code stringlengths 66 870k | docstring stringlengths 19 26.7k | func_name stringlengths 1 138 | language stringclasses 1
value | repo stringlengths 7 68 | path stringlengths 5 324 | url stringlengths 46 389 | license stringclasses 7
values |
|---|---|---|---|---|---|---|---|
def ndgrid(*tensors) -> Tuple[torch.Tensor, ...]:
"""generate N-D grid in dimension order.
The ndgrid function is like meshgrid except that the order of the first two input arguments are switched.
That is, the statement
[X1,X2,X3] = ndgrid(x1,x2,x3)
produces the same result as
[X2,X1,X3] = m... | generate N-D grid in dimension order.
The ndgrid function is like meshgrid except that the order of the first two input arguments are switched.
That is, the statement
[X1,X2,X3] = ndgrid(x1,x2,x3)
produces the same result as
[X2,X1,X3] = meshgrid(x2,x1,x3)
This naming is based on MATLAB, th... | ndgrid | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def drop_block_2d(
x,
drop_prob: float = 0.1,
block_size: int = 7,
gamma_scale: float = 1.0,
with_noise: bool = False,
inplace: bool = False,
batchwise: bool = False,
):
"""DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
DropBlock with an experimental gaussian noise option. This... | DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
DropBlock with an experimental gaussian noise option. This layer has been tested on a few training
runs with success, but needs further validation and possibly optimization for lower runtime impact.
| drop_block_2d | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def drop_block_fast_2d(
x: torch.Tensor,
drop_prob: float = 0.1,
block_size: int = 7,
gamma_scale: float = 1.0,
with_noise: bool = False,
inplace: bool = False,
):
"""DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
DropBlock with an experimental gaussian noise option. Simplied from ... | DropBlock. See https://arxiv.org/pdf/1810.12890.pdf
DropBlock with an experimental gaussian noise option. Simplied from above without concern for valid
block mask at edges.
| drop_block_fast_2d | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def drop_path(x, drop_prob: float = 0.0, training: bool = False, scale_by_keep: bool = True):
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
the original name is misleadi... | Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
See discussion: https:... | drop_path | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def forward(
self,
hidden_states: torch.Tensor,
) -> Tuple[torch.FloatTensor, Optional[torch.FloatTensor], Optional[Tuple[torch.FloatTensor]]]:
"""
Args:
hidden_states (`Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]`): input to the layer of shape `(batch, seq_len,... |
Args:
hidden_states (`Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]`): input to the layer of shape `(batch, seq_len, embed_dim)`
| forward | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def forward(
self,
inputs_embeds,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, BaseModelOutput]:
r"""
Args:
inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
... |
Args:
inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
Embedded representation of the inputs. Should be float, not int tokens.
output_hidden_states (`bool`, *optional*):
Whether or not to return the hidden states ... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/intern/modeling_intern_vit.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/intern/modeling_intern_vit.py | Apache-2.0 |
def trunc_normal_tf_(
tensor: torch.Tensor, mean: float = 0.0, std: float = 1.0, a: float = -2.0, b: float = 2.0
) -> torch.Tensor:
"""Fills the input Tensor with values drawn from a truncated
normal distribution. The values are effectively drawn from the
normal distribution :math:`\\mathcal{N}(\text{me... | Fills the input Tensor with values drawn from a truncated
normal distribution. The values are effectively drawn from the
normal distribution :math:`\mathcal{N}( ext{mean}, ext{std}^2)`
with values outside :math:`[a, b]` redrawn until they are within
the bounds. The method used for generating the random... | trunc_normal_tf_ | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def interpolate_pos_encoding(self, embeddings: torch.Tensor, height: int, width: int) -> torch.Tensor:
"""
This method is an adapted method for SigLIP (due to SigLIP not having class embedding unlike other ViTs)
that allows the model to interpolate the pre-trained position encodings such that it... |
This method is an adapted method for SigLIP (due to SigLIP not having class embedding unlike other ViTs)
that allows the model to interpolate the pre-trained position encodings such that it can be usable on
higher resolution images.
Source:
https://github.com/facebookresearch/d... | interpolate_pos_encoding | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
hidden_states: torch.Tensor,
attention_mask: torch.Tensor,
output_attentions: Optional[bool] = False,
) -> Tuple[torch.FloatTensor]:
"""
Args:
hidden_states (`torch.FloatTensor`):
Input to the layer of shape `(batch, seq_... |
Args:
hidden_states (`torch.FloatTensor`):
Input to the layer of shape `(batch, seq_len, embed_dim)`.
attention_mask (`torch.FloatTensor`):
Attention mask of shape `(batch, 1, q_len, k_v_seq_len)` where padding elements are indicated by very large negativ... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
inputs_embeds,
attention_mask: Optional[torch.Tensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, BaseModelOutput]:
r"""
Args:
... |
Args:
inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
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_... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
input_ids: Optional[torch.Tensor] = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.Tensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bo... |
Returns:
Examples:
```python
>>> from transformers import AutoTokenizer, SiglipTextModel
>>> model = SiglipTextModel.from_pretrained("google/siglip-base-patch16-224")
>>> tokenizer = AutoTokenizer.from_pretrained("google/siglip-base-patch16-224")
>>> # import... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
pixel_values,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
interpolate_pos_encoding: bool = False,
) -> Union[Tuple, BaseModelOutputWithPooling]:
r"""
Retu... |
Returns:
Examples:
```python
>>> from PIL import Image
>>> import requests
>>> from transformers import AutoProcessor, SiglipVisionModel
>>> model = SiglipVisionModel.from_pretrained("google/siglip-base-patch16-224")
>>> processor = AutoProcessor.from_... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def get_text_features(
self,
input_ids: Optional[torch.Tensor] = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.Tensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: O... |
Returns:
text_features (`torch.FloatTensor` of shape `(batch_size, output_dim`): The text embeddings obtained by
applying the projection layer to the pooled output of [`SiglipTextModel`].
Examples:
```python
>>> from transformers import AutoTokenizer, AutoModel... | get_text_features | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def get_image_features(
self,
pixel_values: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
interpolate_pos_encoding: bool = False,
) -> torch.FloatTensor:
... |
Returns:
image_features (`torch.FloatTensor` of shape `(batch_size, output_dim`): The image embeddings obtained by
applying the projection layer to the pooled output of [`SiglipVisionModel`].
