[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73AHHYA7JKNB9PW67TY9GMN":3,"skill-same:01M73AHHYA7JKNB9PW67TY9GMN":78},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":42,"commitSha":43,"skillDirectory":44,"forksCount":45,"repositoryHtmlUrl":46,"repositoryHomepageUrl":6,"ownerHtmlUrl":47,"ownerBlogUrl":48,"ownerDescription":6,"updatedAt":49,"install":50,"resources":63},"01M73AHHYA7JKNB9PW67TY9GMN","model-merging",null,"Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task Arithmetic, linear merging, and production deployment strategies.","qhjqhj00\u002Fresearch-skills-pool","qhjqhj00","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F29684092?v=4","01M72AE0ZGXVM5C57N512DGQG8","Tommy Chien","01M72AE107ESJMHT7QC5SXX144",3,"MIT","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Fskills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: model-merging\ndescription: Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task Arithmetic, linear merging, and production deployment strategies.\nversion: 1.0.0\nauthor: Orchestra Research\nlicense: MIT\ntags: [Emerging Techniques, Model Merging, Mergekit, SLERP, TIES, DARE, Task Arithmetic, Model Fusion, No Retraining, Multi-Capability, Arcee AI]\ndependencies: [mergekit, transformers, torch]\n---\n\n# Model Merging: Combining Pre-trained Models\n\n## When to Use This Skill\n\nUse Model Merging when you need to:\n- **Combine capabilities** from multiple fine-tuned models without retraining\n- **Create specialized models** by blending domain-specific expertise (math + coding + chat)\n- **Improve performance** beyond single models (often +5-10% on benchmarks)\n- **Reduce training costs** - no GPUs needed, merges run on CPU\n- **Experiment rapidly** - create new model variants in minutes, not days\n- **Preserve multiple skills** - merge without catastrophic forgetting\n\n**Success Stories**: Marcoro14-7B-slerp (best on Open LLM Leaderboard 02\u002F2024), many top HuggingFace models use merging\n\n**Tools**: mergekit (Arcee AI), LazyMergekit, Model Soup\n\n## Installation\n\n```bash\n# Install mergekit\ngit clone https:\u002F\u002Fgithub.com\u002Farcee-ai\u002Fmergekit.git\ncd mergekit\npip install -e .\n\n# Or via pip\npip install mergekit\n\n# Optional: Transformer library\npip install transformers torch\n```\n\n## Quick Start\n\n### Simple Linear Merge\n\n```yaml\n# config.yml - Merge two models with equal weights\nmerge_method: linear\nmodels:\n  - model: mistralai\u002FMistral-7B-v0.1\n    parameters:\n      weight: 0.5\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      weight: 0.5\ndtype: bfloat16\n```\n\n```bash\n# Run merge\nmergekit-yaml config.yml .\u002Fmerged-model --cuda\n\n# Use merged model\npython -m transformers.models.auto --model_name_or_path .\u002Fmerged-model\n```\n\n### SLERP Merge (Best for 2 Models)\n\n```yaml\n# config.yml - Spherical interpolation\nmerge_method: slerp\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 32]\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [0, 32]\nparameters:\n  t: 0.5  # Interpolation factor (0=model1, 1=model2)\ndtype: bfloat16\n```\n\n## Core Concepts\n\n### 1. Merge Methods\n\n**Linear (Model Soup)**\n- Simple weighted average of parameters\n- Fast, works well for similar models\n- Can merge 2+ models\n\n```python\nmerged_weights = w1 * model1_weights + w2 * model2_weights + w3 * model3_weights\n# where w1 + w2 + w3 = 1\n```\n\n**SLERP (Spherical Linear Interpolation)**\n- Interpolates along sphere in weight space\n- Preserves magnitude of weight vectors\n- Best for merging 2 models\n- Smoother than linear\n\n```python\n# SLERP formula\nmerged = (sin((1-t)*θ) \u002F sin(θ)) * model1 + (sin(t*θ) \u002F sin(θ)) * model2\n# where θ = arccos(dot(model1, model2))\n# t ∈ [0, 1]\n```\n\n**Task Arithmetic**\n- Extract \"task vectors\" (fine-tuned - base)\n- Combine task vectors, add to base\n- Good for merging multiple specialized models\n\n```python\n# Task vector\ntask_vector = finetuned_model - base_model\n\n# Merge multiple task vectors\nmerged = base_model + α₁*task_vector₁ + α₂*task_vector₂\n```\n\n**TIES-Merging**\n- Task arithmetic + sparsification\n- Resolves sign conflicts in parameters\n- Best for merging many task-specific models\n\n**DARE (Drop And REscale)**\n- Randomly drops fine-tuned parameters\n- Rescales remaining parameters\n- Reduces redundancy, maintains performance\n\n### 2. Configuration Structure\n\n```yaml\n# Basic structure\nmerge_method: \u003Cmethod>  # linear, slerp, ties, dare_ties, task_arithmetic\nbase_model: \u003Cpath>      # Optional: base model for task arithmetic\n\nmodels:\n  - model: \u003Cpath\u002Fto\u002Fmodel1>\n    parameters:\n      weight: \u003Cfloat>   # Merge weight\n      density: \u003Cfloat>  # For TIES\u002FDARE\n\n  - model: \u003Cpath\u002Fto\u002Fmodel2>\n    parameters:\n      weight: \u003Cfloat>\n\nparameters:\n  # Method-specific parameters\n\ndtype: \u003Cdtype>  # bfloat16, float16, float32\n\n# Optional\nslices:  # Layer-wise merging\ntokenizer:  # Tokenizer configuration\n```\n\n## Merge Methods Guide\n\n### Linear Merge\n\n**Best for**: Simple model combinations, equal weighting\n\n```yaml\nmerge_method: linear\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      weight: 0.4\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      weight: 0.3\n  - model: NousResearch\u002FNous-Hermes-2-Mistral-7B-DPO\n    parameters:\n      weight: 0.3\ndtype: bfloat16\n```\n\n### SLERP Merge\n\n**Best for**: Two models, smooth interpolation\n\n```yaml\nmerge_method: slerp\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 32]\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [0, 32]\nparameters:\n  t: 0.5  # 0.0 = first model, 1.0 = second model\ndtype: bfloat16\n```\n\n**Layer-specific SLERP:**\n\n```yaml\nmerge_method: slerp\nslices:\n  - sources:\n      - model: model_a\n        layer_range: [0, 32]\n      - model: model_b\n        layer_range: [0, 32]\nparameters:\n  t:\n    - filter: self_attn    # Attention layers\n      value: 0.3\n    - filter: mlp          # MLP layers\n      value: 0.7\n    - value: 0.5           # Default for other layers\ndtype: bfloat16\n```\n\n### Task Arithmetic\n\n**Best for**: Combining specialized skills\n\n```yaml\nmerge_method: task_arithmetic\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1  # Math\n    parameters:\n      weight: 0.5\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B  # Chat\n    parameters:\n      weight: 0.3\n  - model: ajibawa-2023\u002FCode-Mistral-7B  # Code\n    parameters:\n      weight: 0.2\ndtype: bfloat16\n```\n\n### TIES-Merging\n\n**Best for**: Many models, resolving conflicts\n\n```yaml\nmerge_method: ties\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      density: 0.5  # Keep top 50% of parameters\n      weight: 1.0\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      density: 0.5\n      weight: 1.0\n  - model: NousResearch\u002FNous-Hermes-2-Mistral-7B-DPO\n    parameters:\n      density: 0.5\n      weight: 1.0\nparameters:\n  normalize: true\ndtype: bfloat16\n```\n\n### DARE Merge\n\n**Best for**: Reducing redundancy\n\n```yaml\nmerge_method: dare_ties\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      density: 0.5    # Drop 50% of deltas\n      weight: 0.6\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      density: 0.5\n      weight: 0.4\nparameters:\n  int8_mask: true  # Use int8 for masks (saves memory)\ndtype: bfloat16\n```\n\n## Advanced Patterns\n\n### Layer-wise Merging\n\n```yaml\n# Different models for different layers\nmerge_method: passthrough\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 