Update README.md

README.md

@@ -10,17 +10,17 @@ pipeline_tag: text-generation
 # 🌿 Shurale7B-v1: Narrative based chit-chat model
 
 Developed
-by [@BobaZooba](https://www.linkedin.com/in/boriszubarev/) | [CV](https://docs.google.com/document/d/1BhFvIHQ1mpm81P-n2A-lhNac-U2wOGc6F2uS9gKvk88/edit?usp=sharing) | [LinkedIn](https://www.linkedin.com/in/boriszubarev/) | [[email protected]](mailto:[email protected])
+by [@BobaZooba](https://t.me/BobaZooba) | [CV](https://docs.google.com/document/d/1BhFvIHQ1mpm81P-n2A-lhNac-U2wOGc6F2uS9gKvk88/edit?usp=sharing) | [LinkedIn](https://www.linkedin.com/in/boriszubarev/) | [[email protected]](mailto:[email protected])
 
-[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/JudU3rrPP5i87CfwINANO.png" alt="Powered by X—LLM" width="175" height="32"/>](https://github.com/KompleteAI/xllm)
+[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/JudU3rrPP5i87CfwINANO.png" alt="Powered by X—LLM" width="175" height="32"/>](https://github.com/BobaZooba/xllm)
 
 # 🪄 About
 
 Model based on [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1)
 
-[GitHub Repo](https://github.com/KompleteAI/shurale) | [Detailed step-by-step guide how to train this model](https://github.com/KompleteAI/shurale/blob/main/STEP-BY-STEP-GUIDE.md)
+[GitHub Repo](https://github.com/BobaZooba/shurale) | [Detailed step-by-step guide how to train this model](https://github.com/BobaZooba/shurale/blob/main/STEP-BY-STEP-GUIDE.md)
 
-[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/4y7RfOdhxvh1Tim99uLkW.png" alt="Chat with Shurale" width="120" height="40"/>](https://t.me/ShuraleAIBot)
+[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/4y7RfOdhxvh1Tim99uLkW.png" alt="Chat with Shurale" width="120" height="40"/>](https://t.me/TaleQuestBot)
 
 | **HuggingFace Hub** | **7B**                                                        | **7B-GPTQ**                                                 |
 |---------------------|---------------------------------------------------------------|-------------------------------------------------------------|
@@ -42,7 +42,7 @@ Model based on [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.
 
 > Shurale [/ʃʊrɑˈlʲe/] is a forest spirit in Bashkir and Tatar mythology.
 
-[Do you want models as cool as this one?](https://huggingface.co/KompleteAI/Shurale7B-v1#🚀-call-to-action)
+[Do you want models as cool as this one?](https://www.linkedin.com/in/boriszubarev/)
 
 </div>
 
@@ -128,7 +128,15 @@ don't you dare let me down!
 
 # 🔧 How to use
 
-Recommended **top_p** for sampling: 0.9
+Recommended generation parameters for sampling:
+
+| Param     | Value |
+|-----------|-------|
+| top_p |  0.75     |
+| typical_p |    0.95   |
+| top_k |   50    |
+| temperature |  0.75     |
+| repetition_penalty | 1.05     |
 
 ## Transformers
 
@@ -137,8 +145,8 @@ Recommended **top_p** for sampling: 0.9
 ```python
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-tokenizer = AutoTokenizer.from_pretrained("KompleteAI/Shurale7B-v1")
-model = AutoModelForCausalLM.from_pretrained("KompleteAI/Shurale7B-v1")
+tokenizer = AutoTokenizer.from_pretrained("BobaZooba/Shurale7B-v1")
+model = AutoModelForCausalLM.from_pretrained("BobaZooba/Shurale7B-v1")
 ```
 