Examples:
```python
>>> from PIL import Image
>>> import requ... | get_image_features | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
pixel_values: Optional[torch.FloatTensor] = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
return_loss: Optional[bool] = None,
output_attentions... |
Returns:
Examples:
```python
>>> from PIL import Image
>>> import requests
>>> from transformers import AutoProcessor, AutoModel
>>> import torch
>>> model = AutoModel.from_pretrained("google/siglip-base-patch16-224")
>>> processor = AutoProces... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def forward(
self,
pixel_values: Optional[torch.Tensor] = None,
labels: Optional[torch.Tensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
interpolate_pos_encoding: bool = False... |
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for computing the image classification/regression loss. Indices should be in `[0, ...,
config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
`config.n... | forward | python | NVlabs/VILA | llava/model/multimodal_encoder/siglip/modeling_siglip.py | https://github.com/NVlabs/VILA/blob/master/llava/model/multimodal_encoder/siglip/modeling_siglip.py | Apache-2.0 |
def is_mm_model(model_path):
"""
Check if the model at the given path is a visual language model.
Args:
model_path (str): The path to the model.
Returns:
bool: True if the model is an MM model, False otherwise.
"""
config = AutoConfig.from_pretrained(model_path)
architectur... |
Check if the model at the given path is a visual language model.
Args:
model_path (str): The path to the model.
Returns:
bool: True if the model is an MM model, False otherwise.
| is_mm_model | python | NVlabs/VILA | llava/model/utils/utils.py | https://github.com/NVlabs/VILA/blob/master/llava/model/utils/utils.py | Apache-2.0 |
def process_vision_info(
conversations: list[dict] | list[list[dict]],
return_video_kwargs: bool = False,
) -> tuple[list[PIL.Image.Image] | None, list[torch.Tensor | list[PIL.Image.Image]] | None, Optional[dict]]:
"""
referernce from qwen_vl_utils
NVILA does not depend on th... |
referernce from qwen_vl_utils
NVILA does not depend on the function, but the interface is the same.
| process_vision_info | python | NVlabs/VILA | llava/remote_code/auto_processor.py | https://github.com/NVlabs/VILA/blob/master/llava/remote_code/auto_processor.py | Apache-2.0 |
def __call__(
self,
conversation=None,
**kwargs: Unpack[VILAProcessorKwargs],
) -> BatchFeature:
"""
The `conv` will be look like
[
{
'from': 'human',
'value': [
<transformers_modules.NVILA-Lite-2B-hf-pre... |
The `conv` will be look like
[
{
'from': 'human',
'value': [
<transformers_modules.NVILA-Lite-2B-hf-preview.media.Image object at 0x154e68e4c460>,
'What are the common elements in these pictures?'
]
... | __call__ | python | NVlabs/VILA | llava/remote_code/auto_processor.py | https://github.com/NVlabs/VILA/blob/master/llava/remote_code/auto_processor.py | Apache-2.0 |
def post_process_image_text_to_text(self, generated_outputs):
"""
Post-process the output of the model to decode the text.
Args:
generated_outputs (`torch.Tensor` or `np.ndarray`):
The output of the model `generate` function. The output is expected to be a tensor of ... |
Post-process the output of the model to decode the text.
Args:
generated_outputs (`torch.Tensor` or `np.ndarray`):
The output of the model `generate` function. The output is expected to be a tensor of shape `(batch_size, sequence_length)`
or `(sequence_lengt... | post_process_image_text_to_text | python | NVlabs/VILA | llava/remote_code/auto_processor.py | https://github.com/NVlabs/VILA/blob/master/llava/remote_code/auto_processor.py | Apache-2.0 |
def check_dot_in_model_path(model_path: str):
"""Check if the model path contains dot, which will affect the remote code loading."""
if osp.isdir(model_path): # local model
if "." in osp.abspath(model_path):
return True
else: # remote model
if "." in model_path:
ret... | Check if the model path contains dot, which will affect the remote code loading. | check_dot_in_model_path | python | NVlabs/VILA | llava/remote_code/modeling_vila.py | https://github.com/NVlabs/VILA/blob/master/llava/remote_code/modeling_vila.py | Apache-2.0 |
def generate(
self,
input_ids: Optional[torch.FloatTensor] = None,
media: Optional[Dict[str, List[torch.Tensor]]] = None,
media_config: Dict[str, Dict[str, Any]] = None,
attention_mask: Optional[torch.LongTensor] = None,
return_output_ids_only: bool = False,
**gen... |
input_tokens: <image> describe the image
media: [Tensor(1, 3, 384, 384), ]
----------->
input_tokens: 36000 001 002 003 004
input_emds: <media emd> 001 002 003 004
| generate | python | NVlabs/VILA | llava/remote_code/modeling_vila.py | https://github.com/NVlabs/VILA/blob/master/llava/remote_code/modeling_vila.py | Apache-2.0 |
def split_to_even_chunks(indices, lengths, num_chunks):
"""
Split a list of indices into `chunks` chunks of roughly equal lengths.
"""
if len(indices) % num_chunks != 0:
return [indices[i::num_chunks] for i in range(num_chunks)]
num_indices_per_chunk = len(indices) // num_chunks
chunk... |
Split a list of indices into `chunks` chunks of roughly equal lengths.
| split_to_even_chunks | python | NVlabs/VILA | llava/train/llava_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/train/llava_trainer.py | Apache-2.0 |
def create_optimizer(self):
"""
Setup the optimizer.
We provide a reasonable default that works well. If you want to use something else, you can pass a tuple in the
Trainer's init through `optimizers`, or subclass and override this method in a subclass.
"""
if is_sagemak... |
Setup the optimizer.
We provide a reasonable default that works well. If you want to use something else, you can pass a tuple in the
Trainer's init through `optimizers`, or subclass and override this method in a subclass.
| create_optimizer | python | NVlabs/VILA | llava/train/llava_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/train/llava_trainer.py | Apache-2.0 |
def log(self, logs: Dict[str, float]) -> None:
"""
Log `logs` on the various objects watching training.
Subclass and override this method to inject custom behavior.
Args:
logs (`Dict[str, float]`):
The values to log.
"""
if self.state.epoch i... |
Log `logs` on the various objects watching training.
Subclass and override this method to inject custom behavior.
Args:
logs (`Dict[str, float]`):
The values to log.
| log | python | NVlabs/VILA | llava/train/llava_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/train/llava_trainer.py | Apache-2.0 |
def dice_loss(
inputs: torch.Tensor,
targets: torch.Tensor,
scale=100, # 100000.0,
eps=1e-6,
):
"""
Compute the DICE loss, similar to generalized IOU for masks
Args:
inputs: A float tensor of arbitrary shape.
The predictions fo... |
Compute the DICE loss, similar to generalized IOU for masks
Args:
inputs: A float tensor of arbitrary shape.