16]   # First half\n  - sources:\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [16, 32]  # Second half\ndtype: bfloat16\n```\n\n### MoE from Merged Models\n\n```yaml\n# Create Mixture of Experts\nmerge_method: moe\nbase_model: mistralai\u002FMistral-7B-v0.1\nexperts:\n  - source_model: WizardLM\u002FWizardMath-7B-V1.1\n    positive_prompts:\n      - \"math\"\n      - \"calculate\"\n  - source_model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    positive_prompts:\n      - \"chat\"\n      - \"conversation\"\n  - source_model: ajibawa-2023\u002FCode-Mistral-7B\n    positive_prompts:\n      - \"code\"\n      - \"python\"\ndtype: bfloat16\n```\n\n### Tokenizer Merging\n\n```yaml\nmerge_method: linear\nmodels:\n  - model: mistralai\u002FMistral-7B-v0.1\n  - model: custom\u002Fspecialized-model\n\ntokenizer:\n  source: \"union\"  # Combine vocabularies from both models\n  tokens:\n    \u003C|special_token|>:\n      source: \"custom\u002Fspecialized-model\"\n```\n\n## Best Practices\n\n### 1. Model Compatibility\n\n```python\n# ✅ Good: Same architecture\nmodels = [\n    \"mistralai\u002FMistral-7B-v0.1\",\n    \"teknium\u002FOpenHermes-2.5-Mistral-7B\",  # Both Mistral 7B\n]\n\n# ❌ Bad: Different architectures\nmodels = [\n    \"meta-llama\u002FLlama-2-7b-hf\",  # Llama\n    \"mistralai\u002FMistral-7B-v0.1\",  # Mistral (incompatible!)\n]\n```\n\n### 2. Weight Selection\n\n```yaml\n# ✅ Good: Weights sum to 1.0\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.6\n  - model: model_b\n    parameters:\n      weight: 0.4  # 0.6 + 0.4 = 1.0\n\n# ⚠️  Acceptable: Weights don't sum to 1 (for task arithmetic)\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.8\n  - model: model_b\n    parameters:\n      weight: 0.8  # May boost performance\n```\n\n### 3. Method Selection\n\n```python\n# Choose merge method based on use case:\n\n# 2 models, smooth blend → SLERP\nmerge_method = \"slerp\"\n\n# 3+ models, simple average → Linear\nmerge_method = \"linear\"\n\n# Multiple task-specific models → Task Arithmetic or TIES\nmerge_method = \"ties\"\n\n# Want to reduce redundancy → DARE\nmerge_method = \"dare_ties\"\n```\n\n### 4. Density Tuning (TIES\u002FDARE)\n\n```yaml\n# Start conservative (keep more parameters)\nparameters:\n  density: 0.8  # Keep 80%\n\n# If performance good, increase sparsity\nparameters:\n  density: 0.5  # Keep 50%\n\n# If performance degrades, reduce sparsity\nparameters:\n  density: 0.9  # Keep 90%\n```\n\n### 5. Layer-specific Merging\n\n```yaml\n# Preserve base model's beginning and end\nmerge_method: passthrough\nslices:\n  - sources:\n      - model: base_model\n        layer_range: [0, 2]     # Keep first layers\n  - sources:\n      - model: merged_middle    # Merge middle layers\n        layer_range: [2, 30]\n  - sources:\n      - model: base_model\n        layer_range: [30, 32]   # Keep last layers\n```\n\n## Evaluation & Testing\n\n### Benchmark Merged Models\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\n# Load merged model\nmodel = AutoModelForCausalLM.from_pretrained(\".\u002Fmerged-model\")\ntokenizer = AutoTokenizer.from_pretrained(\".\u002Fmerged-model\")\n\n# Test on various tasks\ntest_prompts = {\n    \"math\": \"Calculate: 25 * 17 =\",\n    \"code\": \"Write a Python function to reverse a string:\",\n    \"chat\": \"What is the capital of France?\",\n}\n\nfor task, prompt in test_prompts.items():\n    inputs = tokenizer(prompt, return_tensors=\"pt\")\n    outputs = model.generate(**inputs, max_length=100)\n    print(f\"{task}: {tokenizer.decode(outputs[0])}\")\n```\n\n### Common Benchmarks\n\n- **Open LLM Leaderboard**: General capabilities\n- **MT-Bench**: Multi-turn conversation\n- **MMLU**: Multitask accuracy\n- **HumanEval**: Code generation\n- **GSM8K**: Math reasoning\n\n## Production Deployment\n\n### Save and Upload\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\n# Load merged model\nmodel = AutoModelForCausalLM.from_pretrained(\".