kevin009 / merge-17-20

Merged-Llama-Adapters-317-320

A merged LoRA adapter combining four fine-tuned adapters (317-320) for the Llama-3.1-8B language model.

Model Details

• Base Model: meta-llama/Llama-3.1-8B-instruct
• Adaptation Method: Merged LoRA
• Source Adapters:
  • https://huggingface.co/kevin009/llama317
  • https://huggingface.co/kevin009/llama318
  • https://huggingface.co/kevin009/llama319
  • https://huggingface.co/kevin009/llama320

Merger Configuration

Source Adapters

All source adapters share the following configuration:

• Rank (r): 16
• Alpha: 16
• Target Modules:
  • q_proj (Query projection)
  • k_proj (Key projection)
  • v_proj (Value projection)
  • o_proj (Output projection)
  • up_proj (Upsampling projection)
  • down_proj (Downsampling projection)
  • gate_proj (Gate projection)

Merger Details

• Merger Method: Linear interpolation
• Merger Weights: Equal weights (0.25) for each adapter
• Combined Rank: 16 (maintained from source adapters)

Usage

This merged adapter must be used with the base Llama-3.1-8B-instruct model.

Loading the Model

from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load base model
base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B-instruct")
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-3.1-8B-instruct")

# Load merged LoRA adapter
model = PeftModel.from_pretrained(base_model, "path_to_merged_adapter")

Limitations and Biases

• This merged adapter inherits limitations and biases from:
  • The base Llama-3.1-8B-instruct model
  • All four source adapters
• The merging process may result in:
  • Potential loss of specialized capabilities from individual adapters
  • Averaged behavior across different adapter specializations
  • Possible interference between adapter weights

Merging Process

The adapters were merged using the following approach:

1. Linear interpolation of adapter weights
2. Equal weighting (0.25) applied to each source adapter
3. Preservation of original LoRA rank and architecture

Method Used

The adapters were merged using PEFT (Parameter-Efficient Fine-Tuning) library's weighted adapter combination feature. The process combines multiple LoRA adapters using linear interpolation with specified weights.

Step-by-Step Merging Process

1. Load the base model and initial adapter:

from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer

MODEL_NAME = "meta-llama/Meta-Llama-3.1-8B-Instruct"
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME)
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)

# Load first adapter as base
peft_model = PeftModel.from_pretrained(model, "llama319", adapter_name="llama319")

1. Load additional adapters:

# Load remaining adapters
peft_model.load_adapter("llama320", adapter_name="llama320")
peft_model.load_adapter("llama318", adapter_name="llama318")
peft_model.load_adapter("llama317", adapter_name="llama317")

1. Configure and execute the merger:

# Define adapters and their weights
adapters = ["llama319", "llama320", "llama318", "llama317"]
weights = [1.0, 1.0, 1.0, 1.0]  # Equal weights for all adapters

# Merge adapters
peft_model.add_weighted_adapter(
    adapters, 
    weights, 
    "merge",
    combination_type="ties",  # Using ties combination method
    density=0.2              # Density parameter for merger
)

# Set active adapter to merged version
peft_model.set_adapter("merge")

# Save the merged adapter
peft_model.save_pretrained("merged")

Key Parameters

• combination_type="ties" : Uses the TIES (Task Interference Edge Selection) method for combining adapters
• density=0.2 : Controls the sparsity of the merged weights
• weights=[1.0, 1.0, 1.0, 1.0] : Equal weighting for all adapters (0.25 each after normalization)

Notes

• The order of loading adapters may affect the final result
• Equal weights were chosen to maintain balanced influence from each adapter
• The merged adapter maintains the same architecture and rank as the original adapters
• While this adapter merges multiple fine-tunes, each component was developed as part of independent research efforts to explore and language model capabilities as part of R&D process.

License

Licensed under Apache 2.0 License.

This merged adapter is part of independent individual research work. While the code is open-source under the Apache 2.0 license, please note:

• You are free to use, modify, and distribute this adapter following the Apache 2.0 license terms
• This work is provided "as is" without warranties or conditions of any kind
• This is an independent research project and not affiliated with any organization
• Attribution is appreciated but not required
• For full license details, see: https://www.apache.org/licenses/LICENSE-2.0