com/pytorch/pytorch/blob/master/torch/nn/init.py
+This copyright of pytorch/pytorch is a BSD-style license, as found in the LICENSE file.
+"""
... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/initializer.py |
Provide clean and structured docstrings | # Copyright (c) 2024, NVIDIA CORPORATION. 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 appli... | --- +++ @@ -34,6 +34,20 @@ tensor_parallel_output=True,
training=True,
):
+ """
+ Parallel matmul
+ Args:
+ lm_output: x for matmul
+ logit_weights: y for matmul
+ bias (Tensor, optional): Bias tensor. Default is None.
+ transpose_y (bool, optional): Whether to transpose y... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/tensor_parallel_utils.py |
Generate docstrings for script automation | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -38,6 +38,17 @@
class PaddleOCRVLProcessor(ProcessorMixin):
+ r"""
+ [`PaddleOCRVLProcessor`] offers all the functionalities of [`PaddleOCRImageProcessor`] and [`Ernie4_5Tokenizer`]. See the
+ [`~PaddleOCRVLProcessor.__call__`] and [`~PaddleOCRVLProcessor.decode`] for more information.
+ Args... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/paddleocr_vl/processor.py |
Document this code for team use | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -77,6 +77,7 @@
def get_layer_spec(is_vit, normalization) -> LayerSpec:
+ """Transformer Layer Spec."""
attn_mask_type = AttnMaskType.no_mask if is_vit else AttnMaskType.causal
if normalization == "LayerNorm":
norm = LayerNorm
@@ -128,11 +129,18 @@
class Qwen3VLTextTransformerLay... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl/modeling_fleet.py |
Generate documentation strings for clarity | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from i... | --- +++ @@ -18,6 +18,7 @@ # 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.
+"""video processor class for Qwen2-VL."""
import math
from typing import Optional, Union
@@ -97,6 +98,29 @@ ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen2_vl/video_processor.py |
Help me write clear docstrings | # Copyright (c) 2020 PaddlePaddle Authors. 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 appli... | --- +++ @@ -63,6 +63,15 @@
def convert_ndarray_dtype(np_array: np.ndarray, target_dtype: str) -> np.ndarray:
+ """convert ndarray
+
+ Args:
+ np_array (np.ndarray): numpy ndarray instance
+ target_dtype (str): the target dtype
+
+ Returns:
+ np.ndarray: converted numpy ndarray instance... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/utils.py |
Add detailed docstrings explaining each function | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen team, Alibaba Group and 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... | --- +++ @@ -12,6 +12,7 @@ # 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.
+"""Paddle Qwen3Moe model."""
from __future__ import annotations
import copy
@@ -55,6 +56,7 @@
@dataclass
c... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_moe/modeling.py |
Fully document this Python code with docstrings | # Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2025 The Qwen Team and The HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommo... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""Paddle Qwen3_VL model."""
from __future__ import annotations
from dataclasses import dataclass
@@ -166,12 ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/qwen3_vl/modeling.py |
Add docstrings including usage examples | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. 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 re... | --- +++ @@ -80,6 +80,7 @@
@property
def timestamps(self) -> list[float]:
+ "Timestamps of the sampled frames in seconds."
if self.fps is None or self.frames_indices is None:
raise ValueError("Cannot infer video `timestamps` when `fps` or `frames_indices` is None.")
retur... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/video_utils.py |
Help me document legacy Python code | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -13,6 +13,7 @@ # limitations under the License.
+"""Paddle PaddleOCR-VL model."""
from contextvars import ContextVar
from dataclasses import dataclass
@@ -58,6 +59,7 @@
def rotate_half(x):
+ """Rotates half the hidden dims of the input."""
x1 = x[..., : x.shape[-1] // 2]
x2 = x[..., ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/paddleocr_vl/modeling.py |
Write reusable docstrings | # Copyright (c) 2024 PaddlePaddle Authors. 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 appli... | --- +++ @@ -71,17 +71,20 @@
def set_recompute_id(value=-1):
+ """switch recompute id to the given value"""
global _recompute_id
_recompute_id = str(value)
def get_recompute_id():
+ """get current recompute id"""
global _recompute_id
return str(_recompute_id)
@contextlib.contextma... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/refined_recompute.py |
Add docstrings that explain logic | # coding=utf-8
# Copyright (c) 2025 PaddlePaddle Authors. 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 re... | --- +++ @@ -189,6 +189,9 @@ self,
video: "paddle.Tensor",
) -> VideoInput:
+ """
+ Converts a video to RGB format.