\u002Fmerged-model\")\ntokenizer = AutoTokenizer.from_pretrained(\".\u002Fmerged-model\")\n\n# Upload to HuggingFace Hub\nmodel.push_to_hub(\"username\u002Fmy-merged-model\")\ntokenizer.push_to_hub(\"username\u002Fmy-merged-model\")\n```\n\n### Quantize Merged Model\n\n```bash\n# Quantize with GGUF\npython convert.py .\u002Fmerged-model --outtype f16 --outfile merged-model.gguf\n\n# Quantize with GPTQ\npython quantize_gptq.py .\u002Fmerged-model --bits 4 --group_size 128\n```\n\n## Common Pitfalls\n\n### ❌ Pitfall 1: Merging Incompatible Models\n\n```yaml\n# Wrong: Different architectures\nmodels:\n  - model: meta-llama\u002FLlama-2-7b  # Llama architecture\n  - model: mistralai\u002FMistral-7B   # Mistral architecture\n```\n\n**Fix**: Only merge models with same architecture\n\n### ❌ Pitfall 2: Over-weighting One Model\n\n```yaml\n# Suboptimal: One model dominates\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.95  # Too high\n  - model: model_b\n    parameters:\n      weight: 0.05  # Too low\n```\n\n**Fix**: Use more balanced weights (0.3-0.7 range)\n\n### ❌ Pitfall 3: Not Evaluating\n\n```bash\n# Wrong: Merge and deploy without testing\nmergekit-yaml config.yml .\u002Fmerged-model\n# Deploy immediately (risky!)\n```\n\n**Fix**: Always benchmark before deploying\n\n## Resources\n\n- **mergekit GitHub**: https:\u002F\u002Fgithub.com\u002Farcee-ai\u002Fmergekit\n- **HuggingFace Tutorial**: https:\u002F\u002Fhuggingface.co\u002Fblog\u002Fmlabonne\u002Fmerge-models\n- **LazyMergekit**: Automated merging notebook\n- **TIES Paper**: https:\u002F\u002Farxiv.org\u002Fabs\u002F2306.01708\n- **DARE Paper**: https:\u002F\u002Farxiv.org\u002Fabs\u002F2311.03099\n\n## See Also\n\n- `references\u002Fmethods.md` - Deep dive into merge algorithms\n- `references\u002Fexamples.md` - Real-world merge configurations\n- `references\u002Fevaluation.md` - Benchmarking and testing strategies\n\n\n","# Model Merging: Combining Pre-trained Models\n\n## When to Use This Skill\n\nUse Model Merging when you need to:\n- **Combine capabilities** from multiple fine-tuned models without retraining\n- **Create specialized models** by blending domain-specific expertise (math + coding + chat)\n- **Improve performance** beyond single models (often +5-10% on benchmarks)\n- **Reduce training costs** - no GPUs needed, merges run on CPU\n- **Experiment rapidly** - create new model variants in minutes, not days\n- **Preserve multiple skills** - merge without catastrophic forgetting\n\n**Success Stories**: Marcoro14-7B-slerp (best on Open LLM Leaderboard 02\u002F2024), many top HuggingFace models use merging\n\n**Tools**: mergekit (Arcee AI), LazyMergekit, Model Soup\n\n## Installation\n\n```bash\n# Install mergekit\ngit clone https:\u002F\u002Fgithub.com\u002Farcee-ai\u002Fmergekit.git\ncd mergekit\npip install -e .\n\n# Or via pip\npip install mergekit\n\n# Optional: Transformer library\npip install transformers torch\n```\n\n## Quick Start\n\n### Simple Linear Merge\n\n```yaml\n# config.yml - Merge two models with equal weights\nmerge_method: linear\nmodels:\n  - model: mistralai\u002FMistral-7B-v0.1\n    parameters:\n      weight: 0.5\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      weight: 0.5\ndtype: bfloat16\n```\n\n```bash\n# Run merge\nmergekit-yaml config.yml .\u002Fmerged-model --cuda\n\n# Use merged model\npython -m transformers.models.auto --model_name_or_path .