--- language: - en license: apache-2.0 library_name: transformers tags: - mergekit - merge base_model: - arcee-ai/Virtuoso-Small - CultriX/SeQwence-14B-EvolMerge - CultriX/Qwen2.5-14B-Wernicke - sthenno-com/miscii-14b-1028 - underwoods/medius-erebus-magnum-14b - sometimesanotion/lamarck-14b-prose-model_stock - sometimesanotion/lamarck-14b-reason-model_stock metrics: - accuracy pipeline_tag: text-generation model-index: - name: Lamarck-14B-v0.3 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: 50.32 name: strict accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 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: 51.27 name: normalized accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 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: 32.4 name: exact match source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 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: 18.46 name: acc_norm source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 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: 18.0 name: acc_norm source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 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: 49.01 name: accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=sometimesanotion/Lamarck-14B-v0.3 name: Open LLM Leaderboard --- ![Lamarck.webp](https://huggingface.co/sometimesanotion/Lamarck-14B-v0.3/resolve/main/Lamarck.webp) --- # merge Lamarck-14B is a carefully designed merge which emphasizes [arcee-ai/Virtuoso-Small](https://huggingface.co/arcee-ai/Virtuoso-Small) in early and finishing layers, and midway features strong influence on reasoning and prose from [CultriX/SeQwence-14B-EvolMerge](http://huggingface.co/CultriX/SeQwence-14B-EvolMerge) especially, but a number of other models as well through its model_stock. Version 0.3 is the product of a carefully planned and tested sequence of templated merges, produced by a toolchain which wraps around Arcee's mergekit. For GGUFs, [mradermacher/Lamarck-14B-v0.3-i1-GGUF](https://huggingface.co/mradermacher/Lamarck-14B-v0.3-i1-GGUF) has you covered. Thank you @mradermacher! **The merge strategy of Lamarck 0.3 can be summarized as:** - Two model_stocks commence specialized branches for reasoning and prose quality. - For refinement on both model_stocks, DELLA merges re-emphasize selected ancestors. - For smooth instruction following, a SLERP merges Virtuoso with a DELLA merge of the two branches, where reason vs. prose quality are balanced. - For finalization and normalization, a TIES merge. ![graph.png](https://huggingface.co/sometimesanotion/Lamarck-14B-v0.3-experimental/resolve/main/graph.png) The first two layers come entirely from Virtuoso. The choice to leave these layers untouched comes from [arxiv.org/abs/2307.03172](https://arxiv.org/abs/2307.03172) which identifies early attention glitches as a chief cause of hallucinations. Layers 3-8 feature a SLERP gradient into introducing the DELLA merge tree in which the reason branch is emphasized, the prose branch only given a small ranking. ### Thanks go to: - @arcee-ai's team for the ever-capable mergekit, and the exceptional Virtuoso Small model. - @CultriX for the helpful examples of memory-efficient sliced merges and evolutionary merging. Their contribution of tinyevals on version 0.1 of Lamarck did much to validate the hypotheses of the DELLA->SLERP gradient process used here. - The authors behind the capable models that appear in the model_stock. ### Models Merged **Top influences:** These ancestors are base models and present in the model_stocks, but are heavily re-emphasized in the DELLA and SLERP merges. - **[arcee-ai/Virtuoso-Small](https://huggingface.co/arcee-ai/Virtuoso-Small)** - A brand new model from Arcee, refined from the notable cross-architecture Llama-to-Qwen distillation [arcee-ai/SuperNova-Medius](https://huggingface.co/arcee-ai/SuperNova-Medius). The first two layers are nearly exclusively from Virtuoso. It has proven to be a well-rounded performer, and contributes a noticeable boost to the model's prose quality. - **[CultriX/SeQwence-14B-EvolMerge](http://huggingface.co/CultriX/SeQwence-14B-EvolMerge)** - A top contender on reasoning benchmarks. **Reason:** While Virtuoso is the strongest influence the starting ending layers, the reasoning mo - **[CultriX/Qwen2.5-14B-Wernicke](http://huggingface.co/CultriX/Qwen2.5-14B-Wernicke)** - A top performer for Arc and GPQA, Wernicke is re-emphasized in small but highly-ranked portions of the model. - **[VAGOsolutions/SauerkrautLM-v2-14b-DPO](https://huggingface.co/VAGOsolutions/SauerkrautLM-v2-14b-DPO)** - This model's influence is understated, but aids BBH and coding capability. **Prose:** While the prose module is gently applied, its impact is noticeable on Lamarck 0.3's prose quality, and a DELLA merge re-emphasizes the contributions of two models particularly: - **[sthenno-com/miscii-14b-1028](https://huggingface.co/sthenno-com/miscii-14b-1028)** - **[underwoods/medius-erebus-magnum-14b](https://huggingface.co/underwoods/medius-erebus-magnum-14b)** **Model