tabbyAPI

theroyallab/tabbyAPI

Fork 0

mirror of https://github.com/theroyallab/tabbyAPI.git synced 2026-07-13 02:27:11 +00:00

Commit Graph

Select branches

Hide Pull Requests

breaking

main

main_tools

tool-calls

#100

#101

#102

#103

#104

#105

#108

#109

#11

#111

#112

#113

#114

#12

#121

#122

#125

#126

#128

#129

#13

#130

#132

#133

#134

#136

#138

#139

#141

#141

#146

#147

#150

#154

#155

#158

#159

#161

#163

#171

#173

#174

#175

#178

#18

#180

#181

#183

#184

#185

#189

#19

#190

#190

#191

#192

#193

#194

#195

#197

#199

#20

#201

#204

#206

#208

#213

#213

#214

#214

#22

#220

#222

#222

#224

#225

#228

#23

#230

#232

#233

#233

#24

#243

#244

#246

#249

#252

#253

#254

#26

#264

#265

#266

#27

#271

#28

#281

#289

#29

#292

#293

#295

#30

#302

#303

#305

#307

#308

#31

#312

#313

#313

#320

#322

#326

#327

#328

#329

#33

#330

#331

#336

#339

#34

#340

#341

#342

#346

#349

#35

#351

#358

#36

#362

#364

#365

#37

#370

#370

#371

#372

#373

#374

#375

#376

#377

#378

#378

#380

#382

#39

#393

#394

#394

#395

#395

#396

#396

#397

#4

#40

#402

#403

#403

#407

#411

#411

#413

#413

#415

#415

#418

#420

#421

#422

#423

#424

#424

#425

#427

#429

#43

#430

#431

#431

#44

#45

#46

#47

#50

#54

#56

#58

#59

#61

#63

#66

#71

#72

#75

#76

#78

#80

#82

#85

#88

#9

#93

#94

#99

#99

565b0300d6 Dependencies: Update Exllamav2 kingbri 2024-08-22 14:07:28 -04:00
078fbf1080 Model: Add quantized cache support for tensor parallel kingbri 2024-08-21 18:01:42 -04:00
871c89063d Model: Add Tensor Parallel support kingbri 2024-08-16 16:35:19 -04:00
5002617eac Model: Split cache creation into a common function kingbri 2024-08-16 15:17:03 -04:00
ecaddec48a Docker-compose: Add models to bind mounts kingbri 2024-08-19 22:07:53 -04:00
dae394050e Improve docker deployment configuration (#163) Amgad Hasan 2024-08-18 22:19:18 +03:00
a51acb9db4 Templates: Switch to async jinja engine kingbri 2024-08-17 12:01:59 -04:00
b4752c1e62 Templates: Revert to load metadata on runtime kingbri 2024-08-17 11:30:50 -04:00
617ac12150 Update README kingbri 2024-08-17 00:35:42 -04:00
70b9fc95de [WIP] OpenAI Tools Support/Function calling (#154) Ben Gitter 2024-08-17 00:16:25 -04:00
c75e911f07 Merge branch 'main' into main Bartowski 2024-08-14 16:16:15 -04:00
9cc0e70098 Actions: Build kobold docs subpage kingbri 2024-08-08 16:40:50 -04:00
685e3836e9 Args: Add api-servers to parser kingbri 2024-08-08 16:32:29 -04:00
63650d2c3c Model: Disable banned strings if grammar is used kingbri 2024-08-05 11:08:58 -04:00
34281c2e14 Start: Add --force-reinstall argument kingbri 2024-08-04 11:14:38 -04:00
ab6c3a53b9 Start: Remove eager upgrade strategy kingbri 2024-08-04 10:50:14 -04:00
8ff2586d45 Start: Fix pip update, method calls, and logging kingbri 2024-08-04 10:13:19 -04:00
6a0cfd731b Main: Only import psutil when the experimental function is run kingbri 2024-08-03 22:00:15 -04:00
b6d2676f1c Start: Give the user a hint when a module can't be imported kingbri 2024-08-03 21:56:53 -04:00
1aa934664c Issues: Update issue templates kingbri 2024-08-03 21:44:45 -04:00
87b6a31fad Update .gitignore kingbri 2024-08-03 20:59:28 -04:00
4868fc6b10 Update README kingbri 2024-08-03 20:56:40 -04:00