The predictions for each example.
targets: A float tensor with the same shape as inputs. Stores the binary
classification label fo... | dice_loss | python | NVlabs/VILA | llava/train/llava_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/train/llava_trainer.py | Apache-2.0 |
def get_nb_trainable_parameters(model) -> tuple[int, int]:
r"""
Returns the number of trainable parameters and the number of all parameters in the model.
"""
trainable_params = 0
all_param = 0
for _, param in model.named_parameters():
num_params = param.numel()
# if using DS Zero... |
Returns the number of trainable parameters and the number of all parameters in the model.
| get_nb_trainable_parameters | python | NVlabs/VILA | llava/train/train.py | https://github.com/NVlabs/VILA/blob/master/llava/train/train.py | Apache-2.0 |
def __getitem__(self, i) -> Dict[str, torch.Tensor]:
"""
{
'prompt': 'Is there a snowman wearing a green scarf and hat in the background?',
'chosen': 'No, there is no snowman wearing a green scarf and hat in the background of the image. The image features a person ...',
... |
{
'prompt': 'Is there a snowman wearing a green scarf and hat in the background?',
'chosen': 'No, there is no snowman wearing a green scarf and hat in the background of the image. The image features a person ...',
'rejected': 'No, there is no snowman in the background.',
... | __getitem__ | python | NVlabs/VILA | llava/train/train.py | https://github.com/NVlabs/VILA/blob/master/llava/train/train.py | Apache-2.0 |
def get_train_dataloader(self) -> DataLoader:
"""
Returns the training [`~torch.utils.data.DataLoader`].
Will use no sampler if `train_dataset` does not implement `__len__`, a random sampler (adapted to distributed
training if necessary) otherwise.
Subclass and override this method if you want to ... |
Returns the training [`~torch.utils.data.DataLoader`].
Will use no sampler if `train_dataset` does not implement `__len__`, a random sampler (adapted to distributed
training if necessary) otherwise.
Subclass and override this method if you want to inject some custom behavior.
| get_train_dataloader | python | NVlabs/VILA | llava/train/train_llm_to_long.py | https://github.com/NVlabs/VILA/blob/master/llava/train/train_llm_to_long.py | Apache-2.0 |
def patched_normalize(
image: np.ndarray,
mean: Union[float, Iterable[float]],
std: Union[float, Iterable[float]],
data_format: Optional[ChannelDimension] = None,
input_data_format: Optional[Union[str, ChannelDimension]] = None,
) -> np.ndarray:
"""
Normalizes `image` using the mean and stan... |
Normalizes `image` using the mean and standard deviation specified by `mean` and `std`.
image = (image - mean) / std
Args:
image (`np.ndarray`):
The image to normalize.
mean (`float` or `Iterable[float]`):
The mean to use for normalization.
std (`float` or ... | patched_normalize | python | NVlabs/VILA | llava/train/transformer_normalize_monkey_patch.py | https://github.com/NVlabs/VILA/blob/master/llava/train/transformer_normalize_monkey_patch.py | Apache-2.0 |
def training_step(
self, model: nn.Module, inputs: Dict[str, Union[torch.Tensor, Any]], num_items_in_batch=None
) -> torch.Tensor:
"""
Perform a training step on a batch of inputs.
Subclass and override to inject custom behavior.
Args:
model (`nn.Module`):
The model to train.
... |
Perform a training step on a batch of inputs.
Subclass and override to inject custom behavior.
Args:
model (`nn.Module`):
The model to train.
inputs (`Dict[str, Union[torch.Tensor, Any]]`):
The inputs and targets of the model.
The dictionary will be unpac... | training_step | python | NVlabs/VILA | llava/train/transformer_normalize_monkey_patch.py | https://github.com/NVlabs/VILA/blob/master/llava/train/transformer_normalize_monkey_patch.py | Apache-2.0 |
def compute_loss(self, model, inputs, return_outputs=False, num_items_in_batch=None):
"""
How the loss is computed by Trainer. By default, all models return the loss in the first element.
Subclass and override for custom behavior.
"""
if (self.label_smoother is not None or self.compute_loss_func is ... |
How the loss is computed by Trainer. By default, all models return the loss in the first element.
Subclass and override for custom behavior.
| compute_loss | python | NVlabs/VILA | llava/train/transformer_normalize_monkey_patch.py | https://github.com/NVlabs/VILA/blob/master/llava/train/transformer_normalize_monkey_patch.py | Apache-2.0 |
def __init__(
self,
module=None,
data_parallel_group=None,
mem_efficient_linear=True,
remote_device=None,
pin_memory=False,
config_dict_or_path=None,
config=None,
enabled=True,
dtype=None,
mpu=None,
):
"""A context manag... | A context manager to partition the model parameters during the model
construction with MiCS partition strategy. Model states are partitioned
to the number of devices specified via ``mics_shard_size`` field in the
deepspeed config json file. The context manager also introduces
hierarchica... | __init__ | python | NVlabs/VILA | llava/train/deepspeed_replace/runtime/zero/mics.py | https://github.com/NVlabs/VILA/blob/master/llava/train/deepspeed_replace/runtime/zero/mics.py | Apache-2.0 |
def _convert_to_zero_parameters(self, ds_config, module, mpu):
"""overload the parent class function for convert the parameters"""
log_dist(f"Convert to zero parameters from MiCS Offload manager", ranks=[0])
non_zero_params = [p for p in module.parameters() if not is_zero_param(p)]
if no... | overload the parent class function for convert the parameters | _convert_to_zero_parameters | python | NVlabs/VILA | llava/train/deepspeed_replace/runtime/zero/mics.py | https://github.com/NVlabs/VILA/blob/master/llava/train/deepspeed_replace/runtime/zero/mics.py | Apache-2.0 |
def load_state_dict(
self, state_dict_list, load_optimizer_states=True, load_from_fp32_weights=False, checkpoint_folder=None
):
r"""Loading the ZeRO-3/MiCS partitioned checkpoints
Because the self.dp_process_group is replaced with the communicator for
partition group we can call the ... | Loading the ZeRO-3/MiCS partitioned checkpoints
Because the self.dp_process_group is replaced with the communicator for
partition group we can call the load_state_dict logic from ZeRO-3.
