metadata
base_model: malhajar/Mixtral-8x7B-v0.1-turkish
language:
- tr
- en
pipeline_tag: text-generation
license: apache-2.0
model_type: mixtral
library_name: transformers
inference: false
Mixtral 8x7B v0.1 Turkish
- Model creator: malhajar
- Original model: Mixtral-8x7B-v0.1-turkish
Description
This repo contains GGUF format model files for malhajar's Mixtral 8x7B v0.1 Turkish
Original model
- Developed by:
Mohamad Alhajar - Language(s) (NLP): Turkish
- Finetuned from model:
mistralai/Mistral-7B-Instruct-v0.2
Original model description
malhajar/Mixtral-8x7B-v0.1-turkish is a finetuned version of Mixtral-8x7B-v0.1 using SFT Training.
This model can answer information in turkish language as it is finetuned on a turkish dataset specifically alpaca-gpt4-tr
Quantizon types
| quantization method | bits | size | description | recommended |
|---|---|---|---|---|
| Q3_K_S | 3 | 20.4 GB | very small, high quality loss | ❌ |
| Q3_K_L | 3 | 26.4 GB | small, substantial quality loss | ❌ |
| Q4_0 | 4 | 26.4 GB | legacy; small, very high quality loss | ❌ |
| Q4_K_M | 4 | 28.4 GB | medium, balanced quality | ✅ |
| Q5_0 | 5 | 33.2 GB | legacy; medium, balanced quality | ❌ |
| Q5_K_S | 5 | 32.2 GB | large, low quality loss | ✅ |
| Q5_K_M | 5 | 33.2 GB | large, very low quality loss | ✅ |
| Q6_K | 6 | 38.4 GB | very large, extremely low quality loss | ❌ |
| Q8_0 | 8 | 49.6 GB | very large, extremely low quality loss | ❌ |
Prompt Template
### Instruction:
<prompt> (without the <>)
### Response: