Commit Graph

102 Commits

Author SHA1 Message Date
250752c620 fix: new OpenAi api 2024-06-12 21:46:14 +08:00
josc146
3488d22d22 bump webgpu(python) (https://github.com/cryscan/web-rwkv-py) 2024-05-28 21:27:10 +08:00
josc146
d66fd89947 improve dynamic state api 2024-05-16 13:50:48 +08:00
josc146
b24a18cd3a fix a tps error 2024-05-16 13:48:06 +08:00
josc146
a2bbbabee2 add support for dynamic state-tuned models 2024-05-12 21:51:24 +08:00
josc146
2ddcd17d23 add tps console output 2024-05-10 16:19:21 +08:00
josc146
aae7cfe1a2 change the default value of presystem to false 2024-04-30 22:30:06 +08:00
josc146
2d853f92b9 small fix 2024-04-30 21:52:47 +08:00
josc146
16f2201d9f new chat template for /chat/completions (better system support) 2024-03-25 12:52:40 +08:00
josc146
dd89041f72 dep_check.py now ignores GPUtil 2024-03-11 18:55:37 +08:00
josc146
c9513822c9 fix the issue where state cache could be modified leading to inconsistent hit results 2024-03-01 13:35:16 +08:00
josc146
a1ae71d221 fix /update-config can make the default value of unclearly specified fields invalid by passing in None fields 2024-02-05 22:27:02 +08:00
josc146
843840baa0 expose penalty_decay, top_k 2024-02-03 22:03:10 +08:00
josc146
5f94d86558 add better custom tokenizer support and tokenizer-midipiano.json 2024-02-03 13:04:13 +08:00
josc146
7e2380e4ed fix body.state 2023-12-28 23:53:58 +08:00
josc146
7f3cfd54b0 improve state cache performance 2023-12-28 22:15:31 +08:00
josc146
e083f2c629 webgpu(python) state cache 2023-12-28 20:43:57 +08:00
josc146
512c4d0f73 improve role-playing effect 2023-12-22 10:51:09 +08:00
josc146
0ddd2e9fea add WebGPU Python Mode (https://github.com/cryscan/web-rwkv-py) 2023-12-14 18:37:07 +08:00
josc146
e0bf44d82f bump MIDI-LLM-tokenizer (fix note off) 2023-12-14 13:33:27 +08:00
josc146
26b30f0dbe add load failed traceback 2023-12-12 23:16:48 +08:00
josc146
b14fbc29b7 rwkv.cpp(ggml) support 2023-12-12 20:29:55 +08:00
josc146
9b7b651ef9 feat: import midi file 2023-12-10 22:38:31 +08:00
josc146
d9e25ad69f better state cache 2023-12-08 15:28:33 +08:00
josc146
dbf0dccc9d add tokenizer(/switch-model) to /docs 2023-11-20 20:11:45 +08:00
josc146
c8470e77fd fix state_cache of deploy mode 2023-11-17 21:32:11 +08:00
josc146
9ede7d7c6d strict default_stop 2023-11-17 21:18:52 +08:00
josc146
7235e1067b add deployment mode. If /switch-model with deploy: true, will disable /switch-model, /exit and other dangerous APIs (state cache APIs, part of midi APIs) 2023-11-08 23:29:42 +08:00
josc146
cfa3669f6f fix /docs default api params (Pydantic v2) 2023-11-07 22:53:11 +08:00
josc146
ff7306349a improve memory usage of state cache 2023-10-28 23:04:49 +08:00
josc146
c87de93498 allow conversation with some document (.pdf, .txt) 2023-10-27 11:36:29 +08:00
josc146
faf1852012 update stop strategy 2023-10-26 17:47:40 +08:00
josc146
627a20936d RWKVType now no longer relies on the file name 2023-10-26 16:55:33 +08:00
josc146
df9e1f408e add /file-to-text api 2023-10-25 17:14:33 +08:00
josc146
46b3b285f5 upgrade packages 2023-10-25 17:07:40 +08:00
josc146
a25965530c custom tokenizer (#77) 2023-09-16 00:34:11 +08:00
josc146
d7dcc90008 chore 2023-09-15 16:31:14 +08:00
josc146
a24b78e8c3 python-backend: extra ChatCompletionBody params (raw, presystem);
add default_stop when stop is null
2023-08-27 21:21:11 +08:00
josc146
c8025f1cff allow message content to be empty 2023-08-27 21:02:54 +08:00
josc146
02d5d641d1 chore 2023-08-24 22:48:54 +08:00
josc146
61cea2a784 add misc API (/models and /dashboard/billing/credit_grants) 2023-08-14 23:37:55 +08:00
josc146
da68926e9c chore (AddStateBody class) 2023-08-13 21:27:29 +08:00
josc146
e0b7453883 allow multiple systems 2023-08-04 22:27:55 +08:00
josc146
91e2828a95 allow completions input to be null 2023-08-04 22:22:59 +08:00
josc146
b3e35a4cdd allow custom user_name and assistant_name (/chat/completions API) 2023-07-31 22:48:54 +08:00
josc146
aecacde819 remove response field of completions api 2023-07-29 19:20:43 +08:00
josc146
3ef22239eb improve default ChatCompletion stop 2023-07-29 19:19:38 +08:00
josc146
719090cc8c improve python backend startup speed 2023-07-29 19:18:01 +08:00
josc146
9d89b6f4db fix params 2023-07-28 22:13:19 +08:00
josc146
d0fd480bd6 chore 2023-07-26 22:24:26 +08:00