Dinh Long Nguyen
b0eec07a01
Add contributing section for jan ( #6231 ) ( #6232 )
...
* Add contributing section for jan
* Update CONTRIBUTING.md
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
---------
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
2025-08-20 10:18:35 +07:00
Faisal Amir
a6e5273c7a
Merge pull request #6198 from Louis454545/fix/issue-6025-default-color-selection
...
fix: resolve issue #6025 - default color selection in appearance page
2025-08-20 10:12:49 +07:00
Faisal Amir
5481ee9e35
Merge pull request #6134 from menloresearch/feat/attachment-ui
...
feat: attachment UI
2025-08-20 10:04:32 +07:00
Faisal Amir
87af59b65d
chore: update icon image instead paperclip
2025-08-20 09:46:57 +07:00
Louis
6efdd66bbd
Merge pull request #6236 from menloresearch/feat/add-tool-call-cancellation
2025-08-20 09:04:53 +07:00
Faisal Amir
6203a93325
Merge pull request #6233 from menloresearch/fix/import-model
...
fix: improve ux import model
2025-08-20 09:03:33 +07:00
Faisal Amir
109b6e8c3d
Merge pull request #6234 from menloresearch/fix/download-model
...
fix: downloaded model should from variant level instead of the model level
2025-08-20 09:02:50 +07:00
Louis
91f05b8f32
feat: add tool call cancellation
2025-08-19 23:27:12 +07:00
Faisal Amir
5155f19c9b
chore: update data test id chat input
2025-08-19 22:56:16 +07:00
Faisal Amir
07b1101736
chore: enable attachment icon for remote provider
2025-08-19 22:46:45 +07:00
Faisal Amir
80dc491f9d
chore: conditianal attachment and drag file to chat input
2025-08-19 22:46:45 +07:00
Faisal Amir
e3eb8e909b
chore: attachment icon conditional
2025-08-19 22:46:45 +07:00
Akarshan Biswas
e761c439d7
feat: Pass API key via environment variable instead of command line argument ( #6225 )
...
This change modifies how the API key is passed to the llama-server process. Previously, it was sent as a command line argument (--api-key). This approach has been updated to pass the key via an environment variable (LLAMA_API_KEY).
This improves security by preventing the API key from being visible in the process list (ps aux on Linux, Task Manager on Windows, etc.), where it could potentially be exposed to other users or processes on the same system.
The commit also updates the Rust backend to read the API key from the environment variable instead of parsing it from the command line arguments.
2025-08-19 20:57:06 +05:30
Louis
df41bad465
fix: import
2025-08-19 22:16:24 +07:00
Faisal Amir
b828d3f84f
chore: handle toaster failed import model
2025-08-19 22:07:30 +07:00
Louis
a210e2f13a
fix: timeout for completion request
2025-08-19 22:06:26 +07:00
Faisal Amir
6c612d8eba
chore: seperate function handle import model
2025-08-19 22:01:14 +07:00
Faisal Amir
9e3237da38
fix: downloaded model hub screen should from variant level instead of the model level
2025-08-19 21:55:57 +07:00
Faisal Amir
6ee044d106
fix: validation import model
2025-08-19 21:29:18 +07:00
Faisal Amir
067f8b5447
chore: update attachment icon alignment
2025-08-19 20:00:47 +07:00
Akarshan
cffbe1a77b
Increase timeout for fetch in llamacpp
2025-08-19 20:00:10 +07:00
Faisal Amir
fdc8e07f86
chore: update model setting include offload-mmproj
2025-08-19 20:00:08 +07:00
Faisal Amir
00674ec0d5
chore: update test case model service
2025-08-19 19:51:33 +07:00
Faisal Amir
9f39f0cdb8
fix: paste image chat input
2025-08-19 19:51:32 +07:00
Faisal Amir
050d38c0ab
fix: convert search hgf repo include mmproj models
2025-08-19 19:51:32 +07:00
Akarshan
9afeb5e514
feat: Add offload_mmproj option and validation
...
This commit introduces a new configuration option offload_mmproj to the llamacpp extension.
