Skip to content

feat: Remove not used padding_idx in models #5385

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

feat: Remove not used padding_idx in models #5385

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9a95fb9
Remove not used padding_idx
HuiGao-NV Jun 20, 2025
f437248
Remove more padding_idx
HuiGao-NV Jun 20, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_bert.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -192,7 +192,6 @@ def __init__(self,
self.model_config = model_config

config = self.model_config.pretrained_config
self.padding_idx = config.pad_token_id
self.add_pooling_layer = add_pooling_layer

self.embedding = BertEmbeddings(config=config)
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_deepseekv3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -998,7 +998,6 @@ class DeepseekV3Model(DecoderModel):
def __init__(self, model_config: ModelConfig[PretrainedConfig]):
super().__init__(model_config)
config = model_config.pretrained_config
self.padding_idx = config.pad_token_id
self.vocab_size = config.vocab_size
self.num_hidden_layers = config.num_hidden_layers
aux_stream_list = [torch.cuda.Stream() for _ in range(2)]
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_gemma3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,7 +241,6 @@ def __init__(self, model_config: ModelConfig[Gemma3TextConfig]):
super().__init__(model_config)
config = self.model_config
self.hidden_size = config.pretrained_config.hidden_size
self.padding_idx = config.pretrained_config.pad_token_id

self.embed_tokens = Gemma3TextScaledWordEmbedding(
config.pretrained_config.vocab_size,
Expand Down
2 changes: 0 additions & 2 deletions tensorrt_llm/_torch/models/modeling_llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -609,7 +609,6 @@ class Llama4Model(DecoderModel):
def __init__(self, model_config: ModelConfig[LlamaConfig]):
super().__init__(model_config)
config = self.model_config.pretrained_config
self.padding_idx = config.pad_token_id
self.num_hidden_layers = config.num_hidden_layers
self.aux_stream = torch.cuda.Stream()
self.mapping = model_config.mapping
Expand Down Expand Up @@ -687,7 +686,6 @@ class LlamaModel(DecoderModel):
def __init__(self, model_config: ModelConfig[LlamaConfig]):
super().__init__(model_config)
config = self.model_config.pretrained_config
self.padding_idx = config.pad_token_id
self.num_hidden_layers = config.num_hidden_layers

vocab_size = config.vocab_size
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_mistral.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,7 +119,6 @@ class MistralModel(DecoderModel):
def __init__(self, model_config: ModelConfig[MistralConfig]):
super().__init__(model_config)
config = self.model_config.pretrained_config
self.padding_idx = config.pad_token_id

self.embed_tokens = Embedding(
config.vocab_size,
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_mixtral.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -159,7 +159,6 @@ class MixtralModel(DecoderModel):
def __init__(self, model_config: ModelConfig[PretrainedConfig]):
super().__init__(model_config)
config = model_config.pretrained_config
self.padding_idx = config.pad_token_id
self.vocab_size = config.vocab_size
self.aux_stream = torch.cuda.Stream()

Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_nemotron.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -132,7 +132,6 @@ class NemotronModel(DecoderModel):
def __init__(self, model_config: ModelConfig[NemotronConfig]):
super().__init__(model_config)
config = self.model_config.pretrained_config
self.padding_idx = config.pad_token_id

self.embed_tokens = Embedding(
config.vocab_size,
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_qwen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -118,7 +118,6 @@ class QwenModel(DecoderModel):
def __init__(self, model_config: ModelConfig[Qwen2Config]):
super().__init__(model_config)
config = self.model_config
self.padding_idx = config.pretrained_config.pad_token_id

self.embed_tokens = Embedding(
config.pretrained_config.vocab_size,
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_qwen3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -179,7 +179,6 @@ class Qwen3Model(DecoderModel):
def __init__(self, model_config: ModelConfig[Qwen3Config]):
super().__init__(model_config)
config = self.model_config
self.padding_idx = config.pretrained_config.pad_token_id

self.embed_tokens = Embedding(
config.pretrained_config.vocab_size,
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_qwen3_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -308,7 +308,6 @@ class Qwen3MoEModel(DecoderModel):
def __init__(self, model_config: ModelConfig[Qwen3MoeConfig]):
super().__init__(model_config)
config = self.model_config
self.padding_idx = config.pretrained_config.pad_token_id
self.aux_stream = torch.cuda.Stream()

if model_config.mapping.enable_attention_dp:
Expand Down
1 change: 0 additions & 1 deletion tensorrt_llm/_torch/models/modeling_qwen_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -191,7 +191,6 @@ class QwenMoeModel(DecoderModel):
def __init__(self, model_config: ModelConfig[Qwen2MoeConfig]):
super().__init__(model_config)
config = self.model_config
self.padding_idx = config.pretrained_config.pad_token_id
self.aux_stream = torch.cuda.Stream()

self.embed_tokens = Embedding(
Expand Down