Dependencies: Update ExllamaV2

v0.0.20

ROCm 6.0 is now required

Signed-off-by: kingbri <[email protected]>

backends/exllamav2/utils.py

@@ -6,7 +6,7 @@
 def check_exllama_version():
     """Verifies the exllama version"""
 
-    required_version = version.parse("0.0.18")
+    required_version = version.parse("0.0.20")
     current_version = version.parse(package_version("exllamav2").split("+")[0])
 
     if current_version < required_version:

pyproject.toml

@@ -48,10 +48,10 @@ cu121 = [
     "torch @ https://download.pytorch.org/whl/cu121/torch-2.3.0%2Bcu121-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
 
     # Exl2
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu121-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu121-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu121-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu121-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu121-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu121-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu121-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu121-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
 
     # Windows FA2 from https://github.com/bdashore3/flash-attention/releases
     "flash_attn @ https://github.com/bdashore3/flash-attention/releases/download/v2.5.6/flash_attn-2.5.6+cu122torch2.2.2cxx11abiFALSE-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
@@ -69,10 +69,10 @@ cu118 = [
     "torch @ https://download.pytorch.org/whl/cu118/torch-2.3.0%2Bcu118-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
 
     # Exl2
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu121-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu118-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu118-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+cu118-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu121-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu118-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu118-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+cu118-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
 
     # Linux FA2 from https://github.com/Dao-AILab/flash-attention/releases
     "flash_attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.5.6/flash_attn-2.5.6+cu118torch2.2cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
@@ -88,8 +88,8 @@ amd = [
     "torch @ https://download.pytorch.org/whl/rocm6.0/torch-2.3.0%2Brocm6.0-cp310-cp310-linux_x86_64.whl ; python_version == '3.10'",
 
     # Exl2
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+rocm5.6-cp311-cp311-linux_x86_64.whl ; python_version == '3.11'",
-    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.19/exllamav2-0.0.19+rocm5.6-cp310-cp310-linux_x86_64.whl ; python_version == '3.10'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+rocm6.0-cp311-cp311-linux_x86_64.whl ; python_version == '3.11'",
+    "exllamav2 @ https://github.com/turboderp/exllamav2/releases/download/v0.0.20/exllamav2-0.0.20+rocm6.0-cp310-cp310-linux_x86_64.whl ; python_version == '3.10'",
 ]
 
 # MARK: Ruff options