The offload_mmproj setting allows users to control whether the multimodal projector model is offloaded to the GPU. By default, it's offloaded for better performance. If set to false, the projector model will remain on the CPU, which can be useful in low GPU memory scenarios, though image processing might take longer.
Additionally, this commit adds validate_mmproj_path to ensure the provided --mmproj path is valid and accessible, preventing issues during model loading.
This change also refactors some invoke calls for improved readability.
2025-08-19 19:51:29 +07:00
Faisal Amir
e04bb86171
chore: prevent drag image to replace the window, and enable shortcut copy and paste image
2025-08-19 19:51:02 +07:00
Faisal Amir
f70449fd98
chore: remove pdf attachement
2025-08-19 19:51:02 +07:00
Faisal Amir
ec9925ed5a
chore: update copy validation format type
2025-08-19 19:51:01 +07:00
Faisal Amir
f95e17c047
chore: disable pdf for local model
2025-08-19 19:51:01 +07:00
Faisal Amir
cef3e122ff
chore: send attachment file when send message
2025-08-19 19:51:01 +07:00
Faisal Amir
e9bd0f0bec
chore: update alignment
2025-08-19 19:51:01 +07:00
Faisal Amir
5f1cb67ffc
feat: enable attachment UI
2025-08-19 19:51:01 +07:00
Faisal Amir
926f3ab99a
Merge pull request #6226 from menloresearch/enhancement/offload-model-provider-disabled
...
enhancement: offload model when provider not active
2025-08-19 16:33:49 +07:00
Faisal Amir
f26d78dfe4
Merge pull request #6229 from menloresearch/fix/hide-model-provider-disabled
...
fix: Auto select model only for active provider
2025-08-19 16:33:45 +07:00
Louis
b32c8b372e
Merge pull request #6224 from menloresearch/fix/use-tauri-fetch-for-custom-providers
...
fix: should use custom fetch provided by app
2025-08-19 15:17:27 +07:00
Faisal Amir
cedf42d683
fix: Auto select model only for active provider
2025-08-19 14:47:02 +07:00
Dinh Long Nguyen
9ea9b7d87d
handle abort properly + finally clause to resolve ( #6227 )
2025-08-19 14:45:57 +07:00
Faisal Amir
872284b770
enhancement: offload model when provider not active
2025-08-19 14:18:39 +07:00
Louis
55390de070
Merge pull request #6222 from menloresearch/feat/model-tool-use-detection
...
feat: #5917 - model tool use capability should be auto detected
2025-08-19 13:55:08 +07:00
Louis
65f23fb387
fix: tokenjs should use custom fetch provided by app
2025-08-19 13:46:21 +07:00
Louis
d7cf258a40
fix: tool indicator in hub
2025-08-19 10:40:11 +07:00
Louis
0dc0a9a4a3
fix: tests
2025-08-19 09:57:10 +07:00
Louis
bfe671d7b4
feat: #5917 - model tool use capability should be auto detected
2025-08-19 09:51:36 +07:00
Louis
e43454d518
Merge pull request #6219 from menloresearch/fix/remove-retired-models-from-providers
2025-08-19 09:03:26 +07:00
Louis
3e6757300c
fix: update token.js to remove retired models
2025-08-18 22:33:20 +07:00
Dinh Long Nguyen
2d486d7b3a
feat: add support for reasoning fields (OpenRouter) ( #6206 )
...
* add support for reasoning fields (OpenRouter)
* reformat
* fix linter
* Update web-app/src/utils/reasoning.ts
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
---------
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
2025-08-18 21:59:14 +07:00
Akarshan Biswas
5ad3d282af
fix: re-enable Vulkan backend in integrated GPUs with enough memory ( #6215 )
2025-08-18 17:31:01 +05:30
Faisal Amir
8f2f3ca352
Merge pull request #6214 from menloresearch/enhancement/disabled-form
...
enhancement: mcp disable save button when server name still empty
2025-08-18 14:12:52 +07:00
Faisal Amir
1eb1bb91ce
chore: update placeholder
2025-08-18 12:17:24 +07:00
