---
language:
- ru
license: apache-2.0
datasets:
- 0x7194633/fialka-v3-data
pipeline_tag: text-generation
model-index:
- name: fialka-13B-v4
  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: 29.69
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      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: 47.37
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      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: 25.09
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      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: 43.65
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      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: 58.88
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      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: 0.38
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=0x7194633/fialka-13B-v4
      name: Open LLM Leaderboard
---
# Fialka v4.0 13B

![Violet](https://i.imgur.com/tQKuehR.png)
## Description
Fialka language models are trained to follow instructions and maintain communication in Russian. The fourth version of Fialka is the third version optimized through RLHF. More responsive and more informative.
## Usage
The model has a query format as in zephyr.
```
<|user|>
Что такое мем?</s>
<|assistant|>
Мем (англ. meme) - это единица социального поведения, которая быстро распространяется в интернете или в социальных сетях с целью передачи информации и идей. Обычно мемы являются шутками, стишками, изображениями или видео и имеют юмористический или сатирический характер, но могут содержать и более серьезные идеи, такие как политические или социальные протесты, и даже угрозы. Мемы могут служить для создания и распространения контента и информации, а также для выражения мнения или чувств автора.
```
# [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_0x7194633__fialka-13B-v4)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |34.18|
|AI2 Reasoning Challenge (25-Shot)|29.69|
|HellaSwag (10-Shot)              |47.37|
|MMLU (5-Shot)                    |25.09|
|TruthfulQA (0-shot)              |43.65|
|Winogrande (5-shot)              |58.88|
|GSM8k (5-shot)                   | 0.38|