Hermann Hans Klie
d77a42a776
Merge eb6c5a171e into bd9f2de73a
2025-11-30 14:07:46 +08:00
oobabooga
bd9f2de73a
Merge pull request #7331 from oobabooga/dev
...
Merge dev branch
2025-11-28 23:00:01 -03:00
aidevtime
661e42d2b7
fix(deps): upgrade coqui-tts to >=0.27.0 for transformers 4.55 compatibility ( #7329 )
2025-11-28 22:59:36 -03:00
oobabooga
5327bc9397
Update modules/shared.py
...
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
2025-11-28 22:48:05 -03:00
oobabooga
78b315344a
Update exllamav3
2025-11-28 06:45:05 -08:00
oobabooga
3cad0cd4c1
Update llama.cpp
2025-11-28 03:52:37 -08:00
GodEmperor785
400bb0694b
Add slider for --ubatch-size for llama.cpp loader, change defaults for better MoE performance ( #7316 )
2025-11-21 16:56:02 -03:00
oobabooga
8f0048663d
More modular HTML generator
2025-11-21 07:09:16 -08:00
oobabooga
b0baf7518b
Remove macos x86-64 portable builds (macos-13 runner deprecated by GitHub)
2025-11-19 06:07:15 -08:00
oobabooga
1afe0827ba
Merge pull request #7317 from oobabooga/dev
...
Merge dev branch
2025-11-19 11:04:02 -03:00
oobabooga
0d4eff284c
Add a --cpu-moe model for llama.cpp
2025-11-19 05:23:43 -08:00
oobabooga
d6f39e1fef
Add ROCm portable builds
2025-11-18 16:32:20 -08:00
oobabooga
327a234d23
Add ROCm requirements.txt files
2025-11-18 16:24:56 -08:00
oobabooga
4e4abd0841
Merge remote-tracking branch 'refs/remotes/origin/dev' into dev
2025-11-18 14:07:05 -08:00
oobabooga
c45f35ccc2
Remove the macos 13 wheels (deprecated by GitHub)
2025-11-18 14:06:42 -08:00
oobabooga
d85b95bb15
Update llama.cpp
2025-11-18 14:06:04 -08:00
dependabot[bot]
4a36b7be5b
Bump triton-windows in /requirements/full ( #7311 )
2025-11-18 18:51:26 -03:00
dependabot[bot]
3d7e9856a2
Update peft requirement from ==0.17.* to ==0.18.* in /requirements/full ( #7310 )
2025-11-18 18:51:15 -03:00
oobabooga
a26e28bdea
Update exllamav3 to 0.0.15
2025-11-18 11:24:16 -08:00
oobabooga
6a3bf1de92
Update exllamav3 to 0.0.14
2025-11-09 19:43:53 -08:00
oobabooga
9ad9afad7d
Merge pull request #7296 from oobabooga/dev
...
Merge dev branch
2025-11-06 00:38:25 -03:00
oobabooga
e7534a90d8
Update llama.cpp
2025-11-05 18:46:01 -08:00
oobabooga
6be1bfcc87
Remove the CUDA 11.7 portable builds
2025-11-05 05:45:10 -08:00
oobabooga
92d9cd36a6
Update llama.cpp
2025-11-05 05:43:34 -08:00
oobabooga
67f9288891
Pin huggingface-hub to 0.36.0 (solves #7284 and #7289 )
2025-11-02 14:01:00 -08:00
oobabooga
16f77b74c4
Merge remote-tracking branch 'refs/remotes/origin/dev' into dev
2025-11-01 19:58:53 -07:00
oobabooga
cd645f80f8
Update exllamav3 to 0.0.12
2025-11-01 19:58:18 -07:00
Trenten Miller
6871484398
fix: Rename 'evaluation_strategy' to 'eval_strategy' in training
2025-10-28 16:48:04 -03:00
oobabooga
338ae36f73
Add weights_only=True to torch.load in Training_PRO
2025-10-28 12:43:16 -07:00
dependabot[bot]
c8cd840b24
Bump flash-linear-attention from 0.3.2 to 0.4.0 in /requirements/full ( #7285 )
...