| load_state_dict | python | NVlabs/VILA | llava/train/deepspeed_replace/runtime/zero/mics.py | https://github.com/NVlabs/VILA/blob/master/llava/train/deepspeed_replace/runtime/zero/mics.py | Apache-2.0 |
def all_to_all_4D(input: torch.tensor, scatter_idx: int = 2, gather_idx: int = 1, group=None) -> torch.tensor:
"""
all-to-all for QKV
Args:
input (torch.tensor): a tensor sharded along dim scatter dim
scatter_idx (int): default 1
gather_idx (int): default 2
group : torch pro... |
all-to-all for QKV
Args:
input (torch.tensor): a tensor sharded along dim scatter dim
scatter_idx (int): default 1
gather_idx (int): default 2
group : torch process group
Returns:
torch.tensor: resharded tensor (bs, seqlen/P, hc, hs)
| all_to_all_4D | python | NVlabs/VILA | llava/train/sequence_parallel/all_to_all.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/all_to_all.py | Apache-2.0 |
def all_to_all_5D(input: torch.tensor, scatter_idx: int = 3, gather_idx: int = 1, group=None) -> torch.tensor:
"""
all-to-all for QKV
forward (bs, seqlen/N, 3, hc, hs) -> (bs, seqlen, 3, hc/N, hs)
Args:
input (torch.tensor): a tensor sharded along dim scatter dim
scatter_idx (int): defa... |
all-to-all for QKV
forward (bs, seqlen/N, 3, hc, hs) -> (bs, seqlen, 3, hc/N, hs)
Args:
input (torch.tensor): a tensor sharded along dim scatter dim
scatter_idx (int): default 1
gather_idx (int): default 2
group : torch process group
Returns:
torch.tensor: resh... | all_to_all_5D | python | NVlabs/VILA | llava/train/sequence_parallel/all_to_all.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/all_to_all.py | Apache-2.0 |
def set_pg_manager(sp_degree, sp_ring_degree=1, use_ulysses_low=True, ring_type=None):
"""
Set the process group manager for sequence parallelism.
sp_degree = sp_ring_degree x sp_ulysses_degree
"""
# first check torch distributed group init and set device accordingly;
# (DL) TODO: Whether this ... |
Set the process group manager for sequence parallelism.
sp_degree = sp_ring_degree x sp_ulysses_degree
| set_pg_manager | python | NVlabs/VILA | llava/train/sequence_parallel/globals.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/globals.py | Apache-2.0 |
def forward(
self,
query: Tensor,
key: Tensor,
value: Tensor,
*args: Any,
attention_mask=None,
dropout_p=0.0,
softmax_scale=None,
seqlens_in_batch=None,
causal=False,
window_size=(-1, -1),
alibi_slopes=None,
determin... | forward
Arguments:
query (Tensor): query input to the layer
key (Tensor): key input to the layer
value (Tensor): value input to the layer
args: other args
Returns:
* output (Tensor): context output
| forward | python | NVlabs/VILA | llava/train/sequence_parallel/hybrid_attn.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/hybrid_attn.py | Apache-2.0 |
def forward(
self,
query: Tensor,
key: Tensor,
value: Tensor,
dropout_p=0.0,
softmax_scale=None,
causal=False,
window_size=(-1, -1),
alibi_slopes=None,
deterministic=False,
return_attn_probs=False,
*args: Any,
) -> Tenso... | forward
Arguments:
query (Tensor): query input to the layer (bs, seqlen/P, hc, hs)
key (Tensor): key input to the layer (bs, seqlen/P, hc_kv, hs)
value (Tensor): value input to the layer (bs, seqlen/P, hc_kv, hs)
args: other args
Returns:
* o... | forward | python | NVlabs/VILA | llava/train/sequence_parallel/hybrid_attn.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/hybrid_attn.py | Apache-2.0 |
def _flash_attention_forward(
query_states: torch.Tensor,
key_states: torch.Tensor,
value_states: torch.Tensor,
attention_mask: torch.Tensor,
query_length: int,
is_causal: bool,
dropout: float = 0.0,
position_ids: Optional[torch.Tensor] = None,
softmax_scale: Optional[float] = None,
... |
Calls the forward method of Flash Attention - if the input hidden states contain at least one padding token
first unpad the input, then computes the attention scores and pad the final attention scores.
Args:
query_states (`torch.Tensor`):
Input query states to be passed to Flash Attent... | _flash_attention_forward | python | NVlabs/VILA | llava/train/sequence_parallel/monkey_patch.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/monkey_patch.py | Apache-2.0 |
def backward(ctx, grad_output_k, grad_output_v):
"""
For backward, we sum the copy head inside a query group.
"""
num_head_dim = ctx.num_head_dim
num_heads_kv = ctx.num_heads_kv
# we split the grad into query groups splits.
grad_output_k_splits = torch.chunk(gra... |
For backward, we sum the copy head inside a query group.
| backward | python | NVlabs/VILA | llava/train/sequence_parallel/ulysses_attn.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/ulysses_attn.py | Apache-2.0 |
def unflatten_varlen_lse(lse, cu_seqlens, max_seqlen: int):
"""
Arguments:
lse: (total_seqlen, nheads, 1)
cu_seqlens: (batch_size + 1,)
max_seqlen: int
Return:
unflatten_lse: (batch_size, nheads, max_seqlen)
"""
lse = lse.unsqueeze(dim=-1)
batch_size = len(cu_seql... |
Arguments:
lse: (total_seqlen, nheads, 1)
cu_seqlens: (batch_size + 1,)
max_seqlen: int
Return:
unflatten_lse: (batch_size, nheads, max_seqlen)
| unflatten_varlen_lse | python | NVlabs/VILA | llava/train/sequence_parallel/ring/triton_utils.py | https://github.com/NVlabs/VILA/blob/master/llava/train/sequence_parallel/ring/triton_utils.py | Apache-2.0 |
def top_k_top_p_filtering(
logits: torch.FloatTensor,
top_k: int = 0,
top_p: float = 1.0,
filter_value: float = -float("Inf"),
min_tokens_to_keep: int = 1,
) -> torch.FloatTensor:
"""
Filter a distribution of logits using top-k and/or nucleus (top-p) filtering.
Args:
logits: log... |
Filter a distribution of logits using top-k and/or nucleus (top-p) filtering.
