Downloads · 30 days
27
4% of all-time downloads
Ba2han_-_Cucumber-7b-10k-gguf
RichardErkhov/Ba2han_-_Cucumber-7b-10k-gguf
Ba2han_-_Cucumber-7b-10k-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
602
Public
Repo size
82.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf95.2 GB · 100%
At a glance
- Access
- Public
- Created
- Sep 9, 2024
- Updated
- Sep 9, 2024
- SHA
- ed9b68c0
Model card
From the Hugging Face model README
Quantization made by Richard Erkhov.
Cucumber-7b-10k - GGUF
- Model creator: https://huggingface.co/Ba2han/
- Original model: https://huggingface.co/Ba2han/Cucumber-7b-10k/
| Name | Quant method | Size |
|---|---|---|
| Cucumber-7b-10k.Q2_K.gguf | Q2_K | 2.53GB |
| Cucumber-7b-10k.IQ3_XS.gguf | IQ3_XS | 2.81GB |
| Cucumber-7b-10k.IQ3_S.gguf | IQ3_S | 2.96GB |
| Cucumber-7b-10k.Q3_K_S.gguf | Q3_K_S | 2.95GB |
| Cucumber-7b-10k.IQ3_M.gguf | IQ3_M | 3.06GB |
| Cucumber-7b-10k.Q3_K.gguf | Q3_K | 3.28GB |
| Cucumber-7b-10k.Q3_K_M.gguf | Q3_K_M | 3.28GB |
| Cucumber-7b-10k.Q3_K_L.gguf | Q3_K_L | 3.56GB |
| Cucumber-7b-10k.IQ4_XS.gguf | IQ4_XS | 3.67GB |
| Cucumber-7b-10k.Q4_0.gguf | Q4_0 | 3.83GB |
| Cucumber-7b-10k.IQ4_NL.gguf | IQ4_NL | 3.87GB |
| Cucumber-7b-10k.Q4_K_S.gguf | Q4_K_S | 3.86GB |
| Cucumber-7b-10k.Q4_K.gguf | Q4_K | 4.07GB |
| Cucumber-7b-10k.Q4_K_M.gguf | Q4_K_M | 4.07GB |
| Cucumber-7b-10k.Q4_1.gguf | Q4_1 | 4.24GB |
| Cucumber-7b-10k.Q5_0.gguf | Q5_0 | 4.65GB |
| Cucumber-7b-10k.Q5_K_S.gguf | Q5_K_S | 4.65GB |
| Cucumber-7b-10k.Q5_K.gguf | Q5_K | 4.78GB |
| Cucumber-7b-10k.Q5_K_M.gguf | Q5_K_M | 4.78GB |
| Cucumber-7b-10k.Q5_1.gguf | Q5_1 | 5.07GB |
| Cucumber-7b-10k.Q6_K.gguf | Q6_K | 5.53GB |
| Cucumber-7b-10k.Q8_0.gguf | Q8_0 | 7.17GB |
Original model description:
license: cc-by-4.0 datasets:
- HuggingFaceH4/no_robots
- Ba2han/Reddit-instruct-curated_rated-1.2k
- Ba2han/databricks-dolly_rated
- Open-Orca/OpenOrca
The training dataset consists of 2k (longest) examples from no_robots, reddit_instruct, dolly, OpenOrca plus two other personal datasets.
Please use with ChatML and the default system message or enter your own. It was trained with various system messages, the one in the config being the default one.
The model is:
-
Very good at generating long and coherent text.
-
Creative due to data from Reddit ELI5 and a few other sources.
-
Better at handling longer input.
-
Not great with short text both in input and generation.
The aim is to see how the "Long is More for Alignment" paper holds. This is basically a combination of LIMA + LMA. There should be no benchmark contamination as far as I am aware of. Around 70% of the data is from the mentioned datasets. I am happy with how it turned out.
- Created
- Sep 9, 2024
- Updated
- Sep 9, 2024