24x24bit

24x24bit is a merge of the following models using LazyMergekit:

• liminerity/Bitnet-Mistral.0.2-v6.9

🧩 Configuration

slices:
 - sources:
 - model: liminerity/Bitnet-Mistral.0.2-v6.9
 layer_range: [0, 12]
 - model: liminerity/Bitnet-Mistral.0.2-v6.9
 layer_range: [0, 12]
 - model: liminerity/Bitnet-Mistral.0.2-v6.9
 layer_range: [12, 24]
 - model: liminerity/Bitnet-Mistral.0.2-v6.9
 layer_range: [12, 24]



merge_method: passthrough
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "gate369/24x24bit"
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"])