Args:
logits: logits distribution shape (batch size, vocabulary size)
top_k (`int`, *optional*, defaults to 0):
If > 0, only keep the top k tokens with highest probability (top-k filtering)
top_p ... | top_k_top_p_filtering | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def flatten_dict(nested: Dict, sep: str = "/") -> Dict:
"""Flatten dictionary and concatenate nested keys with separator."""
def recurse(nest: Dict, prefix: str, into: Dict) -> None:
for k, v in nest.items():
if sep in k:
raise ValueError(f"separator '{sep}' not allowed to b... | Flatten dictionary and concatenate nested keys with separator. | flatten_dict | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def convert_to_scalar(stats: Dict) -> Dict:
"""
Converts the stats from a flattened dict to single scalar dicts
"""
tensorboard_stats = {}
for k, v in stats.items():
# for tensorboard compatibility - arrays and tensors are ignored with tensorboard
# therefore we convert single elemen... |
Converts the stats from a flattened dict to single scalar dicts
| convert_to_scalar | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def logprobs_from_logits(logits: torch.Tensor, labels: torch.Tensor, gather: bool = True) -> torch.Tensor:
"""
See: https://github.com/pytorch/pytorch/issues/563#issuecomment-330103591
"""
logp = F.log_softmax(logits, dim=2)
if not gather:
return logp
logpy = torch.gather(logp, 2, label... |
See: https://github.com/pytorch/pytorch/issues/563#issuecomment-330103591
| logprobs_from_logits | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def masked_mean(values: torch.Tensor, mask: torch.Tensor, axis: Optional[bool] = None) -> torch.Tensor:
"""Compute mean of tensor with a masked values."""
if axis is not None:
return (values * mask).sum(axis=axis) / mask.sum(axis=axis)
else:
return (values * mask).sum() / mask.sum() | Compute mean of tensor with a masked values. | masked_mean | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def masked_var(values: torch.Tensor, mask: torch.Tensor, unbiased: bool = True) -> torch.Tensor:
"""Compute variance of tensor with masked values."""
mean = masked_mean(values, mask)
centered_values = values - mean
variance = masked_mean(centered_values**2, mask)
if unbiased:
mask_sum = mask... | Compute variance of tensor with masked values. | masked_var | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def average_torch_dicts(list_of_dicts: List[Dict]) -> Dict:
"""Average values of a list of dicts with torch tensors."""
average_dict = dict()
for key in list_of_dicts[0].keys():
average_dict[key] = torch.mean(torch.stack([d[key] for d in list_of_dicts]), axis=0)
return average_dict | Average values of a list of dicts with torch tensors. | average_torch_dicts | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def stats_to_np(stats_dict: Dict) -> Dict:
"""Cast all torch.tensors in dict to numpy arrays."""
new_dict = dict()
for k, v in stats_dict.items():
if isinstance(v, torch.Tensor):
new_dict[k] = v.detach().cpu()
if new_dict[k].dtype == torch.bfloat16:
new_dict[k... | Cast all torch.tensors in dict to numpy arrays. | stats_to_np | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def set_seed(seed: int) -> None:
"""
Helper function for reproducible behavior to set the seed in `random`, `numpy`, and `torch`.
Args:
seed (`int`): The seed to set.
"""
random.seed(seed)
np.random.seed(seed)
torch.manual_seed(seed)
if is_xpu_available():
torch.xpu.manu... |
Helper function for reproducible behavior to set the seed in `random`, `numpy`, and `torch`.
Args:
seed (`int`): The seed to set.
| set_seed | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def randn_tensor(
shape: Union[Tuple, List],
generator: Optional[Union[List[torch.Generator], torch.Generator]] = None,
device: Optional[torch.device] = None,
dtype: Optional[torch.dtype] = None,
layout: Optional[torch.layout] = None,
) -> torch.Tensor:
"""A helper function to create random tens... | A helper function to create random tensors on the desired `device` with the desired `dtype`. When
passing a list of generators, you can seed each batch size individually. If CPU generators are passed, the tensor
is always created on the CPU.
| randn_tensor | python | NVlabs/VILA | llava/trl/core.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/core.py | Apache-2.0 |
def is_npu_available() -> bool:
"""Checks if `torch_npu` is installed and potentially if a NPU is in the environment"""
if importlib.util.find_spec("torch") is None or importlib.util.find_spec("torch_npu") is None:
return False
import torch
import torch_npu # noqa: F401
return hasattr(tor... | Checks if `torch_npu` is installed and potentially if a NPU is in the environment | is_npu_available | python | NVlabs/VILA | llava/trl/import_utils.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/import_utils.py | Apache-2.0 |
def __call__(self, input_ids, scores, **kwargs):
"""Returns true if all generated sequences contain any of the stop strings."""
if self.first_call:
self.generated_tokens = [1 for _ in range(input_ids.shape[0])]
self.start_length = input_ids.shape[-1] - 1
self.first_ca... | Returns true if all generated sequences contain any of the stop strings. | __call__ | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def __init__(self, text, tokens, system=True):
"""
Initialize TextHistory.
args:
text (`str`): The text of the first segment.
tokens (`torch.LongTensor`): The tokens of the first segment.
system (`bool`, *optional*): Whether the first segment is a system or u... |
Initialize TextHistory.
args:
text (`str`): The text of the first segment.
tokens (`torch.LongTensor`): The tokens of the first segment.
system (`bool`, *optional*): Whether the first segment is a system or user segment.
| __init__ | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def append_segment(self, text, tokens, system=True):
"""
Append a new segment to the history.
args:
text (`str`): The text of the new segment.
tokens (`torch.LongTensor`): The tokens of the new segment.
system (`bool`, *optional*): Whether the new segment is ... |
Append a new segment to the history.
args:
text (`str`): The text of the new segment.
tokens (`torch.LongTensor`): The tokens of the new segment.
system (`bool`, *optional*): Whether the new segment is a system or user segment.
| append_segment | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def split_query_response_tokens(self):
"""
Split the tokens into query and response tokens.
"""
split_index = self.token_spans[0][1]
query = self.tokens[:split_index]
response = self.tokens[split_index:]
mask = self.token_masks[split_index:]
return query,... |
Split the tokens into query and response tokens.
| split_query_response_tokens | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def __init__(
self,
model=None,
tokenizer=None,
tools=None,
reward_fn=None,
prompt=None,
max_turns=4,
max_tool_reponse=100,
max_length=None,
generation_kwargs=None,
):
"""
Initialize TextEnvironment.
Args:
... |
Initialize TextEnvironment.
Args:
model (`PreTrainedModelWrapper`): The model to use for generation.
tokenizer (`transformers.PreTrainedTokenizer`): The tokenizer to use for generation.
tools (list): A list of tools to use for interaction.
reward_fn (fun... | __init__ | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def run(self, queries, **rewards_kwargs):
"""
Run the environment on a list of queries.
Args:
queries (list[str]): A list of queries to run the model in the environment on.
"""
turns = 0
queries = [self.prompt + task for task in queries]
queries_toke... |
Run the environment on a list of queries.
Args:
queries (list[str]): A list of queries to run the model in the environment on.
| run | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def step(self, history):
"""
Step the environment forward one turn.
Args:
history (`TextHistory`): The history to step forward.
"""
truncated, ended = self.task_end_check(history)
if ended:
history.complete(truncated=truncated)
if history.... |
Step the environment forward one turn.
