ngxson/llama.cpp

Pull Requests Commits
CANN: RoPE operator optimization (#10563)
noemotiovon committed 1 year ago
Verified 938f6087
vulkan: get the first command buffer submitted sooner (#10499)
jeffbolznv committed 1 year ago
Verified f095a649
llava: return false instead of exit (#10546)
tinglou committed 1 year ago
Verified 678d7994
ggml : remove redundant copyright notice + update authors
ggerganov committed 1 year ago
Verified dc223440
llama : add missing model types
ggerganov committed 1 year ago
Verified 4c0a95b1
server : (tests) don't use thread for capturing stdout/stderr, bump openai client library (#10568)
ngxson committed 1 year ago
Verified 6c595676
common: fix warning message when no GPU found (#10564)
JohannesGaessler committed 1 year ago
Verified 89071931
docs: fix outdated usage of llama-simple (#10565)
rand-fly committed 1 year ago
Verified 7281cf13
ci : fix tag name in cuda and hip releases (#10566)
slaren committed 1 year ago
Verified e90688ed
ggml : fix row condition for i8mm kernels (#10561)
ggerganov committed 1 year ago
Verified 76b27d29
cmake : fix ARM feature detection (#10543)
ggerganov committed 1 year ago
Verified eea986f2
ggml-cpu: support IQ4_NL_4_4 by runtime repack (#10541)
FanShupei committed 1 year ago
Verified c202cef1
kompute : improve backend to pass test_backend_ops (#10542)
slp committed 1 year ago
Verified 2025fa67
CANN: Update cann.md to display correctly in CLion (#10538)
HRXWEB committed 1 year ago
Verified c6bc7395
CANN: Fix SOC_TYPE compile bug (#10519)
leo-pony committed 1 year ago
Verified 605fa66c
CANN: ROPE operator optimization (#10540)
noemotiovon committed 1 year ago
Verified b7420131
common : fix duplicated file name with hf_repo and hf_file (#10550)
ngxson committed 1 year ago
Verified 9f912511
Add some minimal optimizations for CDNA (#10498)
IMbackK committed 1 year ago
Verified 3ad5451f
ci : faster CUDA toolkit installation method and use ccache (#10537)
slaren committed 1 year ago
Verified 46c69e0e
metal : fix group_norm support condition (#0)
ggerganov committed 1 year ago
Verified 9e2301f4
sync : ggml
ggerganov committed 1 year ago
Verified fee824a1
Do not include arm_neon.h when compiling CUDA code (ggml/1028)
frankier committed 1 year ago
Verified 9150f8fe
vulkan: define all quant data structures in types.comp (#10440)
jeffbolznv committed 1 year ago
Verified c31ed2ab
vulkan: Handle GPUs with less shared memory (#10468)
jeffbolznv committed 1 year ago
Verified 5b3466be
vulkan: further optimize q5_k mul_mat_vec (#10479)
jeffbolznv committed 1 year ago
Verified 249a7902
vulkan: skip integer div/mod in get_offsets for batch_idx==0 (#10506)
jeffbolznv committed 1 year ago
Verified 71a64989
vulkan: optimize Q2_K and Q3_K mul_mat_vec (#10459)
jeffbolznv committed 1 year ago
Verified 4a57d362
ci : fix cuda releases (#10532)
slaren committed 1 year ago
Verified c9b00a70
Add OLMo 2 model in docs (#10530)
2015aroras committed 1 year ago
Verified de509735
ci : remove nix workflows (#10526)
slaren committed 1 year ago
Verified 5a349f28
Terms Of Use Privacy Policy Refund Policy Impressum