+ """
video = grayscale_to_rgb(video)
if video.shape[-3] == 3 or not (video[..., 3, :, :] < 255).any():
return video
@@ -204,6 +207... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/video_processing_utils.py |
Add docstrings to improve readability | # coding:utf-8
# Copyright (c) 2020 PaddlePaddle Authors. 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 req... | --- +++ @@ -40,6 +40,12 @@
class Logger(object):
+ """
+ Default logger in PaddleFormers
+
+ Args:
+ name(str) : Logger name, default is 'PaddleFormers'
+ """
def __init__(self, name: str = None):
name = "PaddleFormers" if not name else name
@@ -93,6 +99,13 @@
@contextlib.co... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/log.py |
Generate consistent documentation across files | # Copyright (c) 2022 PaddlePaddle Authors. 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 appli... | --- +++ @@ -48,11 +48,20 @@
class SerializationError(Exception):
+ """Exception for serialization"""
pass
def seek_by_string(file_handler: BufferedReader, string: str, file_size: int) -> int:
+ """seek the index of file-handler with target words
+ Args:
+ file_handler (BufferedReader): f... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/serialization.py |
Provide docstrings following PEP 257 | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -365,6 +365,15 @@ axis: int,
DEQUANT_SCALE_ROUNDING_MODE: DequantScaleRoundingMode = DequantScaleRoundingMode.ROUND_UP,
):
+ """
+ Convert the src weights to mx format. The src weight is quantized along the axis dimension.
+
+ If weight_quant_type is paddle.uint8, we output mxfp4 where two... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/upcast_downcast_triton.py |
Expand my code with proper documentation strings | # copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
#
# 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 applic... | --- +++ @@ -26,6 +26,26 @@
class ProfilerOptions(object):
+ """
+ Use a string to initialize a ProfilerOptions.
+ The string should be in the format: "key1=value1;key2=value;key3=value3".
+ For example:
+ "profile_path=model.profile"
+ "batch_range=[50, 60]; profile_path=model.profile"
+ ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/profiler.py |
Fully document this Python code with docstrings | # Copyright (c) 2024 PaddlePaddle Authors. 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 appli... | --- +++ @@ -74,6 +74,10 @@
def _cache_commit_hash_for_specific_revision(storage_folder: str, revision: str, commit_hash: str) -> None:
+ """Cache reference between a revision (tag, branch or truncated commit hash) and the corresponding commit hash.
+
+ Does nothing if `revision` is already a proper `commit_ha... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/download/common.py |
Write docstrings for data processing functions | # Copyright (c) 2022 PaddlePaddle Authors. 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 appli... | --- +++ @@ -37,6 +37,18 @@
def static_params_to_dygraph(model, static_tensor_dict):
+ """Simple tool for convert static parameters to dygraph parameters dict.
+
+ **NOTE** The model must both support static graph and dygraph mode.
+
+ Args:
+ model (nn.Layer): the model of a neural network.
+ ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/tools.py |
Add return value explanations in docstrings | # Copyright (c) 2024 PaddlePaddle Authors. 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 appli... | --- +++ @@ -38,6 +38,7 @@
class PDCErrorCode(Enum):
+ """Error Code For PDCTools usage"""
# success
Success = 0
@@ -61,14 +62,27 @@
class PDCTools:
+ """PDCTools"""
def __init__(self):
+ """ """
self._pdc_agent_bin = PDC_AGENT_BIN
self._hash_sum_bin = HASH_SUM_B... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/pdc_sdk.py |
Write clean docstrings for readability | # Copyright (c) 2022 PaddlePaddle Authors. 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 appli... | --- +++ @@ -22,6 +22,9 @@
class _LazyModule(ModuleType):
+ """
+ Module class that surfaces all objects but only performs associated imports when the objects are requested.
+ """
# Very heavily inspired by optuna.integration._IntegrationModule
# https://github.com/optuna/optuna/blob/master/optu... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/lazy_import.py |
Add docstrings for better understanding | # Copyright (c) 2023 PaddlePaddle Authors. 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 appli... | --- +++ @@ -25,6 +25,16 @@
def convert_file_size_to_int(size: Union[int, str]):
+ """
+ Converts a size expressed as a string with digits an unit (like `"5MB"`) to an integer (in bytes).
+ Args:
+ size (`int` or `str`): The size to convert. Will be directly returned if an `int`.
+ Example:
+ `... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/distributed.py |
Document my Python code with docstrings | # Copyright (c) 2025, Tri Dao.
# Ported from Flash Attention: https://github.com/Dao-AILab/flash-attention/blob/main/flash_attn/ops/triton/rotary.py
# Original code licensed under BSD-3-Clause license.
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (t... | --- +++ @@ -16,6 +16,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""Triton Rotary Position Embedding for PaddlePaddle."""
import paddle
@@ -64,6 +65,22 @@ BLOCK_H: tl.constexpr,
BLOCK_M: tl.constexpr,
):
+ """
+ Triton kernel for r... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/triton_kernels/rope_triton.py |
Add docstrings to improve readability | # Copyright (c) 2020 PaddlePaddle Authors. 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 app... | --- +++ @@ -20,6 +20,60 @@
class DistributedBatchSampler(paddle.io.BatchSampler):
+ """Sampler that restricts data loading to a subset of the dataset.