Args:
history (`TextHistory`): The history to step forward.
| step | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def parse_tool_call(self, text):
"""
Parse request string. Expected format: <request><tool_name>query<call>
"""
result = re.search(f"(?<={self.request_token}).*?(?={self.call_token})", text, re.DOTALL)
# if we can't find a <request>/<call> span we return none
if result i... |
Parse request string. Expected format: <request><tool_name>query<call>
| parse_tool_call | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def compute_reward(self, histories, **reward_kwargs):
"""
Compute the reward for a list of histories.
"""
rewards = self.reward_fn([history.last_text_segment for history in histories], **reward_kwargs)
for history, reward in zip(histories, rewards):
history.reward = r... |
Compute the reward for a list of histories.
| compute_reward | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def generate(self, histories):
"""
Generate responses for a list of histories.
"""
active_histories = [i for i, history in enumerate(histories) if not history.completed]
query_tensors = [histories[i].tokens for i in active_histories]
response_tensors = self._generate_bat... |
Generate responses for a list of histories.
| generate | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def tasks_end_check(self, histories, model_turn=True):
"""
Check if the current generation sequences have finished.
"""
for history in histories:
if not history.completed:
truncated, ended = self.task_end_check(history, model_turn=model_turn)
i... |
Check if the current generation sequences have finished.
| tasks_end_check | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def task_end_check(self, history, model_turn=True):
"""
Check if the current generation sequence has finished.
"""
truncated = False
ended = False
if history.completed:
return truncated, ended
if self.max_length is not None and len(self.tokenizer(histo... |
Check if the current generation sequence has finished.
| task_end_check | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def _generate_batched(
self,
query_tensors,
batch_size: int = 16,
pad_to_multiple_of: int = None,
):
"""
Generate responses for a list of query tensors.
args:
query_tensors (list[torch.Tensor]): A list of query tensors to generate responses for.
... |
Generate responses for a list of query tensors.
args:
query_tensors (list[torch.Tensor]): A list of query tensors to generate responses for.
batch_size (int): The batch size to use for generation.
pad_to_multiple_of (int): The padding length to use for generation.
... | _generate_batched | python | NVlabs/VILA | llava/trl/environment/base_environment.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/environment/base_environment.py | Apache-2.0 |
def __init__(
self,
model: PreTrainedModelWrapper,
tokenizer: Union[PreTrainedTokenizer, PreTrainedTokenizerFast],
queries_to_scores: Callable[[List[str]], List[float]],
length_sampler: Any,
sample_size: int = 4,
seed: Optional[int] = None,
n_candidates: i... |
Initialize the sampler for best-of-n generation
Args:
model (`PreTrainedModelWrapper`):
The pretrained model to use for generation
tokenizer (`PreTrainedTokenizer` or `PreTrainedTokenizerFast`):
Tokenizer associated with the pretrained model
... | __init__ | python | NVlabs/VILA | llava/trl/extras/best_of_n_sampler.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/extras/best_of_n_sampler.py | Apache-2.0 |
def generate(
self,
tokenized_query: Union[List[int], torch.Tensor, List[torch.Tensor], List[List[int]]],
skip_special_tokens: bool = True,
device: Optional[Union[str, torch.device]] = None,
**generation_kwargs,
) -> List[List[str]]:
r"""
Generate the best of ... |
Generate the best of n samples for input queries
Args:
tokenized_query (`List[int]` or `torch.Tensor` or `List[torch.Tensor]` or `List[int]`):
represents either a single tokenized query (a single tensor or a list of integers) or a batch of tokenized queries (a list of tenso... | generate | python | NVlabs/VILA | llava/trl/extras/best_of_n_sampler.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/extras/best_of_n_sampler.py | Apache-2.0 |
def conversations_formatting_function(tokenizer: AutoTokenizer, messages_field: Literal["messages", "conversations"]):
r"""
return a callable function that takes in a "messages" dataset and returns a formatted dataset, based on the tokenizer
apply chat template to the dataset
"""
def format_dataset... |
return a callable function that takes in a "messages" dataset and returns a formatted dataset, based on the tokenizer
apply chat template to the dataset
| conversations_formatting_function | python | NVlabs/VILA | llava/trl/extras/dataset_formatting.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/extras/dataset_formatting.py | Apache-2.0 |
def instructions_formatting_function(tokenizer: AutoTokenizer):
r"""
return a callable function that takes in an "instructions" dataset and returns a formatted dataset, based on the tokenizer
apply chat template to the dataset
"""
def format_dataset(examples):
if isinstance(examples["prompt... |
return a callable function that takes in an "instructions" dataset and returns a formatted dataset, based on the tokenizer
apply chat template to the dataset
| instructions_formatting_function | python | NVlabs/VILA | llava/trl/extras/dataset_formatting.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/extras/dataset_formatting.py | Apache-2.0 |
def get_formatting_func_from_dataset(
dataset: Union[Dataset, ConstantLengthDataset], tokenizer: AutoTokenizer
) -> Optional[Callable]:
r"""
Finds the correct formatting function based on the dataset structure. Currently supported datasets are:
- `ChatML` with [{"role": str, "content": str}]
- `inst... |
Finds the correct formatting function based on the dataset structure. Currently supported datasets are:
- `ChatML` with [{"role": str, "content": str}]
- `instruction` with [{"prompt": str, "completion": str}]
Args:
dataset (Dataset): User dataset
tokenizer (AutoTokenizer): Tokenizer u... | get_formatting_func_from_dataset | python | NVlabs/VILA | llava/trl/extras/dataset_formatting.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/extras/dataset_formatting.py | Apache-2.0 |
def from_pretrained(cls, pretrained_model_name_or_path, *model_args, **kwargs):
r"""
Instantiates a new model from a pretrained model from `transformers`. The
pretrained model is loaded using the `from_pretrained` method of the
`transformers.PreTrainedModel` class. The arguments that are... |
Instantiates a new model from a pretrained model from `transformers`. The
pretrained model is loaded using the `from_pretrained` method of the
`transformers.PreTrainedModel` class. The arguments that are specific to the
`transformers.PreTrainedModel` class are passed along this method a... | from_pretrained | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def _get_current_device(cls):
r"""
Get the current device. For GPU, we return the local process index using the `accelerate.PartialState`
object to handle corner cases when running scripts in distributed environments.
Returns:
current_device (`Union[int, str]`):
... |
Get the current device. For GPU, we return the local process index using the `accelerate.PartialState`
object to handle corner cases when running scripts in distributed environments.
Returns:
current_device (`Union[int, str]`):
The current device.
| _get_current_device | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def _split_kwargs(cls, kwargs):
"""
Separate the kwargs from the arguments that we support inside
`supported_args` and the ones that we don't.
