HappyZ happyz
0 Followers · 0 Following
Repositories
18
 Projects Packages Public Activity Starred Repositories
happyz synced commits to refs/pull/20778/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
acc52d5a75 Merge 370cdb9f26 into c46583b86b
370cdb9f26 grammar : fix lazy trigger crash during generation_prompt prefill
Compare 2 commits »
happyz synced commits to refs/tags/b8445 at happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
happyz synced new reference refs/tags/b8445 to happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
happyz synced commits to refs/pull/20775/head at happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
77ff285fbd common : add standard Hugging Face cache support
happyz synced commits to refs/pull/20775/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
84e139639e Merge 77ff285fbd into 21c8045214
77ff285fbd common : add standard Hugging Face cache support
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
Compare 4 commits »
happyz synced commits to refs/pull/20778/head at happyz/llama.cpp from mirror 2026-03-20 07:02:14 -07:00
fba6b87ab2 sampling : handle grammar prefill crash for Functionary v3.2
happyz synced commits to refs/pull/20747/head at happyz/llama.cpp from mirror 2026-03-20 07:02:13 -07:00
0fcc4cc8ba CANN: add RoPE cache preload before ACL graph capture
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
a0bbcdd9b6 ggml: guard KleidiAI DOWNLOAD_EXTRACT_TIMESTAMP for cmake < 3.24 (#20767)
Compare 26 commits »
happyz synced commits to refs/pull/20747/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:13 -07:00
7a63b07c52 Merge 0fcc4cc8ba into c46583b86b
0fcc4cc8ba CANN: add RoPE cache preload before ACL graph capture
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
Compare 4 commits »
happyz synced commits to refs/pull/20758/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:13 -07:00
f6297997f4 Merge ed510c25d0 into 21c8045214
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
Compare 17 commits »
happyz synced commits to refs/pull/20759/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:13 -07:00
3fd7bf402a Merge e6b05bbe5c into 21c8045214
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
Compare 16 commits »
happyz synced commits to refs/pull/20773/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:13 -07:00
c37a428032 Merge 7794488b32 into dc6592431b
dc6592431b context: zero output buffer on allocation (#20781)
3adbef7776 model: assert nextn_predict_layers to prevent underflow (#20783)
ab9d4c3678 server : improve mtmd ctx checkpoints (#20726)
1af9dab32b CANN: add BF16 support for core operators (#20152)
Compare 10 commits »
happyz synced commits to refs/pull/20726/head at happyz/llama.cpp from mirror 2026-03-20 07:02:12 -07:00
92a70fef1a server : fix off-by-one in pos_min_thold
6051df2f2b server : improve mtmd ctx checkpoints
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
Compare 34 commits »
happyz synced commits to refs/pull/20729/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:12 -07:00
5f7923daef Merge 841e09eaf4 into 6d99b44c7e
6d99b44c7e docs : fix Metal backend op support status in ops.md (#20779)
464fd0e71f ai : update find-related action (#20790)
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
Compare 6 commits »
happyz synced commits to refs/pull/20742/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:12 -07:00
4c87057bcf Merge aee5e863ad into 21c8045214
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
Compare 15 commits »
happyz synced commits to refs/pull/20690/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
0cc0558f12 Merge fcad53c54e into ab9d4c3678
ab9d4c3678 server : improve mtmd ctx checkpoints (#20726)
1af9dab32b CANN: add BF16 support for core operators (#20152)
6d99b44c7e docs : fix Metal backend op support status in ops.md (#20779)
464fd0e71f ai : update find-related action (#20790)
Compare 8 commits »
happyz synced commits to refs/pull/20700/head at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
4aeffc690d doc: document MTP attention requirement for higher acceptance
happyz synced commits to refs/pull/20700/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
c5dfcdc1a0 Merge 4aeffc690d into c46583b86b
4aeffc690d doc: document MTP attention requirement for higher acceptance
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
Compare 3 commits »
happyz synced commits to refs/pull/20712/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
3a52ed6101 Merge ce773a7bf6 into dc6592431b
dc6592431b context: zero output buffer on allocation (#20781)
3adbef7776 model: assert nextn_predict_layers to prevent underflow (#20783)
ab9d4c3678 server : improve mtmd ctx checkpoints (#20726)
1af9dab32b CANN: add BF16 support for core operators (#20152)
Compare 11 commits »
happyz synced commits to refs/pull/20716/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
49d0540135 Merge 4af94d9afe into 21c8045214
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
Compare 15 commits »
happyz synced commits to refs/pull/20723/merge at happyz/llama.cpp from mirror 2026-03-20 07:02:11 -07:00
599d8e3c49 Merge 04719ae517 into 21c8045214
21c8045214 jinja : fix heap OOB read in value equality comparison (#20782)
c46583b86b common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
c1b911654a server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
b739738dad docs: Update server README to reflect PR #20297 (#20560)
Compare 8 commits »