llama.cpp
llama : add llm_build helper functions
#3848
Merged
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Overview
Commits
18
Changes
View On
GitHub
llama : add llm_build helper functions
#3848
ggerganov
merged 18 commits into
llama-refactor
from
llama-refactor-norm
ggerganov
added
refactoring
ggerganov
added
need feedback
llama : add llm_build_norm helper function
7db9c96d
ggerganov
force pushed
from
66928991
to
7db9c96d
2 years ago
llama : add llm_build_ffn helper function (#3849)
dbf836bb
ggerganov
changed the title
llama : add llm_build_norm helper function
llama : add llm_build helper functions
2 years ago
llama : add llm_build_k_shift helper
38728a0b
ggerganov
force pushed
from
c6ae530e
to
38728a0b
2 years ago
llama : fix offloading after recent changes
909d6447
ggerganov
force pushed
from
61ca4777
to
c82880d2
2 years ago
ggerganov
force pushed
from
c82880d2
to
85ec6b02
2 years ago
ggerganov
force pushed
from
85ec6b02
to
ca15e4ac
2 years ago
llama : add llm_build_kv_store helper
3e046259
ggerganov
force pushed
from
ca15e4ac
to
3e046259
2 years ago
llama : remove obsolete offload names
59908619
llama : fix llm_build_k_shift to use n_head_kv instead of n_head
31a12f3d
llama : simplify falcon Q, K, V computation
a104abea
llama : remove obsolete comments in build graphs
c9121fdd
llama : add llm_build_kqv helper
f39e6075
ggerganov
force pushed
from
d82a5c02
to
f39e6075
2 years ago
ggerganov
marked this pull request as ready for review
2 years ago
llama : minor
792d1a1b
ggerganov
force pushed
from
4d115ea1
to
792d1a1b
2 years ago
llama : add LLAMA_OFFLOAD_DEBUG + fix starcoder offloading
a3f80013
KerfuffleV2
commented on 2023-10-30
llama : fix input allocation logic
2926ef63
llama : update offload functions for KQ tensors
6669cd83
llama : normalize tensor names
0bfdcdd0
llama : enable warning about not offloaded tensors
fc5a26aa
ggerganov
requested a review
from
slaren
2 years ago
ggerganov
requested a review
from
KerfuffleV2
2 years ago
slaren
commented on 2023-10-31
slaren
commented on 2023-10-31
slaren
commented on 2023-10-31
slaren
commented on 2023-10-31
slaren
commented on 2023-10-31
KerfuffleV2
commented on 2023-10-31
llama : remove extra ; + deduplicate gate_b logic
2073347e
llama : add llm_build_inp_embd helper
7923b70c
ggerganov
merged
5baefef4
into llama-refactor
2 years ago
Login to write a write a comment.
Login via GitHub
Reviewers
slaren
KerfuffleV2
Assignees
No one assigned
Labels
refactoring
need feedback
Milestone
No milestone
Login to write a write a comment.
Login via GitHub