9380da130e
fix: fix tests
2023-11-10 14:57:15 -08:00
4068d6e81d
refactor: extract BoxCodeSearch as interface to CodeSearch ( #756 )
2023-11-10 22:55:51 +00:00
73a76a3d8e
feat(scheduler): add a tqdm bar for scheduler job to better present the remaining time. ( #754 )
...
* feat(scheduler): add a tqdm bar for scheduler job to better present the
remaining time.
* update
* add changelog
2023-11-10 19:52:07 +00:00
3600ef77fc
fix: clippy warnings in CodeSearchSchema
2023-11-10 11:48:52 -08:00
27ed9c2cc4
fix: tantivy requires more memory for indexing in new version. ( #753 )
2023-11-10 11:41:50 -08:00
b510f61aca
refactor: extract tabby_common::api::code / tabby_common::index::CodeSearchSchema ( #743 )
...
* refactor: extract tabby_common::api::code
mark CodeSearch being Send + Sync
* extract CodeSearchSchema
2023-11-10 10:11:13 -08:00
138b7459c5
feat: add LLAMA_CPP_N_THREADS env ( #742 )
...
* feat: add LLAMA_CPP_N_THREADS and LLAMA_CPP_N_THREADS_BATCH envs
* apply format
* improve: use LLAMA_CPP_N_THREADS for both n_threads and n_threads_batch
* Update crates/llama-cpp-bindings/src/engine.cc
---------
Co-authored-by: Meng Zhang <meng@tabbyml.com>
2023-11-09 19:54:23 +00:00
8c669dee8e
fix: llama.cpp queuing logic ( #741 )
2023-11-09 08:29:54 +00:00
03ff80efdb
feat: update tabby-ui
2023-11-08 16:07:39 -08:00
cde3602877
feat: sync llama.cpp to latest
2023-11-08 16:06:09 -08:00
b51520062a
refactor: extract ChatState -> ChatService ( #730 )
2023-11-08 22:12:29 +00:00
72d1d9f0bb
refactor: extract IndexServer into CodeSearchService ( #728 )
...
* refactor: extract IndexServer into CodeSearchService
* refactor: make CodeSearchService interface to be async
2023-11-08 21:42:03 +00:00
8ab35b2639
feat: add --parallelism to control throughput and vram usage ( #727 )
...
* feat: add --parallelism to control throughput and vram usage
* update default
* Revert "update default"
This reverts commit 349792c0d48d913dcd8be4ce1c9d7ce887918f29.
* cargo fmt
2023-11-08 18:31:22 +00:00
3fb8445747
feat: supports java ( #715 )
...
* feat: add java language configuration
* feat: add java repository context support
* Update programming-languages.md
* added rev to tree-sitter-java
Signed-off-by: Lei Wen <wenlei03@qiyi.com>
Co-authored-by: Lei Wen <wenlei03@qiyi.com>
2023-11-07 23:59:00 -08:00
1ad0d39903
fix: deadlock between background job and requests ( #720 )
...
* fix: deadlock between background job and requests
* refactor: extract LlamaService
2023-11-07 13:11:28 -08:00
3c3b14c9f5
fix: cuda serialization
2023-11-07 00:55:38 -08:00
ca52ac4b01
fix: support cpu only run in llama.cpp cuda build
2023-11-06 22:59:24 -08:00
eb7ae96157
fix: llama.cpp requires kv cache to be N_CTX * parallelism ( #714 )
2023-11-07 06:16:36 +00:00
9344c32b31
fix: when there's an error happens in background inference loop, it should exit the process ( #713 )
2023-11-06 20:41:49 +00:00
00e0c4fddc
chore: add machete check to ensure no unused dependencies ( #701 )
...
* refactor: remove useless dependencies
* add machete
2023-11-05 02:48:05 +00:00
33ef27ba30
feat: support downloading resume ( #700 )
2023-11-05 02:38:06 +00:00
64e0abb8cc
fix(llama.cpp): wrongly index for n_seq in warmup
2023-11-04 17:53:22 -07:00
c7c67c2f90
fix: llama.cpp warmp logic
2023-11-04 14:28:04 -07:00
fc9c9f644b
Release 0.6.0-dev
...
http-api-bindings@0.6.0-dev
llama-cpp-bindings@0.6.0-dev
tabby@0.6.0-dev
tabby-common@0.6.0-dev
tabby-download@0.6.0-dev
tabby-inference@0.6.0-dev
tabby-scheduler@0.6.0-dev
Generated by cargo-workspaces
2023-11-03 18:04:12 -07:00
ec8d88de0d
chore: release 0.5.0 ( #697 )
...
