Astridboros-3B-GGUF / README.md
afrideva's picture
Upload README.md with huggingface_hub
888eab1
|
raw
history blame
2.05 kB
---
base_model: Aryanne/Astridboros-3B
inference: false
language:
- en
library_name: transformers
license: cc-by-sa-4.0
model_creator: Aryanne
model_name: Astridboros-3B
pipeline_tag: text-generation
quantized_by: afrideva
tags:
- gpt
- llm
- large language model
- gguf
- ggml
- quantized
- q2_k
- q3_k_m
- q4_k_m
- q5_k_m
- q6_k
- q8_0
---
# Aryanne/Astridboros-3B-GGUF
Quantized GGUF model files for [Astridboros-3B](https://huggingface.co/Aryanne/Astridboros-3B) from [Aryanne](https://huggingface.co/Aryanne)
| Name | Quant method | Size |
| ---- | ---- | ---- |
| [astridboros-3b.fp16.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.fp16.gguf) | fp16 | 5.59 GB |
| [astridboros-3b.q2_k.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q2_k.gguf) | q2_k | 1.20 GB |
| [astridboros-3b.q3_k_m.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q3_k_m.gguf) | q3_k_m | 1.39 GB |
| [astridboros-3b.q4_k_m.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q4_k_m.gguf) | q4_k_m | 1.71 GB |
| [astridboros-3b.q5_k_m.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q5_k_m.gguf) | q5_k_m | 1.99 GB |
| [astridboros-3b.q6_k.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q6_k.gguf) | q6_k | 2.30 GB |
| [astridboros-3b.q8_0.gguf](https://huggingface.co/afrideva/Astridboros-3B-GGUF/resolve/main/astridboros-3b.q8_0.gguf) | q8_0 | 2.97 GB |
## Original Model Card:
This model is a merge/fusion of [PAIXAI/Astrid-3B](https://huggingface.co/PAIXAI/Astrid-3B) and [jondurbin/airoboros-3b-3p0](https://huggingface.co/jondurbin/airoboros-3b-3p0) , 16 layers of each glued together(see Astridboros.yml or below).
```yaml
slices:
- sources:
- model: PAIXAI/Astrid-3B
layer_range: [0, 16]
- sources:
- model: jondurbin/airoboros-3b-3p0
layer_range: [16, 32]
merge_method: passthrough
dtype: float16
```