ggerganov/llama.cpp

Pull Requests Commits

awq-py : remove (#5768)

ggerganov committed 1 year ago

Verified 78aacf36

ggerganov committed 1 year ago

Verified 8c0e8f4e

add google magika inference example (ggml/748)

slaren committed 1 year ago

Verified 2774b0c9

Introduce backend GUIDs (ggml/743)

uextm committed 1 year ago

Verified 5f706718

server : hit Ctrl+C twice to exit (#5734)

ngxson committed 1 year ago

Verified a693bea1

llama : fix non-quantization of expert gating tensors (#5754)

compilade committed 1 year ago

Verified adcb12a9

llama : improve BERT tokenization (#5740)

iamlemec committed 1 year ago

Verified 177628bf

readme : add link to LLaVA 1.6 models (#5758)

danbev committed 1 year ago

Verified 6c441686

server : add "/chat/completions" alias for "/v1/...` (#5722)

jorgealias committed 1 year ago

Verified efc72253

ggml : make i-quants work with super-blocks of 64 (CPU,Metal) (#5760)

ikawrakow committed 1 year ago

Verified 7c4263d4

Attempt to fix android build (#5752)

ikawrakow committed 1 year ago

Verified cb49e0f8

IQ4_XS: a 4.25 bpw quantization (#5747)

ikawrakow committed 1 year ago

Verified 0becb22a

cuda : replace remaining shfl_xor with calls to warp_reduce functions (#5744)

Engininja2 committed 1 year ago

Verified c24a2a6e

ggml-quants : fix avx2 iq1_s vec_dot when compiled with gcc (#5742)

Engininja2 committed 1 year ago

Verified 1f30b7a9

llama : fix defrag bugs + add parameter (#5735)

ggerganov committed 1 year ago

Verified 9d533a77

Makefile: use variables for cublas (#5689)

lindeer committed 1 year ago

Verified cbbd1efa

fix server hangs on empty prompt (#5733)

ngxson committed 1 year ago

Verified b11a93df

Adding IQ2_S and IQ2_M to complete coverage of the 2-3 bit quantization range (#5721)

ikawrakow committed 1 year ago

Verified a33e6a0d

CUDA: fix DEBUG_CUDA_MALLOC (#5729)

JohannesGaessler committed 1 year ago

Verified 47bb7b48

readme : update ui list (#5731)

guinmoon committed 1 year ago

Verified c4d7f817

[SYCL] Add support for soft_max ALiBi (#5639)

AidanBeltonS committed 1 year ago

Verified e849078c

unicode : reuse iterator (#5726)

ggerganov committed 1 year ago

Verified 67fd3313

server: CI fix trailing space (#5728)

phymbert committed 1 year ago

Verified 4804215c

server: CI tests reduce build matrix (#5725)

phymbert committed 1 year ago

Verified 8a533f0d

llama : fix Gemma rope type (#5691)

ggerganov committed 1 year ago

Verified 269de86b

flake.lock: Update

github-actions[bot] committed 1 year ago

c3937339

server: tests - slow inference causes timeout on the CI (#5715)

phymbert committed 1 year ago

Verified e3965cf3

server: docs - refresh and tease a little bit more the http server (#5718)

phymbert committed 1 year ago

Verified 8b350356

llama : refactor k-shift implementation + KV defragmentation (#5691)

ggerganov committed 1 year ago

Verified bf08e006

server : fix crash when system prompt is bigger than batch size (#5714)

compilade committed 1 year ago

Verified f7625019

Newer Older