fix: lint and refactor import check and avoid model enum as global names

server/text_generation_server/layers/bnb.py

@@ -1,5 +1,4 @@
 from dataclasses import dataclass
-from functools import lru_cache
 
 import bitsandbytes as bnb
 import torch

server/text_generation_server/layers/fp8.py

@@ -12,17 +12,23 @@ from text_generation_server.utils.weights import (
     Weights,
 )
 from text_generation_server.utils.log import log_master, log_once
+import importlib.util
+
 
 FBGEMM_MM_AVAILABLE = False
 FBGEMM_DYN_AVAILABLE = False
-try:
-    import fbgemm_gpu.experimental.gen_ai
 
+
+def is_fbgemm_gpu_available():
+    return importlib.util.find_spec("fbgemm_gpu.experimental.gen_ai") is not None
+
+
+if is_fbgemm_gpu_available():
     if SYSTEM == "cuda":
         major, _ = torch.cuda.get_device_capability()
         FBGEMM_MM_AVAILABLE = major == 9
         FBGEMM_DYN_AVAILABLE = major >= 8
-except (ImportError, ModuleNotFoundError):
+else:
     log_master(logger.warning, "FBGEMM fp8 kernels are not installed.")
 
 

server/text_generation_server/layers/rotary.py

@@ -4,7 +4,6 @@ import torch
 from torch import nn
 
 # Inverse dim formula to find dim based on number of rotations
-import math
 
 
 from text_generation_server.utils.import_utils import SYSTEM

server/text_generation_server/models/__init__.py

@@ -298,10 +298,8 @@ class ModelType(enum.Enum):
         "multimodal": True,
     }
 
-
+    def __str__(self):
-__GLOBALS = locals()
+        return self.value["type"]
-for data in ModelType:
-    __GLOBALS[data.name] = data.value["type"]
 
 
 def get_model(
@@ -716,6 +714,7 @@ def get_model(
         or model_type == ModelType.BAICHUAN
         or model_type == ModelType.PHI3
     ):
+        print(f">>> model_type: {model_type}")
         if FLASH_ATTENTION:
             return FlashCausalLM(
                 model_id=model_id,

server/text_generation_server/models/custom_modeling/flash_llama_modeling.py

@@ -46,7 +46,6 @@ from text_generation_server.layers.layernorm import (
     FastRMSNorm,
 )
 from text_generation_server.utils.weights import (
-    UnquantizedWeight,
     Weights,
 )
 from text_generation_server.layers.fp8 import HybridFP8UnquantLoader

server/text_generation_server/utils/quantization.py

@@ -6,7 +6,6 @@ from typing import Optional
 from huggingface_hub import hf_hub_download
 from text_generation_server.utils.weights import (
     DefaultWeightsLoader,
-    UnquantizedWeight,
     WeightsLoader,
 )