mirror of
https://github.com/mudler/LocalAI.git
synced 2026-09-12 22:33:54 -04:00
feat(gallery): add TwIL-LM3 variants (#11478)
Add the publisher-recommended Q4_K_M build and a near-lossless Q8_0 variant for the 3B formal-logic reasoning model. Assisted-by: Codex:gpt-5 Co-authored-by: localai-org-maint-bot <306269227+localai-org-maint-bot@users.noreply.github.com>
This commit is contained in:
1 parent
d50a1b9c98
commit
d54067d4c6
1 file changed
+61
@@ -1,4 +1,65 @@
|
||||
---
|
||||
- &twil-lm3
|
||||
name: "twil-lm3-q4"
|
||||
variants:
|
||||
- model: twil-lm3-q8
|
||||
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
|
||||
urls:
|
||||
- https://huggingface.co/HuggingFaceTB/SmolLM3-3B
|
||||
- https://huggingface.co/webAI-Official/TwIL-LM3
|
||||
description: |
|
||||
TwIL-LM3 is a 3B SmolLM3-based reasoning model specialized for formal
|
||||
logic, entailment, semantic parsing, and Lean formalization. This default
|
||||
entry uses the publisher's recommended Q4_K_M GGUF and supports a 65K-token
|
||||
context window. A higher-quality Q8_0 build is available as a variant.
|
||||
license: "webai-non-commercial-license-ver.-1.0"
|
||||
tags:
|
||||
- llm
|
||||
- gguf
|
||||
- cpu
|
||||
- gpu
|
||||
- reasoning
|
||||
- formal-logic
|
||||
- coding
|
||||
last_checked: "2026-08-12"
|
||||
overrides:
|
||||
backend: llama-cpp
|
||||
context_size: 65536
|
||||
known_usecases:
|
||||
- chat
|
||||
options:
|
||||
- use_jinja:true
|
||||
parameters:
|
||||
model: llama-cpp/models/twil-lm3/TwIL-LM3-Q4_K_M.gguf
|
||||
temperature: 0
|
||||
template:
|
||||
use_tokenizer_template: true
|
||||
files:
|
||||
- filename: llama-cpp/models/twil-lm3/TwIL-LM3-Q4_K_M.gguf
|
||||
uri: huggingface://webAI-Official/TwIL-LM3/TwIL-LM3-Q4_K_M.gguf
|
||||
sha256: e4eb515cceeae304ca4d5bc77447e7635143888afe2a6b37aac45c942dc36f54
|
||||
- !!merge <<: *twil-lm3
|
||||
name: "twil-lm3-q8"
|
||||
variants: []
|
||||
description: |
|
||||
TwIL-LM3 in the publisher's near-lossless Q8_0 GGUF format for
|
||||
quality-sensitive use on hosts with enough memory.
|
||||
overrides:
|
||||
backend: llama-cpp
|
||||
context_size: 65536
|
||||
known_usecases:
|
||||
- chat
|
||||
options:
|
||||
- use_jinja:true
|
||||
parameters:
|
||||
model: llama-cpp/models/twil-lm3/TwIL-LM3-Q8_0.gguf
|
||||
temperature: 0
|
||||
template:
|
||||
use_tokenizer_template: true
|
||||
files:
|
||||
- filename: llama-cpp/models/twil-lm3/TwIL-LM3-Q8_0.gguf
|
||||
uri: huggingface://webAI-Official/TwIL-LM3/TwIL-LM3-Q8_0.gguf
|
||||
sha256: b6fc209a2b12c6d6d2316d3c3a563276adb7a7f76144c81f12eb2e600b465594
|
||||
- &nemotron-3-5-lightning-30b-a3b
|
||||
name: "nemotron-3.5-lightning-30b-a3b-q4"
|
||||
variants:
|
||||
|
||||
Reference in new issue
Block a user