Bumps [flash-linear-attention](https://github.com/fla-org/flash-linear-attention ) from 0.3.2 to 0.4.0.
- [Release notes](https://github.com/fla-org/flash-linear-attention/releases )
- [Commits](https://github.com/fla-org/flash-linear-attention/compare/v0.3.2...v0.4.0 )
---
updated-dependencies:
- dependency-name: flash-linear-attention
dependency-version: 0.4.0
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2025-10-28 10:07:03 -03:00
Hermann Hans Klie
eb6c5a171e
Update loaders.py
2025-10-24 09:06:29 +03:00
Hermann Hans Klie
779795266f
Update models.py
...
the def load_model(model_name, loader=None) we fill in ktransformers .
before the def unload_model(keep_model_name=False) fill def ktransformers_loader
2025-10-24 08:53:23 +03:00
oobabooga
fc67e5e692
Merge pull request #7279 from oobabooga/dev
...
Merge dev branch
2025-10-23 12:50:31 -03:00
oobabooga
f4c9e67155
Update llama.cpp
2025-10-23 08:19:32 -07:00
Immanuel
9a84a828fc
Fixed python requirements for apple devices with macos tahoe ( #7273 )
2025-10-22 14:59:27 -03:00
reksarka
138cc654c4
Make it possible to run a portable Web UI build via a symlink ( #7277 )
2025-10-22 14:55:17 -03:00
Hermann Hans Klie
02c7049227
Merge pull request #1 from hermannklie/ktransformers_in_textgenwebui
...
Add KTransformers loader integration
2025-10-21 20:47:20 +03:00
Hermann Hans Klie
8fdb1b1e5f
Add KTransformers loader integration
...
This PR adds native support for the KTransformers backend as a selectable loader in Text-Generation-WebUI.
It provides a reproducible installation and integration process compatible with the one-click installer (Conda environment).
The integration is not limited to small models — it has meant to be used with Qwen3-Next-80B-A3B-Instruct-FP8 and other larger architectures using FP8 like DeepSeeK FP8 model and FlashAttention-2.
Smaller models (e.g., Qwen3-4B-Instruct) now run efficiently, confirming broad coverage from laptop to workstation setups.
2025-10-21 20:44:25 +03:00
oobabooga
24fd2b4dec
Update exllamav3 to 0.0.11
2025-10-21 07:26:38 -07:00
oobabooga
be81f050a7
Merge remote-tracking branch 'refs/remotes/origin/dev' into dev
2025-10-20 19:43:36 -07:00
oobabooga
9476123ee6
Update llama.cpp
2025-10-20 19:43:26 -07:00
dependabot[bot]
0d85744205
Bump triton-windows in /requirements/full ( #7274 )
2025-10-20 20:36:55 -03:00
oobabooga
771130532c
Merge pull request #7267 from oobabooga/dev
...
Merge dev branch
2025-10-15 17:15:28 -03:00
oobabooga
a156ebbf76
Lint
2025-10-15 13:15:01 -07:00
oobabooga
c871d9cdbd
Revert "Same as 7f06aec3a1 but for exllamav3_hf"
...
This reverts commit deb37b821b
2025-10-15 13:05:41 -07:00
oobabooga
163d863443
Update llama.cpp
2025-10-15 11:23:10 -07:00
oobabooga
c93d567f97
Update exllamav3 to 0.0.10
2025-10-15 06:41:09 -07:00
oobabooga
b5a6904c4a
Make --trust-remote-code immutable from the UI/API
2025-10-14 20:47:01 -07:00
oobabooga
efaf2aef3d
Update exllamav3 to 0.0.9
2025-10-13 15:32:25 -07:00
oobabooga
047855c591
Update llama.cpp
2025-10-13 15:32:03 -07:00
