Compare commits

...

6 Commits

Author SHA1 Message Date
Ettore Di Giacinto
c27da0a0f6 fix(diffusers): fix float detection (#6313)
There was apparently an oversight, this fixes the float/int detection

Fixes: https://github.com/mudler/LocalAI/issues/6312

Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
2025-09-19 19:09:04 +02:00
Ettore Di Giacinto
ac043ed9ba chore(model gallery): add aquif-3.5-a4b-think (#6311)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2025-09-19 11:16:50 +02:00
Ettore Di Giacinto
2e0d66a1c8 chore(model gallery): add impish_qwen_14b-1m (#6310)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2025-09-19 10:57:33 +02:00
Ettore Di Giacinto
41a0f361eb chore(model gallery): add mistralai_magistral-small-2509 (#6309)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2025-09-19 10:48:13 +02:00
LocalAI [bot]
d3c5c02837 docs: ⬆️ update docs version mudler/LocalAI (#6307)
⬆️ Update docs version mudler/LocalAI

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: mudler <2420543+mudler@users.noreply.github.com>
2025-09-18 23:48:02 +02:00
LocalAI [bot]
ae3d8fb0c4 chore: ⬆️ Update ggml-org/llama.cpp to 3edd87cd055a45d885fa914d879d36d33ecfc3e1 (#6308)
⬆️ Update ggml-org/llama.cpp

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: mudler <2420543+mudler@users.noreply.github.com>
2025-09-18 21:09:14 +00:00
4 changed files with 57 additions and 3 deletions

View File

@@ -1,5 +1,5 @@
LLAMA_VERSION?=0320ac5264279d74f8ee91bafa6c90e9ab9bbb91
LLAMA_VERSION?=3edd87cd055a45d885fa914d879d36d33ecfc3e1
LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
CMAKE_ARGS?=

View File

@@ -177,7 +177,7 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
key, value = opt.split(":")
# if value is a number, convert it to the appropriate type
if is_float(value):
if value.is_integer():
if float(value).is_integer():
value = int(value)
else:
value = float(value)

View File

@@ -1,3 +1,3 @@
{
"version": "v3.5.1"
"version": "v3.5.2"
}

View File

@@ -2638,6 +2638,39 @@
- filename: Alibaba-NLP_Tongyi-DeepResearch-30B-A3B-Q4_K_M.gguf
sha256: 1afefb3b369ea2de191f24fe8ea22cbbb7b412357902f27bd81d693dde35c2d9
uri: huggingface://bartowski/Alibaba-NLP_Tongyi-DeepResearch-30B-A3B-GGUF/Alibaba-NLP_Tongyi-DeepResearch-30B-A3B-Q4_K_M.gguf
- !!merge <<: *qwen3
name: "impish_qwen_14b-1m"
icon: https://huggingface.co/SicariusSicariiStuff/Impish_QWEN_14B-1M/resolve/main/Images/Impish_Qwen_14B.png
urls:
- https://huggingface.co/SicariusSicariiStuff/Impish_QWEN_14B-1M
- https://huggingface.co/mradermacher/Impish_QWEN_14B-1M-GGUF
description: |
Supreme context One million tokens to play with.
Strong Roleplay internet RP format lovers will appriciate it, medium size paragraphs.
Qwen smarts built-in, but naughty and playful Maybe it's even too naughty.
VERY compliant with low censorship.
VERY high IFeval for a 14B RP model: 78.68.
overrides:
parameters:
model: Impish_QWEN_14B-1M.Q4_K_M.gguf
files:
- filename: Impish_QWEN_14B-1M.Q4_K_M.gguf
sha256: d326f2b8f05814ea3943c82498f0cd3cde64859cf03f532855c87fb94b0da79e
uri: huggingface://mradermacher/Impish_QWEN_14B-1M-GGUF/Impish_QWEN_14B-1M.Q4_K_M.gguf
- !!merge <<: *qwen3
name: "aquif-3.5-a4b-think"
urls:
- https://huggingface.co/aquif-ai/aquif-3.5-A4B-Think
- https://huggingface.co/QuantFactory/aquif-3.5-A4B-Think-GGUF
description: |
The aquif-3.5 series is the successor to aquif-3, featuring a simplified naming scheme, expanded Mixture of Experts (MoE) options, and across-the-board performance improvements. This release streamlines model selection while delivering enhanced capabilities across reasoning, multilingual support, and general intelligence tasks.
overrides:
parameters:
model: aquif-3.5-A4B-Think.Q4_K_M.gguf
files:
- filename: aquif-3.5-A4B-Think.Q4_K_M.gguf
sha256: 1650b72ae1acf12b45a702f2ff5f47205552e494f0d910e81cbe40dfba55a6b9
uri: huggingface://QuantFactory/aquif-3.5-A4B-Think-GGUF/aquif-3.5-A4B-Think.Q4_K_M.gguf
- &gemma3
url: "github:mudler/LocalAI/gallery/gemma.yaml@master"
name: "gemma-3-27b-it"
@@ -15175,6 +15208,27 @@
- filename: Impish_Longtail_12B-Q4_K_M.gguf
sha256: 2cf0cacb65d71cfc5b4255f3273ad245bbcb11956a0f9e3aaa0e739df57c90df
uri: huggingface://SicariusSicariiStuff/Impish_Longtail_12B_GGUF/Impish_Longtail_12B-Q4_K_M.gguf
- !!merge <<: *mistral03
name: "mistralai_magistral-small-2509"
urls:
- https://huggingface.co/mistralai/Magistral-Small-2509
- https://huggingface.co/bartowski/mistralai_Magistral-Small-2509-GGUF
description: |
Magistral Small 1.2
Building upon Mistral Small 3.2 (2506), with added reasoning capabilities, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters.
Magistral Small can be deployed locally, fitting within a single RTX 4090 or a 32GB RAM MacBook once quantized.
Learn more about Magistral in our blog post.
The model was presented in the paper Magistral.
overrides:
parameters:
model: mistralai_Magistral-Small-2509-Q4_K_M.gguf
files:
- filename: mistralai_Magistral-Small-2509-Q4_K_M.gguf
sha256: 1d638bc931de30d29fc73ad439206ff185f76666a096e7ad723866a20f78728d
uri: huggingface://bartowski/mistralai_Magistral-Small-2509-GGUF/mistralai_Magistral-Small-2509-Q4_K_M.gguf
- &mudler
url: "github:mudler/LocalAI/gallery/mudler.yaml@master" ### START mudler's LocalAI specific-models
name: "LocalAI-llama3-8b-function-call-v0.2"