Tags
Tags give the ability to mark specific points in history as being important
master-2e6cd4b
2e6cd4b0
·
OpenCL Token Generation Acceleration (#1459)
·
May 23, 2023
master-7e4ea5b
7e4ea5be
·
examples : add server example with REST API (#1443)
·
May 21, 2023
master-7780e4f
7780e4f4
·
make : .PHONY clean (#1553)
·
May 21, 2023
master-265db98
265db983
·
ggml : output 3d sizes in ggml_graph_dump_dot()
·
May 21, 2023
master-fab49c6
fab49c68
·
ggml : update WASM SIMD
·
May 20, 2023
master-b8ee340
b8ee340a
·
feature : support blis and other blas implementation (#1536)
·
May 20, 2023
master-9ecb30f
9ecb30f9
·
OpenCL: Fixes for older devices. (#1435)
·
May 20, 2023
master-29cf559
29cf5596
·
llama : define magic numbers as integer constants (#1518) (#1520)
·
May 20, 2023
master-3de84b2
3de84b26
·
ggml : add ggml_clamp() (#1539)
·
May 20, 2023
master-affc76e
affc76ed
·
cuda : loading models directly into VRAM, norm calculation on GPU,...
·
May 20, 2023
master-ea60007
ea600071
·
Revert "feature : add blis and other BLAS implementation support (#1502)"
·
May 20, 2023
master-07e9ace
07e9ace0
·
feature : add blis and other BLAS implementation support (#1502)
·
May 20, 2023
master-ec2e10c
ec2e10c4
·
llama : add llama_init_backend() API (close #1527)
·
May 20, 2023
master-d2c59b8
d2c59b8b
·
Fix for mingw (#1462)
·
May 20, 2023
master-8a203f9
8a203f9f
·
llama : fix compile warnings in llama_set_state_data()
·
May 20, 2023
master-2d5db48
2d5db483
·
ggml : use F16 instead of F32 in Q4_0, Q4_1, Q8_0 (#1508)
·
May 19, 2023
master-6986c78
6986c783
·
tests : add missing header
·
May 19, 2023
master-5ea4339
5ea43392
·
make kv_f16 the default for api users (#1517)
·
May 18, 2023
master-ee96541
ee965413
·
Fixes #1511 lambda issue for w64devkit (mingw) (#1513)
·
May 18, 2023
master-dc271c5
dc271c52
·
Remove unused n_parts parameter (#1509)
·
May 17, 2023
1
…
18
19
20
21
22
23
24
25
26
…
38