+
+ In such case, each process can pass a DistributedBatchSampler instance
+ as a DataLoader sampler, and load a subset of the original dataset that
+ is... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/batch_sampler.py |
Add docstrings explaining edge cases | # Copyright (c) 2020 PaddlePaddle Authors. 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 applic... | --- +++ @@ -11,6 +11,13 @@ # 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.
+"""
+This module is used to store environmental variables in PaddleFormers.
+PF_HOME --> the root ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/env.py |
Auto-generate documentation strings for this file | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -32,6 +32,20 @@ tensor_model_parallel_size: int,
logging_enabled: bool = True,
) -> int:
+ """Calculate padded vocab size for tensor parallelism.
+
+ This function pads the vocabulary size to ensure it's divisible by the required
+ multiple for efficient tensor parallel operations.
+
+ ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/transformers/vocab_utils.py |
Add detailed documentation for each class | # Copyright (c) 2025 PaddlePaddle Authors. 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 appli... | --- +++ @@ -26,6 +26,34 @@ dtype: Optional[paddle.dtype] = paddle.bfloat16,
is_causal: Optional[bool] = None,
):
+ """
+ Recover 4-D attention_mask from attn_mask_startend_row_indices.
+
+ Args:
+ attn_mask_startend_row_indices (paddle.Tensor):
+ A column-wise sparse attention mask ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/masking_utils.py |
Write proper docstrings for these functions | # Copyright (c) 2024 PaddlePaddle Authors. 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 appli... | --- +++ @@ -89,9 +89,14 @@
class LocalTokenNotFoundError(EnvironmentError):
+ """Raised if local token is required but not found."""
def _clean_token(token: Optional[str]) -> Optional[str]:
+ """Clean token by removing trailing and leading spaces and newlines.
+
+ If token is an empty string, return N... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/download/aistudio_hub_download.py |
Generate docstrings for script automation | # Copyright (c) 2020 PaddlePaddle Authors. 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 appli... | --- +++ @@ -82,10 +82,29 @@
def is_url(path):
+ """
+ Whether path is URL.
+ Args:
+ path (string): URL string or not.
+ """
return path.startswith("http://") or path.startswith("https://")
def get_weights_path_from_url(url, md5sum=None):
+ """Get weights path from WEIGHT_HOME, if no... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/downloader.py |
Can you add docstrings to this Python file? | # Copyright (c) 2020 PaddlePaddle Authors. 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 appli... | --- +++ @@ -30,10 +30,12 @@
def is_git_repo(dir: str) -> bool:
+ """Is the given directory version-controlled with git?"""
return os.path.exists(os.path.join(dir, ".git"))
def have_git() -> bool:
+ """Can we run the git executable?"""
try:
subprocess.check_output(["git", "--help"])
... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/setup.py |
Create documentation strings for testing functions | # Copyright (c) 2023 PaddlePaddle Authors. 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 appli... | --- +++ @@ -12,6 +12,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
+"""Git utilities."""
# https://github.com/python/mypy/blob/2c2d126cc742f2467045d36780c33bb8fb77a614/mypy/git.py#L1-L34
# Used also from setup.py, so don't pull in anything additional he... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/version/git.py |
Generate missing documentation strings | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -18,6 +18,18 @@
def parse_loss_values(log_content):
+ """Parse loss values from training log
+
+ The log format can be either:
+ - loss: 7.11594725 learning_rate: 5e-07 global_step: 1 ...
+ - loss: 12.24401569, learning_rate: 2.5e-06, global_step: 5, ...
+
+ Args:
+ log_content (str... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/scripts/ci_utils/log_analyzer.py |
Document functions with detailed explanations | # Copyright (c) 2022 PaddlePaddle Authors. 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 appli... | --- +++ @@ -31,6 +31,13 @@ self._id = name + "_" + str(uuid.uuid4())[-6:]
def __call__(self, sample, context=None):
+ """Process a sample.
