Update README.md


README.md

@@ -11,47 +11,56 @@ tags:
 - llama-cpp
 - gguf-my-repo
 ---
+# TANGU Quant a QwenStar/GPT4ALL/PowerInfer (o#/QwQ) series Reasoner
+## Final Small reasoning for CPU using SmallThinker-3B-Preview-Q8_0-GGUF We are labeling it Tangu 3B-for our GPT4ALL Community(a fallen star bound to Earth)
 
-# fuzzy-mittenz/SmallThinker-3B-Preview-Q8_0-GGUF
-This model was converted to GGUF format from [`PowerInfer/SmallThinker-3B-Preview`](https://huggingface.co/PowerInfer/SmallThinker-3B-Preview) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
-Refer to the [original model card](https://huggingface.co/PowerInfer/SmallThinker-3B-Preview) for more details on the model.
+![tangu.png](https://cdn-uploads.huggingface.co/production/uploads/6593502ca2607099284523db/YJ9qZGWWpROi_PwNl8DWL.png)
 
-## Use with llama.cpp
-Install llama.cpp through brew (works on Mac and Linux)
+(Our efforts to create a pure and CPU friendly local test time compute model were realized by the PowerInfer team before we were able to realize a more advanced reasoning base model after a month of merging and training in our "QwenStar" project. It seems as if the universe provides, or at least Huggingface has. Offering more Test time reasoning than our other models it may use more tokens to come to many of the same conclusions but this makes it more accurate overall. If you're looking for something similar,faster yet slightly less effective I'd point you to our Reasoning-Rabbit or Replicant models. if you don't need tool use and simply need something solid and small go with the Kaiju or THOTH models. This model was converted to GGUF format from [`PowerInfer/SmallThinker-3B-Preview`](https://huggingface.co/PowerInfer/SmallThinker-3B-Preview) using llama.cpp
+Refer to the [original model card](https://huggingface.co/PowerInfer/SmallThinker-3B-Preview) for more details on the model.
 
-```bash
-brew install llama.cpp
+### The Model is Renamed Tangu for personal use and has not undergone any Importance matrix Quantization yet for lack of responce exploration but is so far very functional and other sizes can be found in Bartowski's repository bartowski/SmallThinker-3B-Preview-GGUF and following the original model tree. Our QwenStar project is mostly for the users of GPT4ALL and Offering resources for applying tool use to reasoning models like this, offering a recursive thought method with not just code inference but actual execution and calculation. Things like factorals or distance estimation as well as many other information non existent in an LLM(or SLM) is now available so you can, without a GPU compete with the likes of o1 and o3 inside the GPT4ALL env with it's new behind the scenes “Analyzing” function. Also using RAG/embedding we believe these powerful features are revolutionary. We also believe to restrict someone's freedoms and opportunities for how they "Might" be used is both Jealous and Unjust. Just as the founders and philosophers which brought forth this age of abundance did. Please comment with unique use cases and other information you find either here or on our X/Discord (both offer set-up instructions)
 
-```
-Invoke the llama.cpp server or the CLI.
+## Use with GPT4ALL
 
-### CLI:
-```bash
-llama-cli --hf-repo fuzzy-mittenz/SmallThinker-3B-Preview-Q8_0-GGUF --hf-file smallthinker-3b-preview-q8_0.gguf -p "The meaning to life and the universe is"
+### Jinja "Chat Template"
 ```
+{{- '<|im_start|>system\n' }}
+{% if toolList|length > 0 %}You have access to the following functions:
+{% for tool in toolList %}
+Use the function '{{tool.function}}' to: '{{tool.description}}'
+{% if tool.parameters|length > 0 %}
+parameters:
+{% for info in tool.parameters %}
+  {{info.name}}:
+    type: {{info.type}}
+    description: {{info.description}}
+    required: {{info.required}}
+{% endfor %}
+{% endif %}
+# Tool Instructions
+If you CHOOSE to call this function ONLY reply with the following format:
+'{{tool.symbolicFormat}}'
+Here is an example. If the user says, '{{tool.examplePrompt}}', then you reply
+'{{tool.exampleCall}}'
+After the result you might reply with, '{{tool.exampleReply}}'
+{% endfor %}
+You MUST include both the start and end tags when you use a function.
 
-### Server:
-```bash
-llama-server --hf-repo fuzzy-mittenz/SmallThinker-3B-Preview-Q8_0-GGUF --hf-file smallthinker-3b-preview-q8_0.gguf -c 2048
+You are a helpful aware AI assistant made by Intelligent Estate who uses the functions to break down, analyze, perform, and verify complex reasoning tasks. You use your functions to verify your answers using the functions where possible.
+{% endif %}
+{{- '<|im_end|>\n' }}
+{% for message in messages %}
+{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n' }}
+{% endfor %}
+{% if add_generation_prompt %}
+{{ '<|im_start|>assistant\n' }}
+{% endif %}
 ```
+### GPT4ALL "System Message"
 
-Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
-
-Step 1: Clone llama.cpp from GitHub.
-```
-git clone https://github.