README.md

metadata

language:
  - en
  - fr
  - de
  - es
  - it
  - pt
  - ru
  - zh
  - ja
license: apache-2.0
tags:
  - merge
datasets:
  - Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
  - anthracite-org/stheno-filtered-v1.1
  - PJMixers/hieunguyenminh_roleplay-deduped-ShareGPT
  - Gryphe/Sonnet3.5-Charcard-Roleplay
  - Epiculous/Synthstruct-Gens-v1.1-Filtered-n-Cleaned
  - anthracite-org/kalo-opus-instruct-22k-no-refusal
  - anthracite-org/nopm_claude_writing_fixed
  - anthracite-org/kalo_opus_misc_240827
pipeline_tag: text-generation
model-index:
  - name: Violet_Twilight-v0.2
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 45.32
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 23.94
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 2.72
            name: exact match
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 2.13
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 13.61
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 23.45
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=Epiculous/Violet_Twilight-v0.2
          name: Open LLM Leaderboard

QuantFactory/Violet_Twilight-v0.2-GGUF

This is quantized version of Epiculous/Violet_Twilight-v0.2 created using llama.cpp

Original Model Card

Now for something a bit different, Violet_Twilight-v0.2! This model is a SLERP merge of Azure_Dusk-v0.2 and Crimson_Dawn-v0.2!

Quants!

full / exl2 / gguf

Prompting

The v0.2 models are trained on ChatML, the prompting structure goes a little something like this:

<|im_start|>user
Hi there!<|im_end|>
<|im_start|>assistant
Nice to meet you!<|im_end|>
<|im_start|>user
Can I ask a question?<|im_end|>
<|im_start|>assistant

Context and Instruct

The v0.2 models are trained on ChatML, please use that Context and Instruct template.

Current Top Sampler Settings

Spicy_Temp
Violet_Twilight-Nitral-Special

Merging

The following config was used to merge Azure Dusk and Crimson Dawn

slices:
  - sources:
      - model: Epiculous/Azure_Dusk-v0.2
        layer_range: [0, 40]
      - model: Epiculous/Crimson_Dawn-V0.2
        layer_range: [0, 40]
merge_method: slerp
base_model: Epiculous/Azure_Dusk-v0.2
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 # fallback for rest of tensors
dtype: bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	18.53
IFEval (0-Shot)	45.32
BBH (3-Shot)	23.94
MATH Lvl 5 (4-Shot)	2.72
GPQA (0-shot)	2.13
MuSR (0-shot)	13.61
MMLU-PRO (5-shot)	23.45

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	18.53
IFEval (0-Shot)	45.32
BBH (3-Shot)	23.94
MATH Lvl 5 (4-Shot)	2.72
GPQA (0-shot)	2.13
MuSR (0-shot)	13.61
MMLU-PRO (5-shot)	23.45