Skip to content

Commit 4b1b942

Browse files
authored
chore(model gallery): add sicariussicariistuff_redemption_wind_24b (#4781)
Signed-off-by: Ettore Di Giacinto <[email protected]>
1 parent 230fe00 commit 4b1b942

File tree

1 file changed

+22
-0
lines changed

1 file changed

+22
-0
lines changed

gallery/index.yaml

Lines changed: 22 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -6945,6 +6945,28 @@
69456945
- filename: cognitivecomputations_Dolphin3.0-Mistral-24B-Q4_K_M.gguf
69466946
sha256: 6f193bbf98628140194df257c7466e2c6f80a7ef70a6ebae26c53b2f2ef21994
69476947
uri: huggingface://bartowski/cognitivecomputations_Dolphin3.0-Mistral-24B-GGUF/cognitivecomputations_Dolphin3.0-Mistral-24B-Q4_K_M.gguf
6948+
- !!merge <<: *mistral03
6949+
name: "sicariussicariistuff_redemption_wind_24b"
6950+
url: "github:mudler/LocalAI/gallery/chatml.yaml@master"
6951+
icon: https://huggingface.co/SicariusSicariiStuff/Redemption_Wind_24B/resolve/main/Images/Redemption_Wind_24B.png
6952+
urls:
6953+
- https://huggingface.co/SicariusSicariiStuff/Redemption_Wind_24B
6954+
- https://huggingface.co/bartowski/SicariusSicariiStuff_Redemption_Wind_24B-GGUF
6955+
description: |
6956+
This is a lightly fine-tuned version of the Mistral 24B base model, designed as an accessible and adaptable foundation for further fine-tuning and merging fodder. Key modifications include:
6957+
ChatML-ified, with no additional tokens introduced.
6958+
High quality private instruct—not generated by ChatGPT or Claude, ensuring no slop and good markdown understanding.
6959+
No refusals—since it’s a base model, refusals should be minimal to non-existent, though, in early testing, occasional warnings still appear (I assume some were baked into the pre-train).
6960+
High-quality private creative writing dataset Mainly to dilute baked-in slop further, but it can actually write some stories, not bad for loss ~8.
6961+
Small, high-quality private RP dataset This was done so further tuning for RP will be easier. The dataset was kept small and contains ZERO SLOP, some entries are of 16k token length.
6962+
Exceptional adherence to character cards This was done to make it easier for further tunes intended for roleplay.
6963+
overrides:
6964+
parameters:
6965+
model: SicariusSicariiStuff_Redemption_Wind_24B-Q4_K_M.gguf
6966+
files:
6967+
- filename: SicariusSicariiStuff_Redemption_Wind_24B-Q4_K_M.gguf
6968+
sha256: 40025eb00d83c9e9393555962962a2dfc5251fe7bd70812835ff0bcc55ecc463
6969+
uri: huggingface://bartowski/SicariusSicariiStuff_Redemption_Wind_24B-GGUF/SicariusSicariiStuff_Redemption_Wind_24B-Q4_K_M.gguf
69486970
- &mudler
69496971
url: "github:mudler/LocalAI/gallery/mudler.yaml@master" ### START mudler's LocalAI specific-models
69506972
name: "LocalAI-llama3-8b-function-call-v0.2"

0 commit comments

Comments
 (0)