ExternalVendorCode/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2024-12-24 06:46:39 +00:00
Code Issues Actions 13 Packages Projects Releases Wiki Activity

models(gallery): add meta-llama-3-instruct-12.2b-brainstorm-20x-form-8 (#3097)

Browse Source 
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-07-31 17:06:06 +02:00 committed by GitHub
parent 115b523732
commit 4c7e8f4d54
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 13 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
gallery/index.yaml
View File

@ -2409,6 +2409,19 @@
- filename: L3-Stheno-Maid-Blackroot-Grand-HORROR-16B-D_AU-Q4_K_M.gguf
sha256: ae29f38d73dfb04415821405cf8b319fc42d78d0cdd0da91db147d12e68030fe
uri: huggingface://DavidAU/L3-Stheno-Maid-Blackroot-Grand-HORROR-16B-GGUF/L3-Stheno-Maid-Blackroot-Grand-HORROR-16B-D_AU-Q4_K_M.gguf
- !!merge <<: *llama3
name: "meta-llama-3-instruct-12.2b-brainstorm-20x-form-8"
urls:
- https://huggingface.co/DavidAU/Meta-Llama-3-Instruct-12.2B-BRAINSTORM-20x-FORM-8-GGUF
description: |
Meta-Llama-3-8B Instruct (now at 12.2B) with Brainstorm process that increases its performance at the core level for any creative use case. It has calibrations that allow it to exceed the logic solving abilities of the original model. The Brainstorm process expands the reasoning center of the LLM, reassembles and calibrates it, introducing subtle changes into the reasoning process. This enhances the model's detail, concept, connection to the "world", general concept connections, prose quality, and prose length without affecting instruction following. It improves coherence, description, simile, metaphors, emotional engagement, and takes fewer liberties with instructions while following them more closely. The model's performance is further enhanced by other technologies like "Ultra" (precision), "Neo Imatrix" (custom imatrix datasets), and "X-quants" (custom application of the imatrix process). It has been tested on multiple LLaMA2, LLaMA3, and Mistral models of various parameter sizes.
overrides:
parameters:
model: Meta-Llama-3-8B-Instruct-exp20-8-Q4_K_M.gguf
files:
- filename: Meta-Llama-3-8B-Instruct-exp20-8-Q4_K_M.gguf
sha256: 5568ab6195ab5da703f728cc118108ddcbe97255e3ba4a543b531acdf082b999
uri: huggingface://DavidAU/Meta-Llama-3-Instruct-12.2B-BRAINSTORM-20x-FORM-8-GGUF/Meta-Llama-3-8B-Instruct-exp20-8-Q4_K_M.gguf
- &dolphin
name: "dolphin-2.9-llama3-8b"
url: "github:mudler/LocalAI/gallery/hermes-2-pro-mistral.yaml@master"