RichardErkhov/deepseek-ai_-_deepseek-math-7b-rl-gguf

Quantization made by Richard Erkhov.

deepseek-math-7b-rl - GGUF

Model creator: https://huggingface.co/deepseek-ai/
Original model: https://huggingface.co/deepseek-ai/deepseek-math-7b-rl/

Name	Quant method	Size
deepseek-math-7b-rl.Q2_K.gguf	Q2_K	2.53GB
deepseek-math-7b-rl.Q3_K_S.gguf	Q3_K_S	2.92GB
deepseek-math-7b-rl.Q3_K.gguf	Q3_K	3.22GB
deepseek-math-7b-rl.Q3_K_M.gguf	Q3_K_M	3.22GB
deepseek-math-7b-rl.Q3_K_L.gguf	Q3_K_L	3.49GB
deepseek-math-7b-rl.IQ4_XS.gguf	IQ4_XS	3.56GB
deepseek-math-7b-rl.Q4_0.gguf	Q4_0	3.73GB
deepseek-math-7b-rl.IQ4_NL.gguf	IQ4_NL	3.74GB
deepseek-math-7b-rl.Q4_K_S.gguf	Q4_K_S	3.75GB
deepseek-math-7b-rl.Q4_K.gguf	Q4_K	3.93GB
deepseek-math-7b-rl.Q4_K_M.gguf	Q4_K_M	3.93GB
deepseek-math-7b-rl.Q4_1.gguf	Q4_1	4.1GB
deepseek-math-7b-rl.Q5_0.gguf	Q5_0	4.48GB
deepseek-math-7b-rl.Q5_K_S.gguf	Q5_K_S	4.48GB
deepseek-math-7b-rl.Q5_K.gguf	Q5_K	4.59GB
deepseek-math-7b-rl.Q5_K_M.gguf	Q5_K_M	4.59GB
deepseek-math-7b-rl.Q5_1.gguf	Q5_1	4.86GB
deepseek-math-7b-rl.Q6_K.gguf	Q6_K	5.28GB
deepseek-math-7b-rl.Q8_0.gguf	Q8_0	6.84GB

Original model description:

license: other license_name: deepseek license_link: https://github.com/deepseek-ai/DeepSeek-Math/blob/main/LICENSE-MODEL

[🏠Homepage] | [🤖 Chat with DeepSeek LLM] | [Discord] | [Wechat(微信)]

Paper Link👁️

1. Introduction to DeepSeekMath

See the Introduction for more details.

2. How to Use

Here give some examples of how to use our model.

Chat Completion

❗❗❗ Please use chain-of-thought prompt to test DeepSeekMath-Instruct and DeepSeekMath-RL:

English questions: {question}\nPlease reason step by step, and put your final answer within \boxed{}.
Chinese questions: {question}\n请通过逐步推理来解答问题，并把最终答案放置于\boxed{}中。

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig

model_name = "deepseek-ai/deepseek-math-7b-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
model.generation_config = GenerationConfig.from_pretrained(model_name)
model.generation_config.pad_token_id = model.generation_config.eos_token_id

messages = [
    {"role": "user", "content": "what is the integral of x^2 from 0 to 2?\nPlease reason step by step, and put your final answer within \\boxed{}."}
]
input_tensor = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
outputs = model.generate(input_tensor.to(model.device), max_new_tokens=100)

result = tokenizer.decode(outputs[0][input_tensor.shape[1]:], skip_special_tokens=True)
print(result)

Avoiding the use of the provided function apply_chat_template, you can also interact with our model following the sample template. Note that messages should be replaced by your input.

User: {messages[0]['content']}

Assistant: {messages[1]['content']}<｜end▁of▁sentence｜>User: {messages[2]['content']}

Assistant:

Note: By default (add_special_tokens=True), our tokenizer automatically adds a bos_token (<｜begin▁of▁sentence｜>) before the input text. Additionally, since the system prompt is not compatible with this version of our models, we DO NOT RECOMMEND including the system prompt in your input.

3. License

This code repository is licensed under the MIT License. The use of DeepSeekMath models is subject to the Model License. DeepSeekMath supports commercial use.

See the LICENSE-MODEL for more details.

4. Contact

If you have any questions, please raise an issue or contact us at service@deepseek.com.