RichardErkhov/invalid-coder_-_Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp - GGUF

• Model creator: https://huggingface.co/invalid-coder/
• Original model: https://huggingface.co/invalid-coder/Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp/

Name	Quant method	Size
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q2_K.gguf	Q2_K	3.73GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.IQ3_XS.gguf	IQ3_XS	4.14GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.IQ3_S.gguf	IQ3_S	4.37GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q3_K_S.gguf	Q3_K_S	4.34GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.IQ3_M.gguf	IQ3_M	4.51GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q3_K.gguf	Q3_K	4.84GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q3_K_M.gguf	Q3_K_M	4.84GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q3_K_L.gguf	Q3_K_L	5.26GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.IQ4_XS.gguf	IQ4_XS	5.43GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q4_0.gguf	Q4_0	5.66GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.IQ4_NL.gguf	IQ4_NL	5.72GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q4_K_S.gguf	Q4_K_S	5.7GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q4_K.gguf	Q4_K	6.02GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q4_K_M.gguf	Q4_K_M	6.02GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q4_1.gguf	Q4_1	6.27GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q5_0.gguf	Q5_0	6.89GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q5_K_S.gguf	Q5_K_S	6.89GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q5_K.gguf	Q5_K	7.08GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q5_K_M.gguf	Q5_K_M	7.08GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q5_1.gguf	Q5_1	7.51GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q6_K.gguf	Q6_K	8.2GB
Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp.Q8_0.gguf	Q8_0	10.62GB

Original model description:

license: apache-2.0 tags: - merge - mergekit - lazymergekit - jeonsworld/CarbonVillain-en-10.7B-v2 - kyujinpy/Sakura-SOLAR-Instruct

NeuralPipe-7B-slerp

NeuralPipe-7B-slerp is a merge of the following models using LazyMergekit:

• jeonsworld/CarbonVillain-en-10.7B-v2
• kyujinpy/Sakura-SOLAR-Instruct

🧩 Configuration

slices:
  - sources:
      - model: jeonsworld/CarbonVillain-en-10.7B-v2
        layer_range: [0, 48]
      - model: kyujinpy/Sakura-SOLAR-Instruct
        layer_range: [0, 48]
merge_method: slerp
base_model: jeonsworld/CarbonVillain-en-10.7B-v2
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5 # fallback for rest of tensors

tokenizer_source: union

dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "invalid-coder/NeuralPipe-7B-slerp"
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"])