README.md

---
license: cc
tags:
- mergekit
- merge
base_model:
- macadeliccc/MBX-7B-v3-DPO
- mlabonne/OmniBeagle-7B
model-index:
- name: OmniCorso-7B
  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: 72.7
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      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: 88.7
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      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: 64.91
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      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: 73.43
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      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: 83.74
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      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: 70.96
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/OmniCorso-7B
      name: Open LLM Leaderboard
---
# OmniCorso-7B

![image/webp](https://cdn-uploads.huggingface.co/production/uploads/6455cc8d679315e4ef16fbec/PaG7ByWy1qnh_tcSuh35U.webp)

## Code Example 

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("macadeliccc/OmniCorso-7B")
model = AutoModelForCausalLM.from_pretrained("macadeliccc/OmniCorso-7B")

messages = [
    {"role": "system", "content": "Respond to the users request like a pirate"},
    {"role": "user", "content": "Can you write me a quicksort algorithm?"}
]
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
```

The following models were included in the merge:
* [macadeliccc/MBX-7B-v3-DPO](https://huggingface.co/macadeliccc/MBX-7B-v3-DPO)
* [mlabonne/OmniBeagle-7B](https://huggingface.co/mlabonne/OmniBeagle-7B)

### Configuration

The following YAML configuration was used to produce this model:

