diff nicer

server/text_generation_server/models/flash_cohere.py

@@ -28,7 +28,6 @@ class FlashCohere(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_dbrx.py

@@ -30,7 +30,6 @@ class FlashDbrx(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_gemma.py

@@ -28,7 +28,6 @@ class FlashGemma(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_gpt2.py

@@ -30,7 +30,6 @@ class FlashGPT2(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_llama.py

@@ -32,7 +32,6 @@ class FlashLlama(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_neox.py

@@ -29,7 +29,6 @@ class FlashNeoXSharded(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_phi.py

@@ -29,7 +29,6 @@ class FlashPhi(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_qwen2.py

@@ -34,7 +34,6 @@ class FlashQwen2(BaseFlashMistral):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_rw.py

@@ -30,7 +30,6 @@ class FlashRWSharded(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_santacoder.py

@@ -33,7 +33,6 @@ class FlashSantacoderSharded(FlashCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/flash_starcoder2.py

@@ -33,7 +33,6 @@ class FlashStarcoder2(BaseFlashMistral):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/gpt_neox.py

@@ -28,7 +28,6 @@ class GPTNeoxSharded(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/idefics.py

@@ -35,7 +35,6 @@ class IDEFICSSharded(IdeficsCausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/mpt.py

@@ -47,7 +47,6 @@ class MPTSharded(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/opt.py

@@ -26,7 +26,6 @@ class OPTSharded(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")

server/text_generation_server/models/phi.py

@@ -26,7 +26,6 @@ class Phi(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, _rank, _world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device("cuda")

server/text_generation_server/models/rw.py

@@ -16,7 +16,6 @@ class RW(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         if speculator:
             raise RuntimeError("Medusa decoding is not enabled for AutoModel")
 

server/text_generation_server/models/santacoder.py

@@ -23,7 +23,6 @@ class SantaCoder(CausalLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         if torch.cuda.is_available():
             device = torch.device("cuda")
             dtype = torch.float16 if dtype is None else dtype

server/text_generation_server/models/seq2seq_lm.py

@@ -537,7 +537,6 @@ class Seq2SeqLM(Model):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         if speculator:
             raise RuntimeError("Speculator decoding is not enabled for AutoModel")
 

server/text_generation_server/models/t5.py

@@ -29,7 +29,6 @@ class T5Sharded(Seq2SeqLM):
         dtype: Optional[torch.dtype] = None,
         trust_remote_code: bool = False,
     ):
-
         self.process_group, rank, world_size = initialize_torch_distributed()
         if torch.cuda.is_available():
             device = torch.device(f"cuda:{rank}")