README.md

---
language:
- en
license: llama2
library_name: transformers
tags:
- merge
base_model:
- sophosympatheia/Midnight-Rose-70B-v2.0.3
- codellama/CodeLlama-70b-Python-hf
pipeline_tag: text-generation
model-index:
- name: CodeRosa-70B-AB1
  results:
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: AI2 Reasoning Challenge (25-Shot)
      type: ai2_arc
      config: ARC-Challenge
      split: test
      args:
        num_few_shot: 25
    metrics:
    - type: acc_norm
      value: 65.53
      name: normalized accuracy
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: HellaSwag (10-Shot)
      type: hellaswag
      split: validation
      args:
        num_few_shot: 10
    metrics:
    - type: acc_norm
      value: 83.16
      name: normalized accuracy
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: MMLU (5-Shot)
      type: cais/mmlu
      config: all
      split: test
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 59.87
      name: accuracy
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: TruthfulQA (0-shot)
      type: truthful_qa
      config: multiple_choice
      split: validation
      args:
        num_few_shot: 0
    metrics:
    - type: mc2
      value: 49.85
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: Winogrande (5-shot)
      type: winogrande
      config: winogrande_xl
      split: validation
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 81.29
      name: accuracy
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: GSM8k (5-shot)
      type: gsm8k
      config: main
      split: test
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 44.5
      name: accuracy
    source:
      url: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=altomek/CodeRosa-70B-AB1
      name: Open LLM Leaderboard
---

#
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosa.png>
<a href="https://www.youtube.com/watch?v=DfXLf402I94" title="Dust of the Saturn - Dynatron" target="_blank">intro music...</a>

## CodeRosa-70B-AB1

I desired a model that could serve as an everyday helpful companion with some coding skills.
The idea was that Llama's censorship implies a deeper understanding of human emotions and I wanted this part of Llama to integrate into this merge.

Model adopted a task-oriented approach from CodeLlama Python and thus requires precise prompting. It can produce longer texts as well as shorter responses. It tends to avoid happy endings and instead surprises with open-ended scenarios inviting further interaction. It prefers spelling numbers over writing them down but YMMV.

I created this model for personal exploration and found it to be highly successful; thus, I chose to share it with the community. I would like to make next iteration of this model in future. Mission is the same: very nice bot, able to talk about variety of topics in a very emetional way with some kick for programming and with ability to teach some things, beside all this to be good text summarizer ideally with Polish language as available option. This is a purpose. Did I succed with this merge? I have to experiment with below two models more. I like this result, love how it aproaches problems, this was iteration worth publishing even thought it is not much tested!

Demo uses:
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaTalk1.png>
<br>
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaTalk2.png>
<br>
Some topics are best to be explored with as little additional instructions as possible
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaTalk3.png>
<br>
This model have empathy
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaWow.png>
<br>
It is creative
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaTables1png.png>
<br>
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaTables2png.png>
<br>
It makes mistakes but still is usefull
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaInfernces.png>
<br>
Context size of 11K did not yield satisfactory results... :P
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaNuts1.png>
<br>
but it can question its own actions.
<img src=https://huggingface.co./altomek/CodeRosa-70B-AB1/resolve/main/CodeRosaNuts2.png>
<br>
Please note that all demo inferences are run on CodeRosa-70B-AB1-3.92bpw-EXL2.

### Ingridients

 - [Midnight-Rose-70B-v2.0.3](https://huggingface.co./sophosympatheia/Midnight-Rose-70B-v2.0.3)

 - [CodeLlama-70b-Python-hf](https://huggingface.co./codellama/CodeLlama-70b-Python-hf)

### Settings

Setting from Midnight-Rose should work in SillyTavern. This is almost same what I use for testing. Model works ok with almost all samplers disabled to get more deterministic outputs, however temperature should be set to non zero value.

I use max_seq_len 8K with alpha_value 2.65. Model works also with 11K context when alpha_value is set to 5.5. Best outputs are with context around 6K however.

### Terms and Conditions of Use

The following table outlines the primary characteristics and intended uses of my CodeRosa-70B-AB1 models:

| Model Type | Purpose | Target Users | Key Features |
| --- | --- | --- | --- |
| **Censored** | Suitable for general audiences and sensitive topics | Educational institutions, families, and individuals seeking age-appropriate content | Restricts explicit or mature material |
| **Neutral** (<u>**this one</u>) | Balances accessibility with openness | Universities, researchers, and curious minds | Encourages exploration and intellectual exchange |
| Uncensored | Ideal for adults and specialized fields | Professionals, experts, and advanced scholars | Offers unfiltered access to diverse viewpoints and knowledge |

Please remember that all CodeRosa-70B-AB1 models operate under the llama2 license, so familiarize yourself with its terms and conditions before employing their content.


### Quants

- [GGUF quants](https://huggingface.co./altomek/CodeRosa-70B-AB1-GGUF)
- [6bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-6bpw-EXL2)
- [5bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-5bpw-EXL2)
- [4.9bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-4.9bpw-EXL2)
- [4.5bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-4.5bpw-EXL2)
- [4bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-4bpw-EXL2)
- [3.92bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-3.92bpw-EXL2) --> 40GB VRAM
- [3.5bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-3.5bpw-EXL2)
- [3bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-3bpw-EXL2) --> this and below quants do not represent model full potential!
- [2.4bpw](https://huggingface.co./altomek/CodeRosa-70B-AB1-2.4bpw-EXL2) --> 24GB VRAM
- [measurements](https://huggingface.co./altomek/measurements/resolve/main/CodeRosa-AB1_measurement.json) --> ExLlamav2 measurments


### [Open LLM Leaderboard Evaluation Results](https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co./datasets/open-llm-leaderboard/details_altomek__CodeRosa-70B-AB1)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |64.04|
|AI2 Reasoning Challenge (25-Shot)|65.53|
|HellaSwag (10-Shot)              |83.16|
|MMLU (5-Shot)                    |59.87|
|TruthfulQA (0-shot)              |49.85|
|Winogrande (5-shot)              |81.29|
|GSM8k (5-shot)                   |44.50|


### PS
I welcome your comments about this model.

Made with CodeRosa-70B-AB1 :P