ExternalVendorCode/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-03-13 07:54:34 +00:00
Code Issues Actions 14 Packages Projects Releases Wiki Activity

chore(model gallery): update qihoo360_tinyr1-32b-preview (#4937)

Browse Source 
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2025-03-04 09:53:23 +01:00 committed by GitHub
parent 122970d70d
commit 6f9762049c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
gallery/index.yaml
View File

@ -6673,18 +6673,18 @@
name: "qihoo360_tinyr1-32b-preview" name: "qihoo360_tinyr1-32b-preview"
urls: urls:
- https://huggingface.co/qihoo360/TinyR1-32B-Preview - https://huggingface.co/qihoo360/TinyR1-32B-Preview
- https://huggingface.co/bartowski/qihoo360_TinyR1-32B-Preview-GGUF - https://huggingface.co/bartowski/qihoo360_TinyR1-32B-Preview-v0.2-GGUF
description: | description: |
We introduce our first-generation reasoning model, Tiny-R1-32B-Preview, which outperforms the 70B model Deepseek-R1-Distill-Llama-70B and nearly matches the full R1 model in math. We introduce our first-generation reasoning model, Tiny-R1-32B-Preview, which outperforms the 70B model Deepseek-R1-Distill-Llama-70B and nearly matches the full R1 model in math.
We applied supervised fine-tuning (SFT) to Deepseek-R1-Distill-Qwen-32B across three target domains—Mathematics, Code, and Science — using the 360-LLaMA-Factory training framework to produce three domain-specific models. We used questions from open-source data as seeds. Meanwhile, responses for mathematics, coding, and science tasks were generated by R1, creating specialized models for each domain. Building on this, we leveraged the Mergekit tool from the Arcee team to combine multiple models, creating Tiny-R1-32B-Preview, which demonstrates strong overall performance. We applied supervised fine-tuning (SFT) to Deepseek-R1-Distill-Qwen-32B across three target domains—Mathematics, Code, and Science — using the 360-LLaMA-Factory training framework to produce three domain-specific models. We used questions from open-source data as seeds. Meanwhile, responses for mathematics, coding, and science tasks were generated by R1, creating specialized models for each domain. Building on this, we leveraged the Mergekit tool from the Arcee team to combine multiple models, creating Tiny-R1-32B-Preview, which demonstrates strong overall performance.
overrides: overrides:
parameters: parameters:
model: qihoo360_TinyR1-32B-Preview-Q4_K_M.gguf model: qihoo360_TinyR1-32B-Preview-v0.2-Q4_K_M.gguf
files: files:
- filename: qihoo360_TinyR1-32B-Preview-Q4_K_M.gguf - filename: qihoo360_TinyR1-32B-Preview-v0.2-Q4_K_M.gguf
sha256: 4eb3df9cc3d74e1dd75924a39b2ce7466a1053240375d5c969ea24b6045c2a09 sha256: 250e38d6164798a6aa0d5a9208722f835fc6a1a582aeff884bdedb123d209d47
uri: huggingface://bartowski/qihoo360_TinyR1-32B-Preview-GGUF/qihoo360_TinyR1-32B-Preview-Q4_K_M.gguf uri: huggingface://bartowski/qihoo360_TinyR1-32B-Preview-v0.2-GGUF/qihoo360_TinyR1-32B-Preview-v0.2-Q4_K_M.gguf
- !!merge <<: *deepseek-r1 - !!merge <<: *deepseek-r1
name: "thedrummer_fallen-llama-3.3-r1-70b-v1" name: "thedrummer_fallen-llama-3.3-r1-70b-v1"
icon: https://cdn-uploads.huggingface.co/production/uploads/65f2fd1c25b848bd061b5c2e/7BdBxwafsvzqPC98h_gaA.png icon: https://cdn-uploads.huggingface.co/production/uploads/65f2fd1c25b848bd061b5c2e/7BdBxwafsvzqPC98h_gaA.png