10000 add PLaMo model by okdshin · Pull Request #3557 · ggml-org/llama.cpp · GitHub
[go: up one dir, main page]
Skip to content

add PLaMo model #3557

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 19 commits into from
Dec 24, 2023
Merged

add PLaMo model #3557

Changes from 1 commit
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
feb0966
add plamo mock
okdshin Dec 16, 2023
4c585b4
add tensor loading
okdshin Dec 16, 2023
b2330f5
plamo convert
okdshin Dec 17, 2023
9d49236
update norm
okdshin Dec 17, 2023
4a3ef4f
able to compile
okdshin Dec 17, 2023
a22040a
fix norm_rms_eps hparam
okdshin Dec 17, 2023
86d5348
runnable
okdshin Dec 17, 2023
f76fd39
use inp_pos
okdshin Dec 17, 2023
ca8f698
seems ok
okdshin Dec 17, 2023
febc635
update kqv code
okdshin Dec 17, 2023
907b921
remove develop code
okdshin Dec 18, 2023
9339ffc
update README
okdshin Dec 18, 2023
db1b18d
shuffle attn_q.weight and attn_output.weight for broadcasting
okdshin Dec 24, 2023
26340a1
remove plamo_llm_build_kqv and use llm_build_kqv
okdshin Dec 24, 2023
700f7c6
fix style
okdshin Dec 24, 2023
602cc71
Merge branch 'master' into add_pfnet_plamo_13b_2
okdshin Dec 24, 2023
307481f
update
okdshin Dec 24, 2023
eedd434
llama : remove obsolete KQ_scale
ggerganov Dec 24, 2023
1949c95
plamo : fix tensor names for correct GPU offload
ggerganov Dec 24, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
fix style
  • Loading branch information
@okdshin
okdshin committed Dec 24, 2023
commit 700f7c600a47cca3971acdeff6edf2554121a438
13 changes: 5 additions & 8 deletions llama.cpp
View file
Open in desktop
54A7
Original file line number Diff line number Diff line change
Expand Up @@ -5543,7 +5543,7 @@ struct llm_build_context {
cur = llm_build_norm(ctx0, inpL, hparams,
model.layers[il].attn_norm, NULL,
LLM_NORM_RMS, cb, il);
cb(cur, "attention_norm", il);
cb(cur, "attn_norm", il);

struct ggml_tensor * attention_norm = cur;

Expand Down Expand Up @@ -5603,13 +5603,10 @@ struct llm_build_context {

cur = inpL;

// norm
{
cur = llm_build_norm(ctx0, cur, hparams,
model.output_norm, NULL,
LLM_NORM_RMS, cb, -1);
cb(cur, "result_norm", -1);
}
cur = llm_build_norm(ctx0, cur, hparams,
model.output_norm, NULL,
LLM_NORM_RMS, cb, -1);
cb(cur, "result_norm", -1);

// lm_head
cur = ggml_mul_mat(ctx0, model.output, cur);
Expand Down
0