llama.cpp
Add backend sampler for penalties sampler
#25262
Merged

Add backend sampler for penalties sampler #25262

kmorennv
kmorennv sampling: enhance penalty handling in common_sampler_init
0ce6faf5
kmorennv tests: add backend penalties sampling tests and utility functions
e82f0e7d
kmorennv sampling: add support for top-k penalties in backend sampling
af9cf9c4
kmorennv sampling: add fix to ensure stable numerical results. Preserve maske…
3ffeaeae
kmorennv sampling: enhance penalty comparison tests with masking penalties logic
888f264d
kmorennv add comments on padding
1834d9f6
kmorennv sampling: add comments on modifications
38ad11c7
kmorennv kmorennv requested a review from ggerganov ggerganov 102 days ago
kmorennv kmorennv requested a review 102 days ago
github-actions github-actions added testing
ggml-gh-bot
kmorennv
pwilkin
pwilkin
kmorennv
ORippler
ORippler commented on 2026-07-07
ORippler ORippler assigned ORippler ORippler 97 days ago
ORippler ORippler changed the title Kmoren/add penalties cu backend Add backend sampler for penalties sampler 96 days ago
kmorennv kmorennv changed the title Add backend sampler for penalties sampler CUDA: Add backend sampler for penalties sampler 95 days ago
kmorennv add the unit test to cover masked-out token as -INF
d7b05513
kmorennv validate repeat penalty to ensure it is finite and greater than 0; ad…
2adb9571
kmorennv refactor: test functions to share logic and be less verbose
760b7e55
kmorennv add test to cover case where previously penalized token is not part o…
65180ab4
kmorennv remove comments
98b1bea5
kmorennv remove redundant penalty_last_n initialization and validation in comm…
381226d7
kmorennv add support for penalties in sampler chain with configurable positions
91341fe1
kmorennv add validation for penalty parameters and enhance tests for non-finit…
2e54fe32
kmorennv add context parameter to common_sampler_init and set default for pena…
ec2f07a4
kmorennv add llama_n_ctx parameter to common_sampler_init for improved sampler…
e650b603
kmorennv replace penalty_last_n x n_candidates comparison matrix with a vocabu…
cd20503d
kmorennv kmorennv requested a review 91 days ago
kmorennv kmorennv requested a review from CISC CISC 91 days ago
github-actions github-actions added server
kmorennv
kmorennv add tests for backend penalties sampling without filler entries , tok…
e124d900
kmorennv add test for backend penalties sampling after top-p with large histo…
39cabdce
kmorennv Merge branch 'master' into kmoren/add_penalties_cu_backend
b5757809
kmorennv remove as unused
84326a2e
ORippler
ORippler commented on 2026-07-16
ORippler
ORippler commented on 2026-07-20
ORippler ORippler requested a review from danbev danbev 85 days ago
danbev
danbev commented on 2026-07-29
danbev
danbev commented on 2026-07-29
danbev
danbev commented on 2026-07-29
kmorennv add is_disabled method, tensor logits reshape, add rest review sugges…
6e0e28d1
kmorennv clarify comment
3eb4dfc2
danbev
danbev approved these changes on 2026-07-31
ORippler
ORippler approved these changes on 2026-07-31
ORippler ORippler merged 96278e39 into master 71 days ago
ORippler ORippler changed the title CUDA: Add backend sampler for penalties sampler Add backend sampler for penalties sampler 71 days ago
ggerganov
ggerganov commented on 2026-08-03
ggerganov
ggerganov commented on 2026-08-03
ggerganov

Login to write a write a comment.

Login via GitHub

Assignees
Labels
Milestone