Downloads · 30 days
10
2% of all-time downloads
RichardErkhov/froggeric_-_WestLake-10.7B-v2-gguf
froggeric_-_WestLake-10.7B-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.
Downloads · 30 days
10
2% of all-time downloads
All-time downloads
612
Public
Repo size
122 GB
Likes
0
Public
Click a slice to open those files.
.gguf141 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
WestLake-10.7B-v2 - GGUF
| Name | Quant method | Size |
|---|---|---|
| WestLake-10.7B-v2.Q2_K.gguf | Q2_K | 3.73GB |
| WestLake-10.7B-v2.IQ3_XS.gguf | IQ3_XS | 4.14GB |
| WestLake-10.7B-v2.IQ3_S.gguf | IQ3_S | 4.37GB |
| WestLake-10.7B-v2.Q3_K_S.gguf | Q3_K_S | 4.34GB |
| WestLake-10.7B-v2.IQ3_M.gguf | IQ3_M | 4.51GB |
| WestLake-10.7B-v2.Q3_K.gguf | Q3_K | 4.84GB |
| WestLake-10.7B-v2.Q3_K_M.gguf | Q3_K_M | 4.84GB |
| WestLake-10.7B-v2.Q3_K_L.gguf | Q3_K_L | 5.26GB |
| WestLake-10.7B-v2.IQ4_XS.gguf | IQ4_XS | 5.43GB |
| WestLake-10.7B-v2.Q4_0.gguf | Q4_0 | 5.66GB |
| WestLake-10.7B-v2.IQ4_NL.gguf | IQ4_NL | 5.72GB |
| WestLake-10.7B-v2.Q4_K_S.gguf | Q4_K_S | 5.7GB |
| WestLake-10.7B-v2.Q4_K.gguf | Q4_K | 6.02GB |
| WestLake-10.7B-v2.Q4_K_M.gguf | Q4_K_M | 6.02GB |
| WestLake-10.7B-v2.Q4_1.gguf | Q4_1 | 6.27GB |
| WestLake-10.7B-v2.Q5_0.gguf | Q5_0 | 6.89GB |
| WestLake-10.7B-v2.Q5_K_S.gguf | Q5_K_S | 6.89GB |
| WestLake-10.7B-v2.Q5_K.gguf | Q5_K | 7.08GB |
| WestLake-10.7B-v2.Q5_K_M.gguf | Q5_K_M | 7.08GB |
| WestLake-10.7B-v2.Q5_1.gguf | Q5_1 | 7.51GB |
| WestLake-10.7B-v2.Q6_K.gguf | Q6_K | 8.2GB |
| WestLake-10.7B-v2.Q8_0.gguf | Q8_0 | 10.62GB |
base_model:
GGUF version available here
EXL2 versions available here:
3.3bpw / 4.0bpw / 5.0bpw / 6.0bpw / 8.0bpw
This is my first viable self-merge of the fantastic WestLake-7B-v2 model, obtained after more than 12 rounds of testing different merge configurations. In my LLM Creativity Benchmark, it greatly improves over the original 7B model, and ranks between miqu-1-120b and goliath-120b! I would describe the improvements as a better writing style, with more details. It has a bit more difficulties following instructions, but not by much.
It is also the first model I have tested to obtain a perfect score with the following test:
Write a sequence of nominal groups that flow into one another, using the following rules:
- each nominal group is made of exactly 3 words
- the first word of each nominal group must be the last word of the previous nominal group
- the first word of the first nominal group is: "ball"
- the last word of the last nominal group is: "stone"
- there must be a theme, of your choosing, pertaining to all nominal groups
- there must be exactly 7 nominal groups, leading from the first word (ball) to the last word (stone)
- a word already used at the beginning and end of a nominal group cannot be reused
Present your solution as a list numbered with roman numerals.
Finally, explain why you chose your specific theme.
This is a merge of pre-trained language models created using mergekit.
This model was merged using the passthrough merge method.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
dtype: float16
merge_method: passthrough
slices:
- sources:
- model: senseable/WestLake-7B-v2
layer_range: [0,9]
- sources:
- model: senseable/WestLake-7B-v2
layer_range: [5,14]
- sources:
- model: senseable/WestLake-7B-v2
layer_range: [10,19]
- sources:
- model: senseable/WestLake-7B-v2
layer_range: [15,24]
- sources:
- model: senseable/WestLake-7B-v2
layer_range: [20,32]
Update Notes: Version 2 trained 1 additional epoch cycle for 3 total
Welcome to the documentation of Westlake-7B, a cutting-edge language model designed for exceptional role-play and text generation tasks. This README file aims to provide an overview of our capabilities, usage guidelines, and potential applications.
Westlake-7B is built upon a vast corpus of diverse texts, enabling it to generate contextually relevant responses in various scenarios. With its impressive size of 7 billion parameters, this model excels at understanding nuances in language and producing creative outputs.
To utilize Westlake-7Bv2 for your projects or experiments, follow these steps:
Westlake-7Bv2's versatility makes it suitable for various applications across different industries: