models(gallery): add suzume-orpo (#2932)

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-07-20 16:16:29 +02:00 committed by GitHub
parent 46b86f7e6e
commit 450dbed820
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -3176,6 +3176,28 @@
- filename: llama-3-tulu-2-dpo-70b.i1-Q4_K_M.gguf
sha256: fc309bbdf1e2bdced954c4c8dc1f9a885c547017ee5e750bfde645af89e3d3a5
uri: huggingface://mradermacher/llama-3-tulu-2-dpo-70b-i1-GGUF/llama-3-tulu-2-dpo-70b.i1-Q4_K_M.gguf
- !!merge <<: *llama3
license: cc-by-nc-4.0
name: "suzume-llama-3-8b-multilingual-orpo-borda-top25"
icon: https://cdn-uploads.huggingface.co/production/uploads/64b63f8ad57e02621dc93c8b/kWQSu02YfgYdUQqv4s5lq.png
urls:
- https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25
- https://huggingface.co/RichardErkhov/lightblue_-_suzume-llama-3-8B-multilingual-orpo-borda-top25-gguf
description: |
This is Suzume ORPO, an ORPO trained fine-tune of the lightblue/suzume-llama-3-8B-multilingual model using our lightblue/mitsu dataset.
We have trained several versions of this model using ORPO and so recommend that you use the best performing model from our tests, lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half.
Note that this model has a non-commerical license as we used the Command R and Command R+ models to generate our training data for this model (lightblue/mitsu).
We are currently working on a developing a commerically usable model, so stay tuned for that!
overrides:
parameters:
model: suzume-llama-3-8B-multilingual-orpo-borda-top25.Q4_K_M.gguf
files:
- filename: suzume-llama-3-8B-multilingual-orpo-borda-top25.Q4_K_M.gguf
sha256: ef75a02c5f38e14a8873c7989188dac6974851b4654279fe1921d2c8018cc388
uri: huggingface://RichardErkhov/lightblue_-_suzume-llama-3-8B-multilingual-orpo-borda-top25-gguf/suzume-llama-3-8B-multilingual-orpo-borda-top25.Q4_K_M.gguf
- &command-R
### START Command-r
url: "github:mudler/LocalAI/gallery/command-r.yaml@master"