mirror of
https://github.com/mudler/LocalAI.git
synced 2026-10-28 09:08:54 -04:00
chore(gallery): add Swift Heretic GSQ-RCO variants
Add three llama.cpp quantizations with the matching vision projector,
MTP decoding, and installation examples. Pin downloads to one Hugging
Face revision and verify SHA256 values against LFS metadata and headers.
Assisted-by: Codex:gpt-6
(cherry picked from commit b5f4a2f958, mudler/LocalAI#12498)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
localai-org-maint-bot authored and Ettore Di Giacinto committed
1 parent
53e238a320
commit
f220c8e127
1 file changed
+159
@@ -6163,6 +6163,165 @@
|
||||
- filename: llama-cpp/mmproj/qwen3.8-27b-ridge/mmproj-Qwen3.8-27B-BF16.gguf
|
||||
uri: huggingface://empero-ai/Qwen3.8-27B-Ridge-GGUF/mmproj-Qwen3.8-27B-BF16.gguf
|
||||
sha256: 52228402ce4823f10705d901813cd43ced71859524cf2d8bf83305ad6b7dcbc2
|
||||
- name: "swift-1.5-qwen3.8-27b-heretic-gsq-rco-iq3-s"
|
||||
variants:
|
||||
- model: swift-1.5-qwen3.8-27b-heretic-gsq-rco-iq3-xxs
|
||||
- model: swift-1.5-qwen3.8-27b-heretic-gsq-rco-iq2-s
|
||||
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
|
||||
urls:
|
||||
- https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b
|
||||
- https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF
|
||||
license: "swift-open-license-1.0"
|
||||
description: |
|
||||
A community Heretic modification of Swift 1.5 Qwen3.8-27B for reasoning,
|
||||
coding, and image chat. This IQ3_S GGUF uses GSQ-RCO mixed-precision
|
||||
quantization and includes the matching F16 vision projector.
|
||||
Uses llama.cpp with MTP decoding and the Swift Open License v1.0.
|
||||
tags:
|
||||
- llm
|
||||
- gguf
|
||||
- cpu
|
||||
- gpu
|
||||
- reasoning
|
||||
- vision
|
||||
- multimodal
|
||||
- mtp
|
||||
overrides:
|
||||
backend: llama-cpp
|
||||
context_size: 32768
|
||||
function:
|
||||
automatic_tool_parsing_fallback: true
|
||||
grammar:
|
||||
disable: true
|
||||
known_usecases:
|
||||
- chat
|
||||
- vision
|
||||
mmproj: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
options:
|
||||
- use_jinja:true
|
||||
- spec_type:draft-mtp
|
||||
- spec_n_max:3
|
||||
parameters:
|
||||
min_p: 0
|
||||
model: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_S-mtp.gguf
|
||||
presence_penalty: 0
|
||||
repeat_penalty: 1
|
||||
temperature: 1
|
||||
top_k: 20
|
||||
top_p: 0.95
|
||||
template:
|
||||
use_tokenizer_template: true
|
||||
files:
|
||||
- filename: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_S-mtp.gguf
|
||||
sha256: 00f5763c0bc549478cfe27e76bd939c6008ac8179b256580e647e5c5d9998838
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_S-mtp.gguf
|
||||
- filename: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
sha256: b409acbc885f5c852a8f9149b9bb808014b3aef5ecf91ae72e59599ee8fbce23
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
- name: "swift-1.5-qwen3.8-27b-heretic-gsq-rco-iq3-xxs"
|
||||
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
|
||||
urls:
|
||||
- https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b
|
||||
- https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF
|
||||
license: "swift-open-license-1.0"
|
||||
description: |
|
||||
A community Heretic modification of Swift 1.5 Qwen3.8-27B for reasoning,
|
||||
coding, and image chat. This IQ3_XXS GGUF uses GSQ-RCO mixed-precision
|
||||
quantization and includes the matching F16 vision projector.
|
||||
Uses llama.cpp with MTP decoding and the Swift Open License v1.0.
