|
Download README.md from y68ko/zephyr: direct link, hf CLI and curl.
- Browser
- Download file 3.41 kB
-
https://huggingface.co/y68ko/zephyr/resolve/main/README.md
- Command line
-
hf download hf://y68ko/zephyr/README.md
-
curl -L -o README.md https://huggingface.co/y68ko/zephyr/resolve/main/README.md
3.41 kB
| license: apache-2.0 | |
| base_model: Qwen/Qwen2.5-0.5B-Instruct | |
| tags: | |
| - zephyr | |
| - lora | |
| - instruction-tuned | |
| - chat | |
| language: | |
| - en | |
| - ar | |
| pipeline_tag: text-generation | |
| # Zephyr ⚡ (English) | |
| [](https://discord.gg/nhF8fJympm) | |
| **Zephyr** is an instruction-tuned assistant developed by **youssef**. It is fine-tuned to act as a helpful, direct, and capable AI companion for coding, conversation, and general assistance. | |
| ## Model Details | |
| - **Model Name:** Zephyr | |
| - **Developer / Creator:** youssef | |
| - **Base Architecture:** Qwen2.5-0.5B-Instruct | |
| - **Fine-Tuning Method:** LoRA (PEFT) | |
| - **Primary Languages:** English, Arabic | |
| - **Intended Use:** Chat, coding assistance, and conversational reasoning. | |
| ## How to Use | |
| ```python | |
| import torch | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| from peft import PeftModel | |
| base_model_name = "Qwen/Qwen2.5-0.5B-Instruct" | |
| adapter_name = "y68ko/zephyr" | |
| tokenizer = AutoTokenizer.from_pretrained(base_model_name) | |
| base_model = AutoModelForCausalLM.from_pretrained( | |
| base_model_name, | |
| torch_dtype=torch.float16, | |
| device_map="auto" | |
| ) | |
| model = PeftModel.from_pretrained(base_model, adapter_name) | |
| messages = [ | |
| {"role": "user", "content": "Who are you?"} | |
| ] | |
| text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) | |
| inputs = tokenizer(text, return_tensors="pt").to(model.device) | |
| outputs = model.generate(**inputs, max_new_tokens=128) | |
| response = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True) | |
| print(response) | |
| ``` | |
| ## Community & Bot Usage | |
| [](https://discord.gg/nhF8fJympm) | |
| - **Discord Server:** [Archon Project / Zephyr](https://discord.gg/nhF8fJympm) | |
| *Join the Discord server to use and interact with the Zephyr bot directly inside channels.* | |
| --- | |
| # مشروع Zephyr ⚡ (العربية) | |
| [](https://discord.gg/nhF8fJympm) | |
| **Zephyr** هو مساعد ذكاء اصطناعي تفاعلي ومخصص تم تطويره وضبطه بواسطة **youssef**. صُمم ليكون رفيقاً ذكياً ومساعداً برمجياً وحوارياً فعالاً باللغتين العربية والإنجليزية. | |
| ## تفاصيل النموذج | |
| - **اسم النموذج:** Zephyr | |
| - **المطور:** youssef | |
| - **المعمارية الأساسية:** Qwen2.5-0.5B-Instruct | |
| - **طريقة الضبط (Fine-Tuning):** LoRA (PEFT) | |
| - **اللغات الأساسية:** العربية والإنجليزية | |
| - **الاستخدام المستهدف:** الحوار والمحادثة، المساعدة البرمجية، وإدارة المهام. | |
| ## استخدام البوت ومجتمع المشروع | |
| [](https://discord.gg/nhF8fJympm) | |
| - **سيرفر الديسكورد:** [Archon Project / Zephyr](https://discord.gg/nhF8fJympm) | |
| *انضم إلى السيرفر لاستخدام بوت Zephyr والتفاعل معه مباشرة داخل قنوات الديسكورد ومتابعة التحديثات.* | |