\u002Fmerged-model\n```\n\n### SLERP Merge (Best for 2 Models)\n\n```yaml\n# config.yml - Spherical interpolation\nmerge_method: slerp\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 32]\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [0, 32]\nparameters:\n  t: 0.5  # Interpolation factor (0=model1, 1=model2)\ndtype: bfloat16\n```\n\n## Core Concepts\n\n### 1. Merge Methods\n\n**Linear (Model Soup)**\n- Simple weighted average of parameters\n- Fast, works well for similar models\n- Can merge 2+ models\n\n```python\nmerged_weights = w1 * model1_weights + w2 * model2_weights + w3 * model3_weights\n# where w1 + w2 + w3 = 1\n```\n\n**SLERP (Spherical Linear Interpolation)**\n- Interpolates along sphere in weight space\n- Preserves magnitude of weight vectors\n- Best for merging 2 models\n- Smoother than linear\n\n```python\n# SLERP formula\nmerged = (sin((1-t)*θ) \u002F sin(θ)) * model1 + (sin(t*θ) \u002F sin(θ)) * model2\n# where θ = arccos(dot(model1, model2))\n# t ∈ [0, 1]\n```\n\n**Task Arithmetic**\n- Extract \"task vectors\" (fine-tuned - base)\n- Combine task vectors, add to base\n- Good for merging multiple specialized models\n\n```python\n# Task vector\ntask_vector = finetuned_model - base_model\n\n# Merge multiple task vectors\nmerged = base_model + α₁*task_vector₁ + α₂*task_vector₂\n```\n\n**TIES-Merging**\n- Task arithmetic + sparsification\n- Resolves sign conflicts in parameters\n- Best for merging many task-specific models\n\n**DARE (Drop And REscale)**\n- Randomly drops fine-tuned parameters\n- Rescales remaining parameters\n- Reduces redundancy, maintains performance\n\n### 2. Configuration Structure\n\n```yaml\n# Basic structure\nmerge_method: \u003Cmethod>  # linear, slerp, ties, dare_ties, task_arithmetic\nbase_model: \u003Cpath>      # Optional: base model for task arithmetic\n\nmodels:\n  - model: \u003Cpath\u002Fto\u002Fmodel1>\n    parameters:\n      weight: \u003Cfloat>   # Merge weight\n      density: \u003Cfloat>  # For TIES\u002FDARE\n\n  - model: \u003Cpath\u002Fto\u002Fmodel2>\n    parameters:\n      weight: \u003Cfloat>\n\nparameters:\n  # Method-specific parameters\n\ndtype: \u003Cdtype>  # bfloat16, float16, float32\n\n# Optional\nslices:  # Layer-wise merging\ntokenizer:  # Tokenizer configuration\n```\n\n## Merge Methods Guide\n\n### Linear Merge\n\n**Best for**: Simple model combinations, equal weighting\n\n```yaml\nmerge_method: linear\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      weight: 0.4\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      weight: 0.3\n  - model: NousResearch\u002FNous-Hermes-2-Mistral-7B-DPO\n    parameters:\n      weight: 0.3\ndtype: bfloat16\n```\n\n### SLERP Merge\n\n**Best for**: Two models, smooth interpolation\n\n```yaml\nmerge_method: slerp\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 32]\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [0, 32]\nparameters:\n  t: 0.5  # 0.0 = first model, 1.0 = second model\ndtype: bfloat16\n```\n\n**Layer-specific SLERP:**\n\n```yaml\nmerge_method: slerp\nslices:\n  - sources:\n      - model: model_a\n        layer_range: [0, 32]\n      - model: model_b\n        layer_range: [0, 32]\nparameters:\n  t:\n    - filter: self_attn    # Attention layers\n      value: 0.3\n    - filter: mlp          # MLP layers\n      value: 0.7\n    - value: 0.5           # Default for other layers\ndtype: bfloat16\n```\n\n### Task Arithmetic\n\n**Best for**: Combining specialized skills\n\n```yaml\nmerge_method: task_arithmetic\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1  # Math\n    parameters:\n      weight: 0.5\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B  # Chat\n    parameters:\n      weight: 0.3\n  - model: ajibawa-2023\u002FCode-Mistral-7B  # Code\n    parameters:\n      weight: 0.2\ndtype: bfloat16\n```\n\n### TIES-Merging\n\n**Best for**: Many models, resolving conflicts\n\n```yaml\nmerge_method: ties\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      density: 0.5  # Keep top 50% of parameters\n      weight: 