+ Args:
+ sample (dict): a dict of sample, eg: {'image':xx, 'label': xxx}
+ context (dict): info about this sample processing
+ ... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/image_utils.py |
Add professional docstrings to my codebase | # Copyright (c) 2022 PaddlePaddle Authors. 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 applic... | --- +++ @@ -229,6 +229,10 @@
def is_paddle_available() -> bool:
+ """check if `torch` package is installed
+ Returns:
+ bool: if `torch` is available
+ """
return is_package_available("paddle")
@@ -241,35 +245,70 @@
def is_torch_available() -> bool:
+ """check if `torch` package is i... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/paddleformers/utils/import_utils.py |
Generate descriptive docstrings automatically | # Copyright (c) 2026 PaddlePaddle Authors. 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 appli... | --- +++ @@ -19,17 +19,36 @@
class TrainingTestRunner:
+ """Helper class for training tests - handles log parsing and validation only."""
def __init__(self, baseline_path, tolerance=1e-6):
+ """
+ Initialize the test runner.
+
+ Args:
+ baseline_path: Path to the baseline lo... | https://raw.githubusercontent.com/PaddlePaddle/PaddleFormers/HEAD/scripts/ci_utils/training_utils.py |
Annotate my code with docstrings | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -59,6 +59,38 @@
class LLMEngine(LLMEngine):
+ """An LLM engine that receives requests and generates texts.
+
+ This is the main class for the vLLM engine. It receives requests
+ from clients and generates texts from the LLM. It includes a tokenizer, a
+ language model (possibly distributed ac... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py |
Create docstrings for each class method | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -23,6 +23,7 @@
class TokenizerGroup:
+ """A group of tokenizers that can be used for LoRA adapters."""
def __init__(self, tokenizer: PreTrainedTokenizer, enable_lora: bool, max_num_seqs: int,
max_input_length: Optional[int]):
@@ -32,9 +33,11 @@ self.lora_tokenizers = L... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py |
Add docstrings to existing functions | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -29,6 +29,48 @@
class ModelConfig:
+ """Configuration for the model.
+
+ Args:
+ model: Name or path of the huggingface model to use.
+ tokenizer: Name or path of the huggingface tokenizer to use.
+ tokenizer_mode: Tokenizer mode. "auto" will use the fast tokenizer if
+ ... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_3_1/config.py |
Add docstrings to improve collaboration | # Copyright 2024 Bytedance Ltd. and/or its affiliates
#
# 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 ... | --- +++ @@ -11,6 +11,9 @@ # 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.
+"""
+the class of WorkerGroup
+"""
import logging
import threading
import signal
@@ -55,6 +58,10 @@
class ... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/single_controller/base/worker_group.py |
Include argument descriptions in docstrings | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommodate minor ar... | --- +++ @@ -17,6 +17,7 @@ # 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.
+"""PyTorch LLaMA model with Megatron-style acceleration."""
from typing import Optional, Tuple, Union
@@ -42... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/models/llama/megatron/modeling_llama_megatron.py |
Document this module using docstrings | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -12,6 +12,7 @@ # See the License for the specific language governing permissions and
# limitations under the License.
# Adapted from https://github.com/vllm-project/vllm/blob/main/vllm/worker/worker.py
+"""A GPU worker class."""
import os
import gc
from typing import Dict, List, Tuple, Optional, Union
@... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_4_2/worker.py |
Help me document legacy Python code | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -25,6 +25,7 @@
@dataclass
class EngineArgs:
+ """Arguments for vLLM engine."""
model_hf_config: PretrainedConfig = None
dtype: str = 'auto'
kv_cache_dtype: str = 'auto'
@@ -58,6 +59,7 @@
@staticmethod
def add_cli_args(parser: argparse.ArgumentParser) -> argparse.ArgumentParser... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_3_1/arg_utils.py |
Write docstrings for this repository | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -62,6 +62,7 @@
def get_model_loader(load_config: LoadConfig) -> BaseModelLoader:
+ """Get a model loader based on the load format."""
if isinstance(load_config.load_format, type):
return load_config.load_format(load_config)
@@ -100,6 +101,7 @@
class DummyModelLoader(BaseModelLoader... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py |
Write docstrings including parameters and return values | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -33,6 +33,58 @@
class LLM:
+ """An LLM for generating texts from given prompts and sampling parameters.
+
+ This class includes a tokenizer, a language model (possibly distributed
+ across multiple GPUs), and GPU memory space allocated for intermediate
+ states (aka KV cache). Given a batch o... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_4_2/llm.py |
Insert docstrings into my code | # Copyright 2024 Bytedance Ltd. and/or its affiliates
# Copyright 2023 The vLLM team.
# 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
#
# Unle... | --- +++ @@ -40,6 +40,7 @@
class SPMDGPUExecutor(ExecutorBase):
+ """SPMD-based multi-GPU executor implementations."""
def __init__(
self,
@@ -111,6 +112,15 @@ self.worker.load_model()
def determine_num_available_blocks(self) -> Tuple[int, int]:
+ """Determine the number of a... | https://raw.githubusercontent.com/Jiayi-Pan/TinyZero/HEAD/verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.