Quantization made by Richard Erkhov.
llama2_7b_chat_uncensored - GGUF
- Model creator: https://huggingface.co/georgesung/
- Original model: https://huggingface.co/georgesung/llama2_7b_chat_uncensored/
| Name | Quant method | Size |
|---|---|---|
| llama2_7b_chat_uncensored.Q2_K.gguf | Q2_K | 2.36GB |
| llama2_7b_chat_uncensored.Q3_K_S.gguf | Q3_K_S | 2.75GB |
| llama2_7b_chat_uncensored.Q3_K.gguf | Q3_K | 3.07GB |
| llama2_7b_chat_uncensored.Q3_K_M.gguf | Q3_K_M | 3.07GB |
| llama2_7b_chat_uncensored.Q3_K_L.gguf | Q3_K_L | 3.35GB |
| llama2_7b_chat_uncensored.IQ4_XS.gguf | IQ4_XS | 3.4GB |
| llama2_7b_chat_uncensored.Q4_0.gguf | Q4_0 | 3.56GB |
| llama2_7b_chat_uncensored.IQ4_NL.gguf | IQ4_NL | 3.58GB |
| llama2_7b_chat_uncensored.Q4_K_S.gguf | Q4_K_S | 3.59GB |
| llama2_7b_chat_uncensored.Q4_K.gguf | Q4_K | 3.8GB |
| llama2_7b_chat_uncensored.Q4_K_M.gguf | Q4_K_M | 3.8GB |
| llama2_7b_chat_uncensored.Q4_1.gguf | Q4_1 | 3.95GB |
| llama2_7b_chat_uncensored.Q5_0.gguf | Q5_0 | 4.33GB |
| llama2_7b_chat_uncensored.Q5_K_S.gguf | Q5_K_S | 4.33GB |
| llama2_7b_chat_uncensored.Q5_K.gguf | Q5_K | 4.45GB |
| llama2_7b_chat_uncensored.Q5_K_M.gguf | Q5_K_M | 4.45GB |
| llama2_7b_chat_uncensored.Q5_1.gguf | Q5_1 | 4.72GB |
| llama2_7b_chat_uncensored.Q6_K.gguf | Q6_K | 5.15GB |
| llama2_7b_chat_uncensored.Q8_0.gguf | Q8_0 | 6.67GB |
Original model description:
license: other
datasets:
- georgesung/wizard_vicuna_70k_unfiltered
Overview
Fine-tuned Llama-2 7B with an uncensored/unfiltered Wizard-Vicuna conversation dataset (originally from ehartford/wizard_vicuna_70k_unfiltered).
Used QLoRA for fine-tuning. Trained for one epoch on a 24GB GPU (NVIDIA A10G) instance, took ~19 hours to train.
The version here is the fp16 HuggingFace model.
GGML & GPTQ versions
Thanks to TheBloke, he has created the GGML and GPTQ versions:
- https://huggingface.co/TheBloke/llama2_7b_chat_uncensored-GGML
- https://huggingface.co/TheBloke/llama2_7b_chat_uncensored-GPTQ
Running in Ollama
https://ollama.com/library/llama2-uncensored
Prompt style
The model was trained with the following prompt style:
### HUMAN:
Hello
### RESPONSE:
Hi, how are you?
### HUMAN:
I'm fine.
### RESPONSE:
How can I help you?
...
Training code
Code used to train the model is available here.
To reproduce the results:
git clone https://github.com/georgesung/llm_qlora
cd llm_qlora
pip install -r requirements.txt
python train.py configs/llama2_7b_chat_uncensored.yaml
Fine-tuning guide
https://georgesung.github.io/ai/qlora-ift/
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 43.39 |
| ARC (25-shot) | 53.58 |
| HellaSwag (10-shot) | 78.66 |
| MMLU (5-shot) | 44.49 |
| TruthfulQA (0-shot) | 41.34 |
| Winogrande (5-shot) | 74.11 |
| GSM8K (5-shot) | 5.84 |
| DROP (3-shot) | 5.69 |