mirror/LocalAI
1
0 Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2026-10-28 00:59:28 -04:00
Code Issues Packages Projects Releases Wiki Activity

feat(gallery): add Atomic LFM2.5 24B variants

Browse Source 
Add AtomicChat's LFM2.5-24B-A2B derivative with Q4, Q6, and Q8
GGUF builds for llama-cpp. Pin downloads and document installation.

Assisted-by: Codex:gpt-6
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

(cherry picked from commit f38d175875, mudler/LocalAI#12696)
Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
localai-org-maint-bot authored and Ettore Di Giacinto committed 2026-10-25 09:55:39 +00:00
1 parent d92c5f9646
commit 1a6759a873
2 files changed
+276
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

gallery/index.yaml
+126
View File
@@ -3,6 +3,132 @@
# written by scripts/models/cmd/modelcatalog and stays only so releases before
# the cutover keep working. CI fails when it differs from the generated copy:
# go run ./scripts/models/cmd/modelcatalog --library gallery/models/index.yaml --out gallery/index.yaml
- url: github:mudler/LocalAI/gallery/virtual.yaml@master
name: atomic-lfm2.5-24b-a2b-uncensored
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q4_K_M GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
sha256: 17a5de7f9718d8f6133ef08579d35845dd69bb51f329204ead0fc0ee97a75a41
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
variants:
- model: atomic-lfm2.5-24b-a2b-uncensored-q6
- model: atomic-lfm2.5-24b-a2b-uncensored-q8
- url: github:mudler/LocalAI/gallery/virtual.yaml@master
name: atomic-lfm2.5-24b-a2b-uncensored-q6
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q6_K GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
sha256: 2eea8a0f214304357f45527b2bbc757a810445e8cbb38d7664115ac426aa0711
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
- url: github:mudler/LocalAI/gallery/virtual.yaml@master
name: atomic-lfm2.5-24b-a2b-uncensored-q8
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q8_0 GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
sha256: 852b43bc93dfc776f702c9ef0661c1e8cf92eb2d6914def912c51a2a9d45a4d1
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
- url: github:mudler/LocalAI/gallery/virtual.yaml@master
name: atomic-minicpm-v-4.7-35b-a3b-uncensored-q4
description: |
gallery/models/index.yaml
+150
View File
@@ -4,6 +4,156 @@
# changes and never edit it by hand. CI fails when it is stale. A build from
# master reads it as the default model source; a tagged build reads the gallery
# until the release cutover.
- kind: model
id: atomic-lfm2.5-24b-a2b-uncensored
name: atomic-lfm2.5-24b-a2b-uncensored
version: 0.0.0-gallery.83179d51480f
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q4_K_M GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
channel: gallery
policy:
requireSignature: false
schemaVersion: 1
model:
url: github:mudler/LocalAI/gallery/virtual.yaml@master
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
sha256: 17a5de7f9718d8f6133ef08579d35845dd69bb51f329204ead0fc0ee97a75a41
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q4_K_M.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
variants:
- model: atomic-lfm2.5-24b-a2b-uncensored-q6
- model: atomic-lfm2.5-24b-a2b-uncensored-q8
- kind: model
id: atomic-lfm2.5-24b-a2b-uncensored-q6
name: atomic-lfm2.5-24b-a2b-uncensored-q6
version: 0.0.0-gallery.810d2644dbd4
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q6_K GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
channel: gallery
policy:
requireSignature: false
schemaVersion: 1
model:
url: github:mudler/LocalAI/gallery/virtual.yaml@master
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
sha256: 2eea8a0f214304357f45527b2bbc757a810445e8cbb38d7664115ac426aa0711
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q6_K.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
- kind: model
id: atomic-lfm2.5-24b-a2b-uncensored-q8
name: atomic-lfm2.5-24b-a2b-uncensored-q8
version: 0.0.0-gallery.8ea7aed2c62f
description: |
AtomicChat's LFM2.5-24B-A2B derivative reduces refusal behavior in the
original weights. This text-only mixture-of-experts model has 24B total
parameters and supports multilingual reasoning and tool use.
This entry uses the AD-Q8_0 GGUF build with the edits baked in.
license: other
urls:
- https://huggingface.co/LiquidAI/LFM2.5-24B-A2B
- https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF
tags:
- llm
- gguf
- cpu
- gpu
- reasoning
- uncensored
channel: gallery
policy:
requireSignature: false
schemaVersion: 1
model:
url: github:mudler/LocalAI/gallery/virtual.yaml@master
files:
- filename: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
sha256: 852b43bc93dfc776f702c9ef0661c1e8cf92eb2d6914def912c51a2a9d45a4d1
uri: https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF/resolve/1acb8f6eb027d85b7771dcc9eb68817aa480e9c8/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
overrides:
backend: llama-cpp
context_size: 32768
function:
automatic_tool_parsing_fallback: true
grammar:
disable: true
known_usecases:
- chat
- completion
options:
- use_jinja:true
parameters:
model: llama-cpp/models/atomic-lfm2.5-24b-a2b-uncensored/LFM2.5-24B-A2B-Uncensored-AD-Q8_0.gguf
repeat_penalty: 1.05
temperature: 0.1
top_k: 50
template:
use_tokenizer_template: true
- kind: model
id: atomic-minicpm-v-4.7-35b-a3b-uncensored-q4
name: atomic-minicpm-v-4.7-35b-a3b-uncensored-q4