"""
check_peft_kwargs = False
if is_peft_available():
from peft import prepare_model_for_kbit_training
... |
Separate the kwargs from the arguments that we support inside
`supported_args` and the ones that we don't.
| _split_kwargs | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def add_and_load_reward_modeling_adapter(
cls, pretrained_model, adapter_model_id, adapter_name="reward_model_adapter", token=None
):
r"""
Add and load a reward modeling adapter. This method can only be used if the
model is a `PeftModel` and if you have initialized the model with the... |
Add and load a reward modeling adapter. This method can only be used if the
model is a `PeftModel` and if you have initialized the model with the `reward_modeling_adapter_id`
argument, pointing to the id of the reward modeling adapter. The latest needs also to contain the
score head in ... | add_and_load_reward_modeling_adapter | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def save_pretrained(self, *args, **kwargs):
r"""
Save the pretrained model to a directory. This method is a wrapper around
`transformers.PreTrainedModel.save_pretrained`. Please refer to the documentation
of `transformers.PreTrainedModel.save_pretrained` for more information.
Ar... |
Save the pretrained model to a directory. This method is a wrapper around
`transformers.PreTrainedModel.save_pretrained`. Please refer to the documentation
of `transformers.PreTrainedModel.save_pretrained` for more information.
Args:
*args (`list`, *optional*):
... | save_pretrained | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def compute_reward_score(self, input_ids, attention_mask=None, **kwargs):
r"""
Computes the reward score for a given input. The method has first to enable the adapter
and then compute the reward score. After that the model disables the reward modeling
adapter and enables the default ppo ... |
Computes the reward score for a given input. The method has first to enable the adapter
and then compute the reward score. After that the model disables the reward modeling
adapter and enables the default ppo adapter again.
| compute_reward_score | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def create_reference_model(
model: PreTrainedModelWrapper, num_shared_layers: int = None, pattern: str = None
) -> PreTrainedModelWrapper:
"""
Creates a static reference copy of a model. Note that model will be in `.eval()` mode.
Args:
model (`PreTrainedModelWrapper`): The model to be copied.
... |
Creates a static reference copy of a model. Note that model will be in `.eval()` mode.
Args:
model (`PreTrainedModelWrapper`): The model to be copied.
num_shared_layers (`int`, *optional*): The number of initial layers that are shared between both models and kept frozen.
pattern (`str`... | create_reference_model | python | NVlabs/VILA | llava/trl/models/modeling_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_base.py | Apache-2.0 |
def _left_broadcast(input_tensor, shape):
"""
As opposed to the default direction of broadcasting (right to left), this function broadcasts
from left to right
Args:
input_tensor (`torch.FloatTensor`): is the tensor to broadcast
shape (`Tuple[int]`): is the shape to broadcast ... |
As opposed to the default direction of broadcasting (right to left), this function broadcasts
from left to right
Args:
input_tensor (`torch.FloatTensor`): is the tensor to broadcast
shape (`Tuple[int]`): is the shape to broadcast to
| _left_broadcast | python | NVlabs/VILA | llava/trl/models/modeling_sd_base.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_sd_base.py | Apache-2.0 |
def __init__(self, pretrained_model, **kwargs):
r"""
Initializes the model.
Args:
pretrained_model (`transformers.PreTrainedModel`):
The model to wrap. It should be a causal language model such as GPT2.
or any model mapped inside the `AutoModelForCaus... |
Initializes the model.
Args:
pretrained_model (`transformers.PreTrainedModel`):
The model to wrap. It should be a causal language model such as GPT2.
or any model mapped inside the `AutoModelForCausalLM` class.
kwargs (`dict`, `optional`):
... | __init__ | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def _init_weights(self, **kwargs):
r"""
Initializes the weights of the value head. The default initialization strategy is random.
Users can pass a different initialization strategy by passing the `v_head_init_strategy` argument
when calling `.from_pretrained`. Supported strategies are:
... |
Initializes the weights of the value head. The default initialization strategy is random.
Users can pass a different initialization strategy by passing the `v_head_init_strategy` argument
when calling `.from_pretrained`. Supported strategies are:
- `normal`: initializes the weights with... | _init_weights | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def forward(
self,
input_ids=None,
past_key_values=None,
attention_mask=None,
**kwargs,
):
r"""
Applies a forward pass to the wrapped model and returns the logits of the value head.
Args:
input_ids (`torch.LongTensor` of shape `(batch_size... |
Applies a forward pass to the wrapped model and returns the logits of the value head.
Args:
input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
Indices of input sequence tokens in the vocabulary.
past_key_values (`tuple(tuple(torch.FloatTens... | forward | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def state_dict(self, *args, **kwargs):
r"""
Returns the state dictionary of the model. We add the state dictionary of the value head
to the state dictionary of the wrapped model by prepending the key with `v_head.`.
"""
if not self.is_peft_model:
pretrained_model_stat... |
Returns the state dictionary of the model. We add the state dictionary of the value head
to the state dictionary of the wrapped model by prepending the key with `v_head.`.
| state_dict | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def post_init(self, state_dict):
r"""
We add the state dictionary of the value head to the state dictionary of the wrapped model
by prepending the key with `v_head.`. This function removes the `v_head.` prefix from the
keys of the value head state dictionary.
"""
for k in... |
We add the state dictionary of the value head to the state dictionary of the wrapped model
by prepending the key with `v_head.`. This function removes the `v_head.` prefix from the
keys of the value head state dictionary.
| post_init | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def set_device_hook(module, input, outputs):
r"""
A hook that sets the device of the output of the model to the device of the first
parameter of the model.
Args:
module (`nn.Module`):
The module to which the hoo... |
A hook that sets the device of the output of the model to the device of the first
parameter of the model.
Args:
module (`nn.Module`):
The module to which the hook is attached.
input (`tuple`):
... | set_device_hook | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def _init_weights(self, **kwargs):
r"""
We initialize the weights of the value head.
"""
initializer_range = kwargs.pop("v_head_initializer_range", 0.2)
# random init by default
init_strategy = kwargs.pop("v_head_init_strategy", None)
if init_strategy is None:
... |
We initialize the weights of the value head.
| _init_weights | python | NVlabs/VILA | llava/trl/models/modeling_value_head.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/modeling_value_head.py | Apache-2.0 |
def setup_chat_format(
model: PreTrainedModel,
tokenizer: PreTrainedTokenizer,
format: Optional[Literal["chatml"]] = "chatml",
resize_to_multiple_of: Optional[int] = None,
) -> Tuple[PreTrainedModel, PreTrainedTokenizer]:
"""
Setup chat format by adding special tokens to the tokenizer, setting t... |
Setup chat format by adding special tokens to the tokenizer, setting the correct format, and extending the embedding layer of the model based on the new special tokens.