|
||||
tags:
|
||||
- llm
|
||||
- gguf
|
||||
- cpu
|
||||
- gpu
|
||||
- reasoning
|
||||
- vision
|
||||
- multimodal
|
||||
- mtp
|
||||
overrides:
|
||||
backend: llama-cpp
|
||||
context_size: 32768
|
||||
function:
|
||||
automatic_tool_parsing_fallback: true
|
||||
grammar:
|
||||
disable: true
|
||||
known_usecases:
|
||||
- chat
|
||||
- vision
|
||||
mmproj: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
options:
|
||||
- use_jinja:true
|
||||
- spec_type:draft-mtp
|
||||
- spec_n_max:3
|
||||
parameters:
|
||||
min_p: 0
|
||||
model: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_XXS-mtp.gguf
|
||||
presence_penalty: 0
|
||||
repeat_penalty: 1
|
||||
temperature: 1
|
||||
top_k: 20
|
||||
top_p: 0.95
|
||||
template:
|
||||
use_tokenizer_template: true
|
||||
files:
|
||||
- filename: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_XXS-mtp.gguf
|
||||
sha256: a9d571cd9a072b58eec2f6651df2cbdd68f3736fafadc623af7bdd8521fee4bd
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ3_XXS-mtp.gguf
|
||||
- filename: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
sha256: b409acbc885f5c852a8f9149b9bb808014b3aef5ecf91ae72e59599ee8fbce23
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
- name: "swift-1.5-qwen3.8-27b-heretic-gsq-rco-iq2-s"
|
||||
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
|
||||
urls:
|
||||
- https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b
|
||||
- https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF
|
||||
license: "swift-open-license-1.0"
|
||||
description: |
|
||||
A community Heretic modification of Swift 1.5 Qwen3.8-27B for reasoning,
|
||||
coding, and image chat. This IQ2_S GGUF uses GSQ-RCO mixed-precision
|
||||
quantization and includes the matching F16 vision projector.
|
||||
Uses llama.cpp with MTP decoding and the Swift Open License v1.0.
|
||||
tags:
|
||||
- llm
|
||||
- gguf
|
||||
- cpu
|
||||
- gpu
|
||||
- reasoning
|
||||
- vision
|
||||
- multimodal
|
||||
- mtp
|
||||
overrides:
|
||||
backend: llama-cpp
|
||||
context_size: 32768
|
||||
function:
|
||||
automatic_tool_parsing_fallback: true
|
||||
grammar:
|
||||
disable: true
|
||||
known_usecases:
|
||||
- chat
|
||||
- vision
|
||||
mmproj: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
options:
|
||||
- use_jinja:true
|
||||
- spec_type:draft-mtp
|
||||
- spec_n_max:3
|
||||
parameters:
|
||||
min_p: 0
|
||||
model: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ2_S-mtp.gguf
|
||||
presence_penalty: 0
|
||||
repeat_penalty: 1
|
||||
temperature: 1
|
||||
top_k: 20
|
||||
top_p: 0.95
|
||||
template:
|
||||
use_tokenizer_template: true
|
||||
files:
|
||||
- filename: llama-cpp/models/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ2_S-mtp.gguf
|
||||
sha256: 7f3e481d8ca31d67cc97f4dc3b8c3b469a9d4ee6cfca7f1d099682d11f691fd5
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-IQ2_S-mtp.gguf
|
||||
- filename: llama-cpp/mmproj/swift-1.5-qwen3.8-27b-heretic-gsq-rco/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
sha256: b409acbc885f5c852a8f9149b9bb808014b3aef5ecf91ae72e59599ee8fbce23
|
||||
uri: https://huggingface.co/jegreene/Swift-1.5-Qwen3.8-27B-heretic-GSQ-RCO-GGUF/resolve/a1e8a6359d32cb83ab4e20e8b194a7f78f546f36/Swift-1.5-Qwen3.8-27B-heretic-mmproj-F16.gguf
|
||||
- name: swift-1.5-qwen3.8-27b-gsq-rco
|
||||
variants:
|
||||
- model: swift-1.5-qwen3.8-27b-gsq-rco-iq2-s
|
||||
|
||||
Reference in new issue
Block a user