16 commits

Author	SHA1	Message	Date
oobabooga	d5c407cf35	Use Vulkan instead of ROCm for llama.cpp on AMD	2025-05-01 20:05:36 -07:00
oobabooga	c12a53c998	Use turboderp's exllamav2 wheels	2025-05-01 19:46:56 -07:00
oobabooga	a4bf339724	Bump llama.cpp	2025-04-30 11:13:14 -07:00
oobabooga	e9569c3984	Fixes after c5fe92d152	2025-04-30 06:57:23 -07:00
oobabooga	7f49e3c3ce	Bump ExLlamaV3	2025-04-30 05:25:09 -07:00
oobabooga	c5fe92d152	Bump llama.cpp	2025-04-30 05:24:58 -07:00
oobabooga	fa861de05b	Fix portable builds with Python 3.12	2025-04-26 18:52:44 -07:00
oobabooga	bf2aa19b21	Bump llama.cpp	2025-04-26 16:39:22 -07:00
oobabooga	2c7ff86015	Bump exllamav3 to de83084184	2025-04-25 05:28:22 -07:00
oobabooga	5993ebeb1b	Bump exllamav2 to 0.2.9	2025-04-25 05:27:59 -07:00
oobabooga	8ebe868916	Fix typos in b313adf653	2025-04-24 09:32:17 -07:00
oobabooga	b313adf653	Bump llama.cpp, make the wheels work with any Python >= 3.7	2025-04-24 08:26:12 -07:00
oobabooga	06619e5f03	Add vulkan requirements.txt files	2025-04-22 17:46:54 -07:00
oobabooga	ee09e44c85	Portable version (#6868)	2025-04-22 09:25:57 -03:00
oobabooga	c178ea02fe	Revert "Move the requirements*.txt to a requirements folder" ... This reverts commit 6117ef7d64.	2025-04-20 19:27:38 -07:00
oobabooga	6117ef7d64	Move the requirements*.txt to a requirements folder	2025-04-20 19:12:04 -07:00