Justine Tunney
ca19ecf49c
Fine tune crash reports for llama.com
2023-05-12 06:24:26 -07:00
Justine Tunney
95fab334e4
Use yield on aarch in spin locks
2023-05-11 19:57:09 -07:00
Justine Tunney
1f6f9e6701
Remove division from matrix multiplication
...
This change reduces llama.com CPU cycles systemically by 2.5% according
to the Linux Kernel `perf stat -Bddd` utility.
2023-05-10 21:19:54 -07:00
Justine Tunney
a88290e595
Make sure llama.com terminal cleanup happens
2023-05-10 15:56:01 -07:00
Justine Tunney
bb3ebedfce
Fix load time measurement
2023-05-10 07:54:21 -07:00
Justine Tunney
290a49952e
Fix some more issues with aarch64 and llama.cpp
2023-05-10 07:34:26 -07:00
Justine Tunney
6cb9553706
Fix alignment bug in llama.com
2023-05-10 06:15:32 -07:00
Justine Tunney
ca990ef091
Make llama.com -h print to stdout
2023-05-10 04:55:59 -07:00
Justine Tunney
5f57fc1f59
Upgrade llama.cpp to e6a46b0ed1884c77267dc70693183e3b7164e0e0
2023-05-10 04:20:48 -07:00
Justine Tunney
a0237a017c
Get llama.com working on aarch64
2023-05-10 04:20:47 -07:00
Justine Tunney
4c093155a3
Get llama.com building as an aarch64 native binary
2023-05-10 04:20:47 -07:00
Justine Tunney
d04430f4ef
Get LIBC_MEM and LIBC_STDIO building with aarch64
2023-05-10 04:20:47 -07:00
Justine Tunney
2b73e72d59
Make more code aarch64 friendly
2023-05-10 04:20:46 -07:00
Justine Tunney
3dac9f8999
Use Companion AI in llama.com by default
2023-04-30 23:08:15 -07:00
Justine Tunney
d9e27203d4
Incorporate some fixes and updates for GGML
2023-04-28 20:24:55 -07:00
Justine Tunney
b31ba86ace
Introduce prompt caching so prompts load instantly
...
This change also introduces an ephemeral status line in non-verbose mode
to display a load percentage status when slow operations are happening.
2023-04-28 16:15:26 -07:00
Justine Tunney
1c2da3a55a
Make shell usability improvements to llama.cpp
...
- Introduce -v and --verbose flags
- Don't print stats / diagnostics unless -v is passed
- Reduce --top_p default from 0.95 to 0.70
- Change --reverse-prompt to no longer imply --interactive
- Permit --reverse-prompt specifying custom EOS if non-interactive
2023-04-28 02:54:11 -07:00
Justine Tunney
420f889ac3
Further optimize the math library
...
The sincosf() function is now twice as fast, thanks to ARM Limited. The
same might also be true of logf() and expm1f() which have been updated.
2023-04-28 01:20:47 -07:00
Justine Tunney
e8b43903b2
Import llama.cpp
...
https://github.com/ggerganov/llama.cpp
0b2da20538d01926b77ea237dd1c930c4d20b686
See third_party/ggml/README.cosmo for changes
2023-04-27 14:37:14 -07:00
