Actions: ORippler/llama.cpp
Actions
35 workflow runs
35 workflow runs
two_stage_warp_reduce also in softmax kernel, move smem out of it
Python Type-Check
#10:
Commit 1ebe58d
pushed
by
ORippler
compare_token_data
Python Type-Check
#9:
Commit da27c9b
pushed
by
ORippler
compare_token_data
Python Type-Check
#8:
Commit 7ee23c0
pushed
by
ORippler
.clang-format to use BinPackArguments=true
Python Type-Check
#4:
Commit e7011e3
pushed
by
ORippler
block_size values in rms_norm_f32
Python Type-Check
#3:
Commit bcc6c77
pushed
by
ORippler