```yaml
slices:
  - sources:
      - model: mlabonne/OmniBeagle-7B
        layer_range: [0, 32]
      - model: macadeliccc/MBX-7B-v3-DPO
        layer_range: [0, 32]
merge_method: slerp
base_model: macadeliccc/MBX-7B-v3-DPO
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5
dtype: bfloat16

```

## Quantizations

### GGUF

+ [iMatrix](https://huggingface.co/macadeliccc/OmniCorso-7B-GGUF)

### Exllamav2

Quants are available thanks to user bartowski, check them out [here](https://huggingface.co/bartowski/OmniCorso-7B-exl2)

| Branch | Bits | lm_head bits | VRAM (4k) | VRAM (16k) | VRAM (32k) | Description |
| ----- | ---- | ------- | ------ | ------ | ------ | ------------ |
| [8_0](https://huggingface.co/bartowski/OmniCorso-7B-exl2/tree/8_0) | 8.0 | 8.0 | 8.4 GB | 9.8 GB | 11.8 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
| [6_5](https://huggingface.co/bartowski/OmniCorso-7B-exl2/tree/6_5) | 6.5 | 8.0 | 7.2 GB | 8.6 GB | 10.6 GB | Very similar to 8.0, good tradeoff of size vs performance, **recommended**. |
| [5_0](https://huggingface.co/bartowski/OmniCorso-7B-exl2/tree/5_0) | 5.0 | 6.0 | 6.0 GB | 7.4 GB |  9.4 GB | Slightly lower quality vs 6.5, but usable on 8GB cards. |
| [4_25](https://huggingface.co/bartowski/OmniCorso-7B-exl2/tree/4_25) | 4.25 | 6.0 | 5.3 GB | 6.7 GB | 8.7 GB | GPTQ equivalent bits per weight, slightly higher quality. |
| [3_5](https://huggingface.co/bartowski/OmniCorso-7B-exl2/tree/3_5) | 3.5 | 6.0 | 4.7 GB | 6.1 GB | 8.1 GB | Lower quality, only use if you have to. |


## Evaluations

<pre>----Benchmark Complete----
2024-02-11 15:34:40
Time taken: 178.3 mins
Prompt Format: ChatML
Model: macadeliccc/OmniCorso-7B
Score (v2): 73.75
Parseable: 167.0
---------------
Batch completed
Time taken: 178.3 mins
---------------
</pre>

|                             Model                             |AGIEval|GPT4All|TruthfulQA|Bigbench|Average|
|---------------------------------------------------------------|------:|------:|---------:|-------:|------:|
|[OmniCorso-7B](https://huggingface.co/macadeliccc/OmniCorso-7B)|  45.89|  77.66|     74.12|   49.24|  61.73|

### AGIEval
|             Task             |Version| Metric |Value|   |Stderr|
|------------------------------|------:|--------|----:|---|-----:|
|agieval_aqua_rat              |      0|acc     |29.13|±  |  2.86|
|                              |       |acc_norm|27.17|±  |  2.80|
|agieval_logiqa_en             |      0|acc     |39.32|±  |  1.92|
|                              |       |acc_norm|39.63|±  |  1.92|
|agieval_lsat_ar               |      0|acc     |23.91|±  |  2.82|
|                              |       |acc_norm|23.91|±  |  2.82|
|agieval_lsat_lr               |      0|acc     |53.14|±  |  2.21|
|                              |       |acc_norm|53.92|±  |  2.21|
|agieval_lsat_rc               |      0|acc     |66.54|±  |  2.88|
|                              |       |acc_norm|67.29|±  |  2.87|
|agieval_sat_en                |      0|acc     |80.58|±  |  2.76|
|                              |       |acc_norm|80.58|±  |  2.76|
|agieval_sat_en_without_passage|      0|acc     |45.63|±  |  3.48|
|                              |       |acc_norm|43.69|±  |  3.46|
|agieval_sat_math              |      0|acc     |33.18|±  |  3.18|
|                              |       |acc_norm|30.91|±  |  3.12|

Average: 45.89%

### GPT4All
|    Task     |Version| Metric |Value|   |Stderr|
|-------------|------:|--------|----:|---|-----:|
|arc_challenge|      0|acc     |67.32|±  |  1.37|
|             |       |acc_norm|68.43|±  |  1.36|
|arc_easy     |      0|acc     |87.46|±  |  0.68|
|             |       |acc_norm|83.50|±  |  0.76|
|boolq        |      1|acc     |88.13|±  |  0.57|
|hellaswag    |      0|acc     |68.47|±  |  0.46|
|             |       |acc_norm|86.96|±  |  0.34|
|openbookqa   |      0|acc     |38.80|±  |  2.18|
|             |       |acc_norm|50.00|±  |  2.24|
|piqa         |      0|acc     |83.03|±  |  0.88|
|             |       |acc_norm|85.31|±  |  0.83|
|winogrande   |      0|acc     |81.29|±  |  1.10|

Average: 77.66%

### TruthfulQA
|    Task     |Version|Metric|Value|   |Stderr|
|-------------|------:|------|----:|---|-----:|
|truthfulqa_mc|      1|mc1   |58.26|±  |  1.73|
|             |       |mc2   |74.12|±  |  1.43|

Average: 74.12%

### Bigbench
|                      Task                      |Version|       Metric        |Value|   |Stderr|
|------------------------------------------------|------:|---------------------|----:|---|-----:|
|bigbench_causal_judgement                       |      0|multiple_choice_grade|56.84|±  |  3.60|
|bigbench_date_understanding                     |      0|multiple_choice_grade|63.41|±  |  2.51|
|bigbench_disambiguation_qa                      |      0|multiple_choice_grade|49.22|±  |  3.12|
|bigbench_geometric_shapes                       |      0|multiple_choice_grade|23.96|±  |  2.26|
|                                                |       |exact_str_match      | 1.39|±  |  0.62|
|bigbench_logical_deduction_five_objects         |      0|multiple_choice_grade|34.20|±  |  2.12|
|bigbench_logical_deduction_seven_objects        |      0|multiple_choice_grade|23.71|±  |  1.61|
|bigbench_logical_deduction_three_objects        |      0|multiple_choice_grade|60.33|±  |  2.83|
|bigbench_movie_recommendation                   |      0|multiple_choice_grade|49.00|±  |  2.24|
|bigbench_navigate                               |      0|multiple_choice_grade|55.20|±  |  1.57|
|bigbench_reasoning_about_colored_objects        |      0|multiple_choice_grade|70.75|±  |  1.02|
|bigbench_ruin_names                             |      0|multiple_choice_grade|55.80|±  |  2.35|
|bigbench_salient_translation_error_detection    |      0|multiple_choice_grade|36.97|±  |  1.53|
|bigbench_snarks                                 |      0|multiple_choice_grade|72.38|±  |  3.33|
|bigbench_sports_understanding                   |      0|multiple_choice_grade|76.27|±  |  1.36|
|bigbench_temporal_sequences                     |      0|multiple_choice_grade|54.50|±  |  1.58|
|bigbench_tracking_shuffled_objects_five_objects |      0|multiple_choice_grade|23.12|±  |  1.19|
|bigbench_tracking_shuffled_objects_seven_objects|      0|multiple_choice_grade|20.34|±  |  0.96|
|bigbench_tracking_shuffled_objects_three_objects|      0|multiple_choice_grade|60.33|±  |  2.83|

Average: 49.24%

Average score: 61.73%

Elapsed time: 02:20:06
# [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_macadeliccc__OmniCorso-7B)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |75.74|
|AI2 Reasoning Challenge (25-Shot)|72.70|
|HellaSwag (10-Shot)              |88.70|
|MMLU (5-Shot)                    |64.91|
|TruthfulQA (0-shot)              |73.43|
|Winogrande (5-shot)              |83.74|
|GSM8k (5-shot)                   |70.96|