THIS IS THE ONE FOR SURE.

EXPERIMENTAL

<|im_start|>system
bla bla bla<|im_end|>
<|im_start|>input
bla bla bla<|im_end|>
<|im_start|>output
bla bla bla<|im_end|>

recommended prompt:

[ Series: ...; Chapter: ...; Scenario: ...; Tags: ...]

Picaro-v1-210-0.7-24B

Picaro-v1-210-0.7-24B is a merge of the following models using LazyMergekit:

• Trappu/Picaro-v1-210-24B
• IntervitensInc/Mistral-Small-24B-Instruct-2501-chatml

🧩 Configuration

models:
 - model: Trappu/Picaro-v1-210-24B
   parameters:
     density: 0.7
     weight: 1
 - model: IntervitensInc/Mistral-Small-24B-Instruct-2501-chatml
   parameters:
     density: 0.4
     weight: 0.3
merge_method: ties
base_model: IntervitensInc/Mistral-Small-24B-Instruct-2501-chatml
dtype: float16
tokenizer:
  source: Trappu/Picaro-v1-210-24B
parameters:
  normalize: true
  int8_mask: true

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Trappu/Picaro-v1-210-0.7-24B"
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"])