You can deploy this model and start earning money today!

RichardErkhov/
soniox_-_Soniox-7B-v1.0-gguf

Name: soniox_-_Soniox-7B-v1.0-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Soniox-7B-v1.0 - GGUF

Model creator: https://huggingface.co/soniox/
Original model: https://huggingface.co/soniox/Soniox-7B-v1.0/

Name	Quant method	Size
Soniox-7B-v1.0.Q2_K.gguf	Q2_K	2.53GB
Soniox-7B-v1.0.IQ3_XS.gguf	IQ3_XS	2.81GB
Soniox-7B-v1.0.IQ3_S.gguf	IQ3_S	2.96GB
Soniox-7B-v1.0.Q3_K_S.gguf	Q3_K_S	2.95GB
Soniox-7B-v1.0.IQ3_M.gguf	IQ3_M	3.06GB
Soniox-7B-v1.0.Q3_K.gguf	Q3_K	3.28GB
Soniox-7B-v1.0.Q3_K_M.gguf	Q3_K_M	3.28GB
Soniox-7B-v1.0.Q3_K_L.gguf	Q3_K_L	3.56GB
Soniox-7B-v1.0.IQ4_XS.gguf	IQ4_XS	3.67GB
Soniox-7B-v1.0.Q4_0.gguf	Q4_0	3.83GB
Soniox-7B-v1.0.IQ4_NL.gguf	IQ4_NL	3.87GB
Soniox-7B-v1.0.Q4_K_S.gguf	Q4_K_S	3.86GB
Soniox-7B-v1.0.Q4_K.gguf	Q4_K	4.07GB
Soniox-7B-v1.0.Q4_K_M.gguf	Q4_K_M	4.07GB
Soniox-7B-v1.0.Q4_1.gguf	Q4_1	4.24GB
Soniox-7B-v1.0.Q5_0.gguf	Q5_0	4.65GB
Soniox-7B-v1.0.Q5_K_S.gguf	Q5_K_S	4.65GB
Soniox-7B-v1.0.Q5_K.gguf	Q5_K	4.78GB
Soniox-7B-v1.0.Q5_K_M.gguf	Q5_K_M	4.78GB
Soniox-7B-v1.0.Q5_1.gguf	Q5_1	5.07GB
Soniox-7B-v1.0.Q6_K.gguf	Q6_K	5.53GB
Soniox-7B-v1.0.Q8_0.gguf	Q8_0	7.17GB

Original model description:

license: apache-2.0

Model Card for Soniox-7B-v1.0

Soniox 7B is a powerful large language model. Supports English and code with 8K context. Matches GPT-4 performance on some benchmarks. Built on top of Mistral 7B, enhanced with additional pre-training and fine-tuning for strong problem-solving capabilities. Apache 2.0 License. For more details, please read our blog post.

Usage in Transformers

The model is available in transformers and can be used as follows:

python

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_path = "soniox/Soniox-7B-v1.0"
model = AutoModelForCausalLM.from_pretrained(model_path, torch_dtype=torch.float16)
tokenizer = AutoTokenizer.from_pretrained(model_path)

device = "cuda"
model.to(device)

messages = [
    {"role": "user", "content": "12 plus 21?"},
    {"role": "assistant", "content": "33."},
    {"role": "user", "content": "Five minus one?"},
]
tok_prompt = tokenizer.apply_chat_template(messages, return_tensors="pt")

model_inputs = tok_prompt.to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
decoded = tokenizer.batch_decode(generated_ids)
print(decoded[0])

Inference deployment

Refer to our documentation for inference with vLLM and other deployment options.

RichardErkhov/soniox_-_Soniox-7B-v1.0-gguf