Args:
model (`~transformers.PreTrainedModel`): The model to be modified.
tokenizer (`~transformers.PreTrainedTokenizer`): The tok... | setup_chat_format | python | NVlabs/VILA | llava/trl/models/utils.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/models/utils.py | Apache-2.0 |
def step(self, epoch: int, global_step: int):
"""
Perform a single step of training.
Args:
epoch (int): The current epoch.
global_step (int): The current global step.
Side Effects:
- Model weights are updated
- Logs the statistics to the ... |
Perform a single step of training.
Args:
epoch (int): The current epoch.
global_step (int): The current global step.
Side Effects:
- Model weights are updated
- Logs the statistics to the accelerator trackers.
- If `self.image_sample... | step | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def calculate_loss(self, latents, timesteps, next_latents, log_probs, advantages, embeds):
"""
Calculate the loss for a batch of an unpacked sample
Args:
latents (torch.Tensor):
The latents sampled from the diffusion model, shape: [batch_size, num_channels_latents, h... |
Calculate the loss for a batch of an unpacked sample
Args:
latents (torch.Tensor):
The latents sampled from the diffusion model, shape: [batch_size, num_channels_latents, height, width]
timesteps (torch.Tensor):
The timesteps sampled from the dif... | calculate_loss | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def _generate_samples(self, iterations, batch_size):
"""
Generate samples from the model
Args:
iterations (int): Number of iterations to generate samples for
batch_size (int): Batch size to use for sampling
Returns:
samples (List[Dict[str, torch.Tens... |
Generate samples from the model
Args:
iterations (int): Number of iterations to generate samples for
batch_size (int): Batch size to use for sampling
Returns:
samples (List[Dict[str, torch.Tensor]]), prompt_image_pairs (List[List[Any]])
| _generate_samples | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def _train_batched_samples(self, inner_epoch, epoch, global_step, batched_samples):
"""
Train on a batch of samples. Main training segment
Args:
inner_epoch (int): The current inner epoch
epoch (int): The current epoch
global_step (int): The current global st... |
Train on a batch of samples. Main training segment
Args:
inner_epoch (int): The current inner epoch
epoch (int): The current epoch
global_step (int): The current global step
batched_samples (List[Dict[str, torch.Tensor]]): The batched samples to train on... | _train_batched_samples | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def train(self, epochs: Optional[int] = None):
"""
Train the model for a given number of epochs
"""
global_step = 0
if epochs is None:
epochs = self.config.num_epochs
for epoch in range(self.first_epoch, epochs):
global_step = self.step(epoch, glob... |
Train the model for a given number of epochs
| train | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def create_model_card(self, path: str, model_name: Optional[str] = "TRL DDPO Model") -> None:
"""Creates and saves a model card for a TRL model.
Args:
path (`str`): The path to save the model card to.
model_name (`str`, *optional*): The name of the model, defaults to `TRL DDPO M... | Creates and saves a model card for a TRL model.
Args:
path (`str`): The path to save the model card to.
model_name (`str`, *optional*): The name of the model, defaults to `TRL DDPO Model`.
| create_model_card | python | NVlabs/VILA | llava/trl/trainer/ddpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/ddpo_trainer.py | Apache-2.0 |
def get_train_dataloader(self) -> DataLoader:
"""
Returns the training [`~torch.utils.data.DataLoader`].
Subclass of transformers.src.transformers.trainer.get_train_dataloader to precompute `ref_log_probs`.
"""
if self.precompute_ref_log_probs and not self._precomputed_train_re... |
Returns the training [`~torch.utils.data.DataLoader`].
Subclass of transformers.src.transformers.trainer.get_train_dataloader to precompute `ref_log_probs`.
| get_train_dataloader | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
def get_eval_dataloader(self, eval_dataset: Optional[Dataset] = None) -> DataLoader:
"""
Returns the evaluation [`~torch.utils.data.DataLoader`].
Subclass of transformers.src.transformers.trainer.get_eval_dataloader to precompute `ref_log_probs`.
Args:
eval_dataset (`torch.... |
Returns the evaluation [`~torch.utils.data.DataLoader`].
Subclass of transformers.src.transformers.trainer.get_eval_dataloader to precompute `ref_log_probs`.
Args:
eval_dataset (`torch.utils.data.Dataset`, *optional*):
If provided, will override `self.eval_dataset`... | get_eval_dataloader | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
def build_tokenized_answer(self, prompt, answer):
"""
Llama tokenizer does satisfy `enc(a + b) = enc(a) + enc(b)`.
It does ensure `enc(a + b) = enc(a) + enc(a + b)[len(enc(a)):]`.
Reference:
https://github.com/EleutherAI/lm-evaluation-harness/pull/531#issuecomment-1595586257
... |
Llama tokenizer does satisfy `enc(a + b) = enc(a) + enc(b)`.
It does ensure `enc(a + b) = enc(a) + enc(a + b)[len(enc(a)):]`.
Reference:
https://github.com/EleutherAI/lm-evaluation-harness/pull/531#issuecomment-1595586257
| build_tokenized_answer | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
def tokenize_row(self, feature, model: Optional[Union[PreTrainedModel, nn.Module]] = None) -> Dict:
"""Tokenize a single row from a DPO specific dataset.
At this stage, we don't convert to PyTorch tensors yet; we just handle the truncation
in case the prompt + chosen or prompt + rejected respon... | Tokenize a single row from a DPO specific dataset.
At this stage, we don't convert to PyTorch tensors yet; we just handle the truncation
in case the prompt + chosen or prompt + rejected responses is/are too long. First
we truncate the prompt; if we're still too long, we truncate the chosen/... | tokenize_row | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
def null_ref_context(self):
"""Context manager for handling null reference model (that is, peft adapter manipulation)."""
with (
self.accelerator.unwrap_model(self.model).disable_adapter()
if self.is_peft_model and not self.ref_adapter_name
else nullcontext()
... | Context manager for handling null reference model (that is, peft adapter manipulation). | null_ref_context | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
def compute_reference_log_probs(self, padded_batch: Dict) -> Dict:
"""Computes log probabilities of the reference model for a single padded batch of a DPO specific dataset."""
compte_ref_context_manager = torch.cuda.amp.autocast if self._peft_has_been_casted_to_bf16 else nullcontext
# compute r... | Computes log probabilities of the reference model for a single padded batch of a DPO specific dataset. | compute_reference_log_probs | python | NVlabs/VILA | llava/trl/trainer/dpo_trainer.py | https://github.com/NVlabs/VILA/blob/master/llava/trl/trainer/dpo_trainer.py | Apache-2.0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.