stock:** Two model_stock merges, specialized for specific aspects of performance, are used to mildly influence a large range of the model. - **[sometimesanotion/lamarck-14b-reason-model_stock](https://huggingface.co/sometimesanotion/lamarck-14b-reason-model_stock)** - **[sometimesanotion/lamarck-14b-prose-model_stock](https://huggingface.co/sometimesanotion/lamarck-14b-prose-model_stock)** - This brings in a little influence from [EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2](https://huggingface.co/EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2), [oxyapi/oxy-1-small](https://huggingface.co/oxyapi/oxy-1-small), and [allura-org/TQ2.5-14B-Sugarquill-v1](https://huggingface.co/allura-org/TQ2.5-14B-Sugarquill-v1). **Note on abliteration:** This author believes that adjacent services and not language models themselves are where guardrails are best placed. Effort to de-censor Lamarck will resume after the model has been further studied. ### Configuration The following YAML configuration was used to produce this model: ```yaml name: lamarck-14b-reason-della # This contributes the knowledge and reasoning pool, later to be merged merge_method: della # with the dominant instruction-following model base_model: arcee-ai/Virtuoso-Small tokenizer_source: arcee-ai/Virtuoso-Small parameters: int8_mask: false normalize: true rescale: false density: 0.30 weight: 0.50 epsilon: 0.08 lambda: 1.00 models: - model: CultriX/SeQwence-14B-EvolMerge parameters: density: 0.70 weight: 0.90 - model: sometimesanotion/lamarck-14b-reason-model_stock parameters: density: 0.90 weight: 0.60 - model: CultriX/Qwen2.5-14B-Wernicke parameters: density: 0.20 weight: 0.30 dtype: bfloat16 out_dtype: bfloat16 --- name: lamarck-14b-prose-della # This contributes the prose, later to be merged merge_method: della # with the dominant instruction-following model base_model: arcee-ai/Virtuoso-Small tokenizer_source: arcee-ai/Virtuoso-Small parameters: int8_mask: false normalize: true rescale: false density: 0.30 weight: 0.50 epsilon: 0.08 lambda: 0.95 models: - model: sthenno-com/miscii-14b-1028 parameters: density: 0.40 weight: 0.90 - model: sometimesanotion/lamarck-14b-prose-model_stock parameters: density: 0.60 weight: 0.70 - model: underwoods/medius-erebus-magnum-14b dtype: bfloat16 out_dtype: bfloat16 --- name: lamarck-14b-converge-della # This is the strongest control point to quickly merge_method: della # re-balance reasoning vs. prose base_model: arcee-ai/Virtuoso-Small tokenizer_source: arcee-ai/Virtuoso-Small parameters: int8_mask: false normalize: true rescale: false density: 0.30 weight: 0.50 epsilon: 0.08 lambda: 1.00 models: - model: sometimesanotion/lamarck-14b-reason-della parameters: density: 0.80 weight: 1.00 - model: arcee-ai/Virtuoso-Small parameters: density: 0.40 weight: 0.50 - model: sometimesanotion/lamarck-14b-prose-della parameters: density: 0.10 weight: 0.40 dtype: bfloat16 out_dtype: bfloat16 --- name: lamarck-14b-converge # Virtuoso has good capabilities all-around; it is 100% of the first merge_method: slerp # two layers, and blends into the reasoning+prose convergance base_model: arcee-ai/Virtuoso-Small # for some interesting boosts tokenizer_source: base parameters: t: [ 0.00, 0.60, 0.80, 0.80, 0.80, 0.70, 0.40 ] slices: - sources: - layer_range: [ 0, 2 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 0, 2 ] model: merges/lamarck-14b-converge-della t: [ 0.00, 0.00 ] - sources: - layer_range: [ 2, 8 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 2, 8 ] model: merges/lamarck-14b-converge-della t: [ 0.00, 0.60 ] - sources: - layer_range: [ 8, 16 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 8, 16 ] model: merges/lamarck-14b-converge-della t: [ 0.60, 0.70 ] - sources: - layer_range: [ 16, 24 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 16, 24 ] model: merges/lamarck-14b-converge-della t: [ 0.70, 0.70 ] - sources: - layer_range: [ 24, 32 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 24, 32 ] model: merges/lamarck-14b-converge-della t: [ 0.70, 0.70 ] - sources: - layer_range: [ 32, 40 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 32, 40 ] model: merges/lamarck-14b-converge-della t: [ 0.70, 0.60 ] - sources: - layer_range: [ 40, 48 ] model: arcee-ai/Virtuoso-Small - layer_range: [ 40, 48 ] model: merges/lamarck-14b-converge-della t: [ 0.60, 0.40 ] dtype: bfloat16 out_dtype: bfloat16 --- name: lamarck-14b-finalize merge_method: ties base_model: Qwen/Qwen2.5-14B tokenizer_source: Qwen/Qwen2.5-14B-Instruct parameters: int8_mask: false normalize: true rescale: false density: 1.00 weight: 1.00 models: - model: merges/lamarck-14b-converge dtype: bfloat16 out_dtype: bfloat16 --- ``` # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_sometimesanotion__Lamarck-14B-v0.3) | Metric |Value| |-------------------|----:| |Avg. |36.58| |IFEval (0-Shot) |50.32| |BBH (3-Shot) |51.27| |MATH Lvl 5 (4-Shot)|32.40| |GPQA (0-shot) |18.46| |MuSR (0-shot) |18.00| |MMLU-PRO (5-shot) |49.01|