Downloads · 30 days
50
2% of all-time downloads
athirdpath_-_Iambe-20b-DARE-v2-gguf
RichardErkhov/athirdpath_-_Iambe-20b-DARE-v2-gguf
athirdpath_-_Iambe-20b-DARE-v2-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
All-time downloads
2.1K
Public
Repo size
205 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf262 GB · 100%
At a glance
- Access
- Public
- Created
- Jun 2, 2024
- Updated
- Jun 3, 2024
- SHA
- c9841771
Model card
From the Hugging Face model README
Quantization made by Richard Erkhov.
Iambe-20b-DARE-v2 - GGUF
- Model creator: https://huggingface.co/athirdpath/
- Original model: https://huggingface.co/athirdpath/Iambe-20b-DARE-v2/
| Name | Quant method | Size |
|---|---|---|
| Iambe-20b-DARE-v2.Q2_K.gguf | Q2_K | 6.91GB |
| Iambe-20b-DARE-v2.IQ3_XS.gguf | IQ3_XS | 7.63GB |
| Iambe-20b-DARE-v2.IQ3_S.gguf | IQ3_S | 8.06GB |
| Iambe-20b-DARE-v2.Q3_K_S.gguf | Q3_K_S | 8.06GB |
| Iambe-20b-DARE-v2.IQ3_M.gguf | IQ3_M | 8.53GB |
| Iambe-20b-DARE-v2.Q3_K.gguf | Q3_K | 9.04GB |
| Iambe-20b-DARE-v2.Q3_K_M.gguf | Q3_K_M | 9.04GB |
| Iambe-20b-DARE-v2.Q3_K_L.gguf | Q3_K_L | 9.9GB |
| Iambe-20b-DARE-v2.IQ4_XS.gguf | IQ4_XS | 10.01GB |
| Iambe-20b-DARE-v2.Q4_0.gguf | Q4_0 | 10.52GB |
| Iambe-20b-DARE-v2.IQ4_NL.gguf | IQ4_NL | 10.57GB |
| Iambe-20b-DARE-v2.Q4_K_S.gguf | Q4_K_S | 10.59GB |
| Iambe-20b-DARE-v2.Q4_K.gguf | Q4_K | 11.22GB |
| Iambe-20b-DARE-v2.Q4_K_M.gguf | Q4_K_M | 11.22GB |
| Iambe-20b-DARE-v2.Q4_1.gguf | Q4_1 | 11.67GB |
| Iambe-20b-DARE-v2.Q5_0.gguf | Q5_0 | 12.83GB |
| Iambe-20b-DARE-v2.Q5_K_S.gguf | Q5_K_S | 12.83GB |
| Iambe-20b-DARE-v2.Q5_K.gguf | Q5_K | 13.18GB |
| Iambe-20b-DARE-v2.Q5_K_M.gguf | Q5_K_M | 13.18GB |
| Iambe-20b-DARE-v2.Q5_1.gguf | Q5_1 | 13.98GB |
| Iambe-20b-DARE-v2.Q6_K.gguf | Q6_K | 15.28GB |
| Iambe-20b-DARE-v2.Q8_0.gguf | Q8_0 | 19.79GB |
Original model description:
license: cc-by-nc-4.0 language:
- en
<p align="center"><font size="5"> <i>Strange quirk: This model seems to need a context size of EXACTLY 4096 ONLY. I'm assuming this is a dares_ties effect?</i> </font></p> <p align="center"><img src="https://i.ibb.co/pbpJHpk/iambe-sml.png"/><font size="6"> <b>Iambe-20b-DARE-v2</b> </font></p> <p align="center"><font size="4"> <b>Alpaca prompt formatting</b> </font></p>
Description
Named after a charming daughter of Echo and Pan in Greek myth, Iambe-20b-DARE-v2 is an improved DARE merge building on my recent experiments.
Iambe is intended to have the best realistically possible understanding of anatomy and of a scene's state for a 20b merge, while remaining personable and authentic in "voice".
Update Methodology
Noromaid and the general "no-robots" vibe didn't come through like I'd hoped in v1. My hypothesis is that the "soul" MythoMax and Noromaid have is probably distributed widely over many low-value deltas, due to the "ephemeral" nature of such a thing.
My old base model was likely giving DARE conniption fits, so I replaced that with a truly vanilla 20b base model.
CleverGirl was updated to the DARE version, as Sir Hillary said, simply because it was there.
Without a large base of dare_ties models to compare to, I'm basically feeling my way through this intuitively, so here's to good results!
Recipe
merge_method: dare_ties
-
base_model: athirdpath/BigLlama-20b-v1.1
-
model: Noromaid-20b-v0.1.1
weight: 0.38 / density: 0.60
-
model: athirdpath/athirdpath/Eileithyia-20b
weight: 0.22 / density: 0.40
-
model: athirdpath/CleverGirl-20b-Blended-v1.1-DARE
weight: 0.40 / density: 0.33
int8_mask: true
dtype: bfloat16
- Created
- Jun 2, 2024
- Updated
- Jun 3, 2024