Commit Graph

450 Commits

Author SHA1 Message Date
8c8cf38d4d tests: use 1 core 2023-05-05 23:29:34 +02:00
75b25297fd tests: run with ginkgo 2023-05-05 22:51:30 +02:00
009ee47fe2 Don't allow 0 as thread count 2023-05-05 22:51:20 +02:00
ec2adc2c03 tests: use 3 cores 2023-05-05 22:07:01 +02:00
ad301e6ed7 example(add): document query example 2023-05-05 21:56:31 +02:00
d094381e5d ci: lower fixtures spec 2023-05-05 21:28:38 +02:00
3ff9bbd217 examples: add rwkv script folder 2023-05-05 19:04:52 +02:00
e62ee2bc06 fix: remove trailing 0s from embeddings
This happens when no max_tokens are set, so by default go-llama
allocates more space for the slice and padding happens.
2023-05-05 18:35:03 +02:00
b49721cdd1 fix: respect config from file for backends settings 2023-05-05 18:05:10 +02:00
64c0a7967f fix: pass prediction options when using the model 2023-05-05 15:56:02 +02:00
e96eadab40 feat: support deprecated embeddings API 2023-05-05 15:55:19 +02:00
e73283121b feat: support arrays for prompt and input
Signed-off-by: mudler <mudler@mocaccino.org>
2023-05-05 15:54:59 +02:00
857d13e8d6 debug: wire up go-fiber debugger 2023-05-05 15:53:57 +02:00
91db3d4d5c ⬆️ Update go-skynet/go-llama.cpp (#194)
Signed-off-by: GitHub <noreply@github.com>
Co-authored-by: mudler <mudler@users.noreply.github.com>
2023-05-05 13:45:50 +02:00
961cf29217 feat: expose mirostat to config (#193) 2023-05-05 13:45:37 +02:00
c839b334eb feat: add embeddings for go-llama.cpp backend (#190) 2023-05-05 11:20:06 +02:00
714bfcd45b fix: missing returning error and free callback stream (#187) v1.6.3 2023-05-04 19:49:43 +02:00
77ce8b953e fix(deps): update github.com/donomii/go-rwkv.cpp digest to af62fcc (#171)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-04 18:30:48 +02:00
01ada95941 fix(deps): update github.com/go-skynet/go-llama.cpp digest to 2e6ae12 (#172)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-04 18:30:11 +02:00
eabdc5042a ⬆️ Update go-skynet/go-llama.cpp (#184)
Signed-off-by: GitHub <noreply@github.com>
Co-authored-by: mudler <mudler@users.noreply.github.com>
2023-05-04 18:28:49 +02:00
96267d9437 localai: Include the WebUI project example (#130)
Co-authored-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
2023-05-04 18:27:58 +02:00
9497a24127 fix: hardcode default number of cores to '4' (#186) v1.6.2 2023-05-04 18:14:58 +02:00
fdf75c6d0e rwkv fixes and examples (#185) 2023-05-04 17:32:23 +02:00
6352308882 ci: minor fixups 2023-05-04 15:08:20 +02:00
a8172a0f4e ci: fix typo 2023-05-04 15:04:41 +02:00
ebcd10d66f ci: manually update deps 2023-05-04 15:01:29 +02:00
885642915f ci: add renovate suffix 2023-05-04 12:26:59 +02:00
2e424491c0 ci: lookupNameTemplate -> depNameTemplate 2023-05-04 12:23:05 +02:00
aa6faef8f7 ci: versioning -> versioningTemplate 2023-05-04 12:07:29 +02:00
b3254baf60 ci: add versioning 2023-05-04 12:05:39 +02:00
0a43d27f0e ci: update renovate 2023-05-04 12:02:19 +02:00
3fe11fe24d ci: attempt to configure renovate with custom regexes (#178) 2023-05-04 11:55:14 +02:00
af18fdc749 fix(deps): update module github.com/sashabaranov/go-openai to v1.9.3 (#174)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-04 08:44:02 +02:00
32b5eddd7d fix(deps): update module github.com/onsi/ginkgo/v2 to v2.9.4 (#173)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-04 08:41:51 +02:00
07c3aa1869 Dockerized Langchain / PY example (#175) 2023-05-04 08:41:13 +02:00
e59bad89e7 fix(deps): update module github.com/sashabaranov/go-openai to v1.9.2 (#164)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-03 23:05:50 +02:00
b971807980 Looks for models in $CWD/models/ dir by default (#169) 2023-05-03 23:03:31 +02:00
c974dad799 Return usage in the API responses (#166) v1.6.1 2023-05-03 17:29:18 +02:00
4eae570ef5 Update docs (#163) v1.6.0 2023-05-03 15:51:54 +02:00
67992a7d99 feat: support slices or strings in the prompt completion endpoint (#162)
Signed-off-by: mudler <mudler@mocaccino.org>
2023-05-03 13:13:31 +02:00
0a4899f366 fix(deps): update github.com/go-skynet/go-llama.cpp digest to 8ceb616 (#150)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-03 11:48:06 +02:00
1eb02f6c91 fix(deps): update module github.com/onsi/ginkgo/v2 to v2.9.3 (#161)
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2023-05-03 11:47:54 +02:00
575874e4fb readme: minor update 2023-05-03 11:46:29 +02:00
751b7eca62 feat: add rwkv support (#158)
Signed-off-by: mudler <mudler@mocaccino.org>
2023-05-03 11:45:22 +02:00
1ae7150810 feat: allow to specify default backend for model (#156)
Signed-off-by: mudler <mudler@c3os.io>
2023-05-03 00:31:28 +02:00
70caf9bf8c feat: support stopwords both string and arrays (#154) 2023-05-02 23:30:00 +02:00
0b226ac027 Stop parameter of OpenAIRequest changed to String Array (#153) 2023-05-02 22:02:45 +02:00
220d6fd59b feat: add stream events (#152) 2023-05-02 20:03:35 +02:00
0a00a4b58e adding mac build and example (#151)
Co-authored-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
2023-05-02 19:24:45 +02:00
156e15a4fa Bump llama.cpp, downgrade gpt4all-j (#149) 2023-05-02 16:07:18 +02:00