* Release 0.5.0-rc.0
http-api-bindings@0.5.0-rc.0
llama-cpp-bindings@0.5.0-rc.0
tabby@0.5.0-rc.0
tabby-common@0.5.0-rc.0
tabby-download@0.5.0-rc.0
tabby-inference@0.5.0-rc.0
tabby-scheduler@0.5.0-rc.0
Generated by cargo-workspaces
* fix: docker branch tag should only generate when not empty
* Release 0.5.0-rc.1
http-api-bindings@0.5.0-rc.1
llama-cpp-bindings@0.5.0-rc.1
tabby@0.5.0-rc.1
tabby-common@0.5.0-rc.1
tabby-download@0.5.0-rc.1
tabby-inference@0.5.0-rc.1
tabby-scheduler@0.5.0-rc.1
Generated by cargo-workspaces
* fix: handlebar syntax in meta action
* Release 0.5.0-rc.2
http-api-bindings@0.5.0-rc.2
llama-cpp-bindings@0.5.0-rc.2
tabby@0.5.0-rc.2
tabby-common@0.5.0-rc.2
tabby-download@0.5.0-rc.2
tabby-inference@0.5.0-rc.2
tabby-scheduler@0.5.0-rc.2
Generated by cargo-workspaces
* fix: handlebar syntax in meta action
* Release 0.5.0-rc.3
http-api-bindings@0.5.0-rc.3
llama-cpp-bindings@0.5.0-rc.3
tabby@0.5.0-rc.3
tabby-common@0.5.0-rc.3
tabby-download@0.5.0-rc.3
tabby-inference@0.5.0-rc.3
tabby-scheduler@0.5.0-rc.3
Generated by cargo-workspaces
* docs: update change log and docs
* fix: collect_snippet should handle NotReady error
* Release 0.5.0-rc.4
http-api-bindings@0.5.0-rc.4
llama-cpp-bindings@0.5.0-rc.4
tabby@0.5.0-rc.4
tabby-common@0.5.0-rc.4
tabby-download@0.5.0-rc.4
tabby-inference@0.5.0-rc.4
tabby-scheduler@0.5.0-rc.4
Generated by cargo-workspaces
* Release 0.5.0
http-api-bindings@0.5.0
llama-cpp-bindings@0.5.0
tabby@0.5.0
tabby-common@0.5.0
tabby-download@0.5.0
tabby-inference@0.5.0
tabby-scheduler@0.5.0
Generated by cargo-workspaces
2023-11-03 18:02:03 -07:00
e4efcc4091
fix: avoid special keywords (e.g AND) failed the query parsing ( #695 )
2023-11-03 01:13:28 +00:00
2adcc0726c
feat: support prefix query on name field ( #694 )
...
* feat: support prefix phase query on name field
* update changelog
2023-11-03 01:04:33 +00:00
acb3a33d78
fix: handle non utf-8 / utf-16 error
2023-11-02 16:29:30 -07:00
eb34850a5e
fix: output err if step failed
2023-11-02 16:15:11 -07:00
4c7eae584e
feat: add model warmup logic ( #693 )
2023-11-02 23:07:32 +00:00
0e4a2d2a12
feat: simplify download management, model file should be able to indi… ( #690 )
...
* feat: simplify download management, model file should be able to individually introduced
* fix typo
* update local model support
* update spec back
* update spec
* update
* update
2023-11-02 16:01:04 -07:00
36ffeb63f1
refactor: remove useless rust-cxx-cmake-bridge
2023-10-31 17:58:21 -07:00
296342efd8
refactor: use llama.cpp tokenizer ( #683 )
...
* refactor: switch to llama.cpp tokenizer to simplify implementation
* refactor: remove tokenizer dependency from tabby
* refactor: renaming decoding to stop condition
* refactor: remove tokenizer dependency
* refactor: remove submodule
* chore: update formatting
* move tokenization to c++
2023-10-31 22:16:09 +00:00
f15926f233
chore(ui): update crates/tabby/ui
2023-10-31 08:51:22 -07:00
3f1f8bfd30
fix: add /swagger to tabby ui handler
2023-10-30 17:36:22 -07:00
73758e207d
feat: improve dashboard UI ( #677 )
2023-10-30 21:47:38 +00:00
b4772fbcd0
feat(ui): add dashboard ( #674 )
...
* feat(ui): add dashboard
* handle path
2023-10-30 07:29:50 +00:00
89a63dbf33
fix: when send failed, treat the request as stopped ( #673 )
2023-10-30 06:27:09 +00:00
de827b1e74
Revert "feat: make --model optional ( #668 )" ( #672 )
...
This reverts commit c55e4481ba
2023-10-29 21:44:11 -07:00
b47bdd5d77
fix: align with fastchat mainstream ( #670 )
...
Fastchat mainstream change its return format, and text now is only
string in choices structure.
So make this change, to work with mainstream fastchat.
Signed-off-by: Lei Wen <wenlei03@qiyi.com>
Co-authored-by: Lei Wen <wenlei03@qiyi.com>
2023-10-29 19:31:46 -07:00
88d2617a34
fix: move events writer to individual thread ( #669 )
2023-10-30 01:31:41 +00:00
c55e4481ba
feat: make --model optional ( #668 )
2023-10-30 00:04:42 +00:00
7330d75de6
chore: clear cache when there's no active requests
2023-10-29 16:30:30 -07:00
2ee5dbfd4f
chore: move tabby-ui under ee license. ( #667 )
...
* chore: introduce tabby-ui EE license.
* update
2023-10-29 15:56:57 -07:00
8c680a73fb
feat(ui): add /api page ( #665 )
...
* refactor(tabby-ui): extract tabby-fetcher
* feat(tabby-ui): add /api page
* feat(tabby-ui): add chat model badge
* fix: add components.json for shadcn
* chore: release tabby-ui
2023-10-29 14:55:50 -07:00
7bd99d14c0
feat: support continuous batching in llama.cpp backend ( #659 )
...
* refactor: switch back to llama batch interface
* feat: support cont batching
2023-10-28 23:37:05 -07:00
14d03b6826
fix(ui): handle invalid semver error
2023-10-28 23:30:35 -07:00
8dc5526091
Revert "feat: supports PHP ( #634 )"
...
This reverts commit 688e7d75b5
2023-10-28 23:02:10 -07:00
43cc5f38cc
feat: do not download ctranslate2 files in downloader
2023-10-28 02:27:09 -07:00
444222683a
fix(llama.cpp): bump upstream fix for starcoder model on cuda
2023-10-28 02:03:34 -07:00
Meng Zhang