Downloads · 30 days
83
7% of all-time downloads
RichardErkhov/abacusai_-_Giraffe-13b-32k-v3-gguf
abacusai_-_Giraffe-13b-32k-v3-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
83
7% of all-time downloads
All-time downloads
1.2K
Public
Repo size
147 GB
Likes
0
Public
Click a slice to open those files.
.gguf171 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
Giraffe-13b-32k-v3 - GGUF
| Name | Quant method | Size |
|---|---|---|
| Giraffe-13b-32k-v3.Q2_K.gguf | Q2_K | 4.52GB |
| Giraffe-13b-32k-v3.IQ3_XS.gguf | IQ3_XS | 4.99GB |
| Giraffe-13b-32k-v3.IQ3_S.gguf | IQ3_S | 5.27GB |
| Giraffe-13b-32k-v3.Q3_K_S.gguf | Q3_K_S | 5.27GB |
| Giraffe-13b-32k-v3.IQ3_M.gguf | IQ3_M | 5.57GB |
| Giraffe-13b-32k-v3.Q3_K.gguf | Q3_K | 5.9GB |
| Giraffe-13b-32k-v3.Q3_K_M.gguf | Q3_K_M | 5.9GB |
| Giraffe-13b-32k-v3.Q3_K_L.gguf | Q3_K_L | 6.45GB |
| Giraffe-13b-32k-v3.IQ4_XS.gguf | IQ4_XS | 6.54GB |
| Giraffe-13b-32k-v3.Q4_0.gguf | Q4_0 | 6.86GB |
| Giraffe-13b-32k-v3.IQ4_NL.gguf | IQ4_NL | 6.9GB |
| Giraffe-13b-32k-v3.Q4_K_S.gguf | Q4_K_S | 6.91GB |
| Giraffe-13b-32k-v3.Q4_K.gguf | Q4_K | 7.33GB |
| Giraffe-13b-32k-v3.Q4_K_M.gguf | Q4_K_M | 7.33GB |
| Giraffe-13b-32k-v3.Q4_1.gguf | Q4_1 | 7.61GB |
| Giraffe-13b-32k-v3.Q5_0.gguf | Q5_0 | 8.36GB |
| Giraffe-13b-32k-v3.Q5_K_S.gguf | Q5_K_S | 8.36GB |
| Giraffe-13b-32k-v3.Q5_K.gguf | Q5_K | 8.6GB |
| Giraffe-13b-32k-v3.Q5_K_M.gguf | Q5_K_M | 8.6GB |
| Giraffe-13b-32k-v3.Q5_1.gguf | Q5_1 | 9.1GB |
| Giraffe-13b-32k-v3.Q6_K.gguf | Q6_K | 9.95GB |
| Giraffe-13b-32k-v3.Q8_0.gguf | Q8_0 | 12.88GB |
datasets:

Llama2 13B finetuned to process longer (32K) contexts using interpolation and then further instruct finetuned with ShareGPT and Orca-Chat.
This is next version of the model trained and evaluated as part of the experiments described in the repo http://github.com/abacusai/Long-Context. This version was trained with a scaling factor of 8 and shows better reasoning and math abilites on State of the Art benchmarks
Since the model is instruct finetuned it can also be directly used for various prompted tasks. We have tested it on open book question answering using the long context to supply search results.
The model has not been evaluated for safety and is only intended for research and experiments.