L3-ColdBrew-Arcadia
L3-ColdBrew-Arcadia is a merge of the following models using LazyMergekit:
- SvalTek/L3-ColdBrew-Astrid
- FPHam/L3-8B-Everything-COT
- FPHam/L3-8B-Everything-COT
- FPHam/L3-8B-Everything-COT
🧩 Configuration
merge_method: passthrough
slices:
# Lower Layers (0–11): ColdBrew’s foundation
- sources:
- layer_range: [0, 12]
model: SvalTek/L3-ColdBrew-Astrid
# Reasoning Layers (12–23): Use FPHam for logical depth
- sources:
- layer_range: [12, 24]
model: FPHam/L3-8B-Everything-COT
# Reflection Layers (24–31): Use FPHam for reasoning and reflection
- sources:
- layer_range: [24, 32]
model: FPHam/L3-8B-Everything-COT
# Duplicate Layers (24–31): Add valid parameter growth
- sources:
- layer_range: [24, 32] # First duplicate
model: FPHam/L3-8B-Everything-COT
- layer_range: [24, 32] # Second duplicate
model: FPHam/L3-8B-Everything-COT
💻 Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "SvalTek/L3-ColdBrew-Arcadia"
messages = [{"role": "user", "content": "What is a large language model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
- Downloads last month
- 11