 2. Run generation
@@ -172,11 +180,11 @@ https://github.com/huggingface/text-generation-inference#get-started
 ### Docker
 
 ```bash
-model=KompleteAI/Shurale7B-v1
+model=BobaZooba/Shurale7B-v1
 volume=$PWD/data
 version=1.1.0  # please make sure you are using latest or stable version (>= 1.1.0)
 
-docker run --gpus all --shm-size 1g -p 8080:80 -v \
+docker run --gpus all --shm-size 1g -p 8081:80 -v \
   $volume:/data ghcr.io/huggingface/text-generation-inference:$version \
   --model-id $model --max-batch-prefill-tokens 2048 --dtype bfloat16
 ```
@@ -191,7 +199,7 @@ https://www.runpod.io/console/gpu-cloud
 | Field             | Value                                                                                                                       |
 |-------------------|-----------------------------------------------------------------------------------------------------------------------------|
 | Container Image   | ghcr.io/huggingface/text-generation-inference:1.1.0                                                                         |
-| Docker Command    | --model-id KompleteAI/Shurale7B-v1 --num-shard 1 --port 8081 --max-batch-prefill-tokens 2048 --dtype bfloat16 --json-output |
+| Docker Command    | --model-id BobaZooba/Shurale7B-v1 --num-shard 1 --port 8081 --max-batch-prefill-tokens 2048 --dtype bfloat16 --json-output |
 | Container Disk    | 5                                                                                                                           |
 | Volume Disk       | 15                                                                                                                          |
 | Volume Mount Path | /data                                                                                                                       |
@@ -252,7 +260,7 @@ print(text)
 
 # 🚄 Training Process
 
-[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/JudU3rrPP5i87CfwINANO.png" alt="Powered by X—LLM" width="175" height="32"/>](https://github.com/KompleteAI/xllm)
+[<img src="https://cdn-uploads.huggingface.co/production/uploads/6074d5f1134c000d1ae10d42/JudU3rrPP5i87CfwINANO.png" alt="Powered by X—LLM" width="175" height="32"/>](https://github.com/BobaZooba/xllm)
 
 ## Dataset
 
@@ -462,6 +470,18 @@ while True:
 
 # 📋 Dialog examples
 
+## Tale Quest
+
+`Tale Quest` is my personal project which was built using `xllm` and `Shurale`. It's an interactive text-based game
+in `Telegram` with dynamic AI characters, offering infinite scenarios
+
+You will get into exciting journeys and complete fascinating quests. Chat
+with `George Orwell`, `Tech Entrepreneur`, `Young Wizard`, `Noir Detective`, `Femme Fatale` and many more
+
+Try it now: [https://t.me/talequestbot](https://t.me/PapayaAIBot?start=Z2g)
+
+Default examples (not as interesting as in TaleQuest):
+
 <details>
 <summary>Example #1</summary>
 
@@ -589,29 +609,3 @@ Coming soon... (maybe will be in V2)
 If this model proves successful, I plan to implement an algorithm similar to DeepMind's
 ReST ([link](https://arxiv.org/pdf/2308.08998.pdf)). The mentioned work has great potential but has a number of
 shortcomings, which I've managed to address in my approach.
-
----
-
-# 🚀 Call to action
-
-**Looking for an expert in modern LLMs?** I've got the experience you need. I'll guide you through every step,
-fine-tuning everything from data collection to model training and improvement.
-
-**Why me?** Well, with six years of experience in deep learning R&D projects, I've mastered a range of roles - from
-leading a team to rolling up my sleeves as an engineer. I've built and improved products from scratch and I'm keen to do
-the same for you.
-
-**Worried about your team?** Don't be. With four years as a lecturer at Russia’s best university, I can equip them with
-the skills they need to succeed.
-
-**Want to know more?** Check
-out [my CV](https://docs.google.com/document/d/1BhFvIHQ1mpm81P-n2A-lhNac-U2wOGc6F2uS9gKvk88/edit?usp=sharing), [LinkedIn](https://www.linkedin.com/in/boriszubarev/),
-and [past projects](https://komplete.framer.ai/cases) for the full scoop.
-
-**Ready to start?** Let's arrange a free intro meeting. I'll outline the resources we'll need to make your project a
-success.  
-[Contact me form](https://komplete.framer.ai/#contact)
-
-If you're an engineer, I'd appreciate it if you could pass
-along [my LinkedIn](https://www.linkedin.com/in/boriszubarev/) or [website](https://komplete.framer.ai/) to your
-manager.