Convert formatting to use ruff instead of yapf + isort (#26247)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 07:06:22 -07:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115205 additions and 94107 deletions
+9 -10
View File
@@ -28,12 +28,14 @@ import torch.nn as nn
from vllm.config import VllmConfig
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.models.llama import (LlamaDecoderLayer,
LlamaForCausalLM, LlamaModel)
from vllm.model_executor.models.llama import (
LlamaDecoderLayer,
LlamaForCausalLM,
LlamaModel,
)
class TeleFLMModel(LlamaModel):
def __init__(
self,
*,
@@ -41,9 +43,7 @@ class TeleFLMModel(LlamaModel):
prefix: str = "",
layer_type: type[nn.Module] = LlamaDecoderLayer,
):
super().__init__(vllm_config=vllm_config,
prefix=prefix,
layer_type=layer_type)
super().__init__(vllm_config=vllm_config, prefix=prefix, layer_type=layer_type)
"""
This implementation is based on the µScaling paper presented at
the ICLR 2025 Workshop:
@@ -65,7 +65,6 @@ class TeleFLMModel(LlamaModel):
class TeleFLMForCausalLM(LlamaForCausalLM):
def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
super().__init__(vllm_config=vllm_config, prefix=prefix)
# mup
@@ -74,6 +73,6 @@ class TeleFLMForCausalLM(LlamaForCausalLM):
self.mup_scale_factor = self.config.mup_scale_factor
self.output_mult = self.config.output_mult / self.mup_scale_factor
logit_scale = self.output_mult
self.logits_processor = LogitsProcessor(self.unpadded_vocab_size,
self.config.vocab_size,
logit_scale)
self.logits_processor = LogitsProcessor(
self.unpadded_vocab_size, self.config.vocab_size, logit_scale
)