1.0\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      density: 0.5\n      weight: 1.0\n  - model: NousResearch\u002FNous-Hermes-2-Mistral-7B-DPO\n    parameters:\n      density: 0.5\n      weight: 1.0\nparameters:\n  normalize: true\ndtype: bfloat16\n```\n\n### DARE Merge\n\n**Best for**: Reducing redundancy\n\n```yaml\nmerge_method: dare_ties\nbase_model: mistralai\u002FMistral-7B-v0.1\nmodels:\n  - model: WizardLM\u002FWizardMath-7B-V1.1\n    parameters:\n      density: 0.5    # Drop 50% of deltas\n      weight: 0.6\n  - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    parameters:\n      density: 0.5\n      weight: 0.4\nparameters:\n  int8_mask: true  # Use int8 for masks (saves memory)\ndtype: bfloat16\n```\n\n## Advanced Patterns\n\n### Layer-wise Merging\n\n```yaml\n# Different models for different layers\nmerge_method: passthrough\nslices:\n  - sources:\n      - model: mistralai\u002FMistral-7B-v0.1\n        layer_range: [0, 16]   # First half\n  - sources:\n      - model: teknium\u002FOpenHermes-2.5-Mistral-7B\n        layer_range: [16, 32]  # Second half\ndtype: bfloat16\n```\n\n### MoE from Merged Models\n\n```yaml\n# Create Mixture of Experts\nmerge_method: moe\nbase_model: mistralai\u002FMistral-7B-v0.1\nexperts:\n  - source_model: WizardLM\u002FWizardMath-7B-V1.1\n    positive_prompts:\n      - \"math\"\n      - \"calculate\"\n  - source_model: teknium\u002FOpenHermes-2.5-Mistral-7B\n    positive_prompts:\n      - \"chat\"\n      - \"conversation\"\n  - source_model: ajibawa-2023\u002FCode-Mistral-7B\n    positive_prompts:\n      - \"code\"\n      - \"python\"\ndtype: bfloat16\n```\n\n### Tokenizer Merging\n\n```yaml\nmerge_method: linear\nmodels:\n  - model: mistralai\u002FMistral-7B-v0.1\n  - model: custom\u002Fspecialized-model\n\ntokenizer:\n  source: \"union\"  # Combine vocabularies from both models\n  tokens:\n    \u003C|special_token|>:\n      source: \"custom\u002Fspecialized-model\"\n```\n\n## Best Practices\n\n### 1. Model Compatibility\n\n```python\n# ✅ Good: Same architecture\nmodels = [\n    \"mistralai\u002FMistral-7B-v0.1\",\n    \"teknium\u002FOpenHermes-2.5-Mistral-7B\",  # Both Mistral 7B\n]\n\n# ❌ Bad: Different architectures\nmodels = [\n    \"meta-llama\u002FLlama-2-7b-hf\",  # Llama\n    \"mistralai\u002FMistral-7B-v0.1\",  # Mistral (incompatible!)\n]\n```\n\n### 2. Weight Selection\n\n```yaml\n# ✅ Good: Weights sum to 1.0\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.6\n  - model: model_b\n    parameters:\n      weight: 0.4  # 0.6 + 0.4 = 1.0\n\n# ⚠️  Acceptable: Weights don't sum to 1 (for task arithmetic)\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.8\n  - model: model_b\n    parameters:\n      weight: 0.8  # May boost performance\n```\n\n### 3. Method Selection\n\n```python\n# Choose merge method based on use case:\n\n# 2 models, smooth blend → SLERP\nmerge_method = \"slerp\"\n\n# 3+ models, simple average → Linear\nmerge_method = \"linear\"\n\n# Multiple task-specific models → Task Arithmetic or TIES\nmerge_method = \"ties\"\n\n# Want to reduce redundancy → DARE\nmerge_method = \"dare_ties\"\n```\n\n### 4. Density Tuning (TIES\u002FDARE)\n\n```yaml\n# Start conservative (keep more parameters)\nparameters:\n  density: 0.8  # Keep 80%\n\n# If performance good, increase sparsity\nparameters:\n  density: 0.5  # Keep 50%\n\n# If performance degrades, reduce sparsity\nparameters:\n  density: 0.9  # Keep 90%\n```\n\n### 5. Layer-specific Merging\n\n```yaml\n# Preserve base model's beginning and end\nmerge_method: passthrough\nslices:\n  - sources:\n      - model: base_model\n        layer_range: [0, 2]     # Keep first layers\n  - sources:\n      - model: merged_middle    # Merge middle layers\n        layer_range: [2, 30]\n  - sources:\n      - model: base_model\n        layer_range: [30, 32]   # Keep last layers\n```\n\n## Evaluation & Testing\n\n### Benchmark Merged Models\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\n# Load merged model\nmodel = AutoModelForCausalLM.from_pretrained(\".\u002Fmerged-model\")\ntokenizer = AutoTokenizer.from_pretrained(\".\u002Fmerged-model\")\n\n# Test on various tasks\ntest_prompts = {\n    \"math\": \"Calculate: 25 * 17 =\",\n    \"code\": \"Write a Python function to reverse a string:\",\n    \"chat\": \"What is the capital of France?\",\n}\n\nfor task, prompt in test_prompts.items():\n    inputs = tokenizer(prompt, return_tensors=\"pt\")\n    outputs = model.generate(**inputs, max_length=100)\n    print(f\"{task}: {tokenizer.decode(outputs[0])}\")\n```\n\n### Common Benchmarks\n\n- **Open LLM Leaderboard**: General capabilities\n- **MT-Bench**: Multi-turn conversation\n- **MMLU**: Multitask accuracy\n- **HumanEval**: Code generation\n- **GSM8K**: Math reasoning\n\n## Production Deployment\n\n### Save and Upload\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\n\n# Load merged model\nmodel = AutoModelForCausalLM.from_pretrained(\".\u002Fmerged-model\")\ntokenizer = AutoTokenizer.from_pretrained(\".\u002Fmerged-model\")\n\n# Upload to HuggingFace Hub\nmodel.push_to_hub(\"username\u002Fmy-merged-model\")\ntokenizer.push_to_hub(\"username\u002Fmy-merged-model\")\n```\n\n### Quantize Merged Model\n\n```bash\n# Quantize with GGUF\npython convert.py .\u002Fmerged-model --outtype f16 --outfile merged-model.gguf\n\n# Quantize with GPTQ\npython quantize_gptq.py .\u002Fmerged-model --bits 4 --group_size 128\n```\n\n## Common Pitfalls\n\n### ❌ Pitfall 1: Merging Incompatible Models\n\n```yaml\n# Wrong: Different architectures\nmodels:\n  - model: meta-llama\u002FLlama-2-7b  # Llama architecture\n  - model: mistralai\u002FMistral-7B   # Mistral architecture\n```\n\n**Fix**: Only merge models with same architecture\n\n### ❌ Pitfall 2: Over-weighting One Model\n\n```yaml\n# Suboptimal: One model dominates\nmodels:\n  - model: model_a\n    parameters:\n      weight: 0.95  # Too high\n  - model: model_b\n    parameters:\n      weight: 0.05  # Too low\n```\n\n**Fix**: Use more balanced weights (0.3-0.7 range)\n\n### ❌ Pitfall 3: Not Evaluating\n\n```bash\n# Wrong: Merge and deploy without testing\nmergekit-yaml config.yml .\u002Fmerged-model\n# Deploy immediately (risky!)\n```\n\n**Fix**: Always benchmark before deploying\n\n## Resources\n\n- **mergekit GitHub**: https:\u002F\u002Fgithub.com\u002Farcee-ai\u002Fmergekit\n- **HuggingFace Tutorial**: https:\u002F\u002Fhuggingface.co\u002Fblog\u002Fmlabonne\u002Fmerge-models\n- **LazyMergekit**: Automated merging notebook\n- **TIES Paper**: https:\u002F\u002Farxiv.org\u002Fabs\u002F2306.01708\n- **DARE Paper**: https:\u002F\u002Farxiv.org\u002Fabs\u002F2311.03099\n\n## See Also\n\n- `references\u002Fmethods.md` - Deep dive into merge algorithms\n- `references\u002Fexamples.md` - Real-world merge configurations\n- `references\u002Fevaluation.md` - Benchmarking and testing strategies\n\n\n",{"name":5,"tags":24,"author":36,"license":15,"version":37,"description":7,"dependencies":38},[25,26,27,28,29,30,31,32,33,34,35],"Emerging Techniques","Model Merging","Mergekit","SLERP","TIES","DARE","Task Arithmetic","Model Fusion","No Retraining","Multi-Capability","Arcee AI","Orchestra Research","1.0.0",[39,40,41],"mergekit","transformers","torch","master","c247302c836de6438d746b45670990eb71691166","skills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging",1,"https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool","https:\u002F\u002Fgithub.com\u002Fqhjqhj00","chien.io","2026-05-11 15:07:59",{"source":8,"skillName":5,"commands":51,"prompt":61,"zipAvailable":62},[52,55,58],{"tool":53,"command":54},"npx","npx skills add qhjqhj00\u002Fresearch-skills-pool --skill model-merging",{"tool":56,"command":57},"bunx","bunx skills add qhjqhj00\u002Fresearch-skills-pool --skill model-merging",{"tool":59,"command":60},"pnpm","pnpm dlx skills add qhjqhj00\u002Fresearch-skills-pool --skill model-merging","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Fskills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging\u002FSKILL.md",true,[64,70,74],{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Fevaluation.md","REFERENCE","参考文档",11705,"https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Fskills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging\u002Freferences\u002Fevaluation.md",{"path":71,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Fexamples.md",8813,"https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Fskills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging\u002Freferences\u002Fexamples.md",{"path":75,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Fmethods.md",9562,"https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Fskills_collected\u002FOrchestra-Research--AI-research-SKILLs--19-emerging-techniques__model-merging\u002Freferences\u002Fmethods.md",[79,86,93,100,107,114],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M73JJB6EQPFPPVQK2Y9N67Q0","transformer-lens-interpretability","Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, studying attention patterns, or performing activation patching experiments.","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F130-transformer-lens\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M73JHTANRRWEE9HP98TPEHAY","wiki-vitepress","Packages generated wiki Markdown into a VitePress static site with dark theme, dark-mode Mermaid diagrams with click-to-zoom, and production build output. Use when the user wants to create a browsable website from generated wiki pages.","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F126-wiki-vitepress\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M73JHJBMVFW46F1Y8K3SBF43","phoenix-observability","Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM applications with detailed traces, running evaluations on datasets, or monitoring production AI systems with real-time insights.","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F124-phoenix\u002FSKILL.md",[],[],{"skillNo":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":19},"01M73JHFPA3A2806PMP42ZN86Q","frontend-ui-dark-ts","Build dark-themed React applications using Tailwind CSS with custom theming, glassmorphism effects, and Framer Motion animations. Use when creating dashboards, admin panels, or data-rich interfaces with a refined dark aesthetic.","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F123-frontend-ui-dark-ts\u002FSKILL.md",[],[],{"skillNo":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":113,"publishedAt":19},"01M73JFVBVK57NRJS49TT2VDCT","tensorboard","Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile performance with TensorBoard - Google's ML visualization toolkit","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F113-tensorboard\u002FSKILL.md",[],[],{"skillNo":115,"name":116,"nameZh":6,"description":117,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":118,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":121,"publishedAt":19},"01M73JFF6PK6ZCN41C89J7HS2B","markdown-mermaid-writing","Comprehensive markdown and Mermaid diagram writing skill. Use when creating any scientific document, report, analysis, or visualization. Establishes text-based diagrams as the default documentation standard with full style guides (markdown + mermaid), 24 diagram type references, and 9 document templates.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002Fresearch-skills-pool\u002Fblob\u002Fmaster\u002Ftop100-runs\u002F110-markdown-mermaid-writing\u002FSKILL.md",[],[]]