5fb9cdc2b1 Dependencies: Add Python 3.12 specific dependencies kingbri 2024-08-03 17:43:14 -04:00
2a33ebbf29 Model: Bypass lock checks when shutting down kingbri 2024-08-03 16:05:34 -04:00
65c16f2a7c Merge pull request #161 from theroyallab/new-start-scripts Brian Dashore 2024-08-03 15:21:02 -04:00
8703b23f89 Start: Make linux scripts executable kingbri 2024-08-03 15:19:31 -04:00
b795bfc7b2 Start: Split some prints up kingbri 2024-08-03 15:14:40 -04:00
65e758e134 Tree: Format kingbri 2024-08-03 15:07:01 -04:00
7ce46cc2da Start: Rewrite start scripts kingbri 2024-08-03 13:03:24 -04:00
e66d213aef Revert "Dependencies: Use hosted pip index instead of Github" kingbri 2024-08-03 11:35:26 -04:00
7bf2b07d4c Signals: Exit on async cleanup kingbri 2024-08-02 15:10:27 -04:00
b124797949 Dependencies: Re-add sentence-transformers kingbri 2024-08-02 14:35:58 -04:00
56619810bf Dependencies: Switch sentence-transformers to infinity-emb kingbri 2024-08-02 13:34:47 -04:00
3e42211c3e Config: Embeddings: Make embeddings_device a default when API loading kingbri 2024-08-01 13:59:49 -04:00
54aeebaec1 API: Fix return of current embeddings model kingbri 2024-08-01 13:43:31 -04:00
0bcb4e4a7d Model: Attach request ID to logs kingbri 2024-08-01 00:25:54 -04:00
9390d362dd Model: Log generation params and metrics after the prompt/response kingbri 2024-08-01 00:19:21 -04:00
1bf062559d Merge pull request #158 from AlpinDale/embeddings Brian Dashore 2024-07-31 20:33:12 -04:00
f111052e39 Dependencies: Use hosted pip index instead of Github kingbri 2024-07-30 20:46:37 -04:00
46304ce875 Model: Properly pass in max_batch_size from config kingbri 2024-07-30 18:42:25 -04:00
dc3dcc9c0d Embeddings: Update config, args, and parameter names kingbri 2024-07-30 15:32:26 -04:00
bfa011e0ce Embeddings: Add model management kingbri 2024-07-30 15:19:27 -04:00
f13d0fb8b3 Embeddings: Add model load checks kingbri 2024-07-30 11:17:36 -04:00
01c7702859 Signal: Fix async signal handling kingbri 2024-07-30 11:11:05 -04:00
fbf1455db1 Embeddings: Migrate and organize Infinity kingbri 2024-07-30 11:00:23 -04:00
ac1afcc588 Embeddings: Use response classes instead of dicts kingbri 2024-07-29 14:15:40 -04:00
3f21d9ef96 Embeddings: Switch to Infinity kingbri 2024-07-29 13:42:03 -04:00
c9a5d2c363 OAI: Refactor embeddings kingbri 2024-07-28 14:10:51 -04:00
d85414738d Dependencies: Update Flash Attention 2 kingbri 2024-07-28 13:50:15 -04:00
c79e0832d5 Revert "Dependencies: Update pytorch and flash_attention" kingbri 2024-07-28 13:48:49 -04:00
7b8b3fe23d Kobold: Fix max length type kingbri 2024-07-26 23:00:26 -04:00
e3226ed930 Kobold: Add untracked file kingbri 2024-07-26 22:57:55 -04:00
3038f668e8 Kobold: Add extra routes for horde compatability kingbri 2024-07-26 22:50:01 -04:00
2773517a16 API: Add setup function to routers kingbri 2024-07-26 22:24:33 -04:00
6365427d38 Merge pull request #155 from Vhallo/main Brian Dashore 2024-07-26 21:35:50 -04:00
884b6f5ecd API: Add log options for initialization kingbri 2024-07-26 21:32:05 -04:00
e8fc13a1f6 Tree: Format kingbri 2024-07-26 18:33:04 -04:00
ea80b62e30 Sampling: Reorder aliased params and add kobold aliases kingbri 2024-07-26 18:32:33 -04:00
7522b1447b Model: Add support for HuggingFace config and bad_words_ids kingbri 2024-07-26 18:23:22 -04:00
545e26608f Kobold: Move params to aliases kingbri 2024-07-26 16:45:29 -04:00
b7cb6f0b91 API: Add KoboldAI server kingbri 2024-07-26 16:37:30 -04:00
4e808cbed7 Auth: Fix disable auth when checking for key permissions kingbri 2024-07-26 15:04:29 -04:00
f47d96790c Dependencies: Update pytorch and flash_attention kingbri 2024-07-25 23:39:52 -04:00
5adfab1cbd ruff: formatting AlpinDale 2024-07-26 02:53:14 +00:00
765d3593b3 remove submodule AlpinDale 2024-07-26 02:52:18 +00:00
f20cd330ef feat: add embeddings support via sentence-transformers AlpinDale 2024-07-26 02:45:07 +00:00
a1c3f6cc1c Dependencies: Update ExllamaV2 kingbri 2024-07-24 22:00:43 -04:00
27f9559d83 Dependencies: Switch to fastapi-slim kingbri 2024-07-24 21:50:16 -04:00
42bc4adcfb Config: Add option to set priority to realtime kingbri 2024-07-24 21:47:47 -04:00
5c082b7e8c Async: Add option to use Uvloop/Winloop kingbri 2024-07-24 18:56:28 -04:00
71de3060bb Downloader: Make timeout configurable kingbri 2024-07-23 21:42:38 -04:00
8c02fe9771 Downloader: Disable timeout kingbri 2024-07-23 21:37:53 -04:00
b2064bbfb4 Typo fix in completion.py Vhallo 2024-07-23 23:49:43 +02:00
88e4b108b4 Typo fix in chat_completion.py Vhallo 2024-07-23 23:48:50 +02:00
3e8ffebdd3 Tree: Format kingbri 2024-07-23 14:32:50 -04:00
300f034233 API: Add config option to select servers kingbri 2024-07-23 14:26:15 -04:00
9ad69e8ab6 API: Migrate universal routes to core kingbri 2024-07-23 14:08:48 -04:00
64c2cc85c9 OAI: Migrate model depends into proper file kingbri 2024-07-23 12:40:32 -04:00
d1706fb067 OAI: Remove double logging if request is cancelled kingbri 2024-07-22 21:48:59 -04:00
14dfaf600a Args: Add request logging kingbri 2024-07-22 21:41:42 -04:00
3826815edb API: Add request logging kingbri 2024-07-22 21:33:10 -04:00
522999ebb4 Config: Change from gen_logging to logging kingbri 2024-07-22 21:15:16 -04:00
191600a150 Revert "Model: Skip empty token chunks" kingbri 2024-07-22 18:34:00 -04:00
15f891b277 Args: Update to latest config.yml kingbri 2024-07-22 16:25:26 -04:00
ad4d17bca2 Tree: Format kingbri 2024-07-22 12:24:34 -04:00
21516bd7b5 Model: Skip empty token chunks kingbri 2024-07-22 12:23:49 -04:00
0eedc8ca14 API: Switch from request ID middleware to depends kingbri 2024-07-22 12:19:46 -04:00
cae94b920c API: Add ability to use request IDs kingbri 2024-07-21 21:01:05 -04:00
38185a1ff4 Auth: Fix key check coalesce kingbri 2024-07-19 10:08:57 -04:00
e20a2d504b API: Fix pydantic validation errors on disconnect poll returns kingbri 2024-07-15 14:39:55 -04:00
933404c185 Model: Warn user if terminating jobs kingbri 2024-07-15 11:33:44 -04:00
9dae461142 Model: Attempt to recreate generator on a fatal error kingbri 2024-07-15 01:09:49 -04:00
6019c93637 Networking: Gate sending tracebacks over the API kingbri 2024-07-13 17:59:58 -04:00
ddad422d8b Merge pull request #150 from AmgadHasan/fixDockerWorkflow Brian Dashore 2024-07-12 14:38:37 -04:00
ae4ba76fab Merge pull request #147 from ai-and-i/stream_options Brian Dashore 2024-07-12 14:38:20 -04:00
c1b61441f4 OAI: Fix usage chunk return kingbri 2024-07-12 14:35:48 -04:00
5917515696 Dependencies: Update flash-attention kingbri 2024-07-12 10:09:49 -04:00
2e5cf0ea3f Fix docker compose volume mount Amgad Hasan 2024-07-12 13:23:58 +00:00
b149d3398d OAI: support stream_options argument Volodymyr Kuznetsov 2024-07-08 13:42:54 -07:00
073e9fa6f0 Dependencies: Bump ExllamaV2 kingbri 2024-07-11 14:11:37 -04:00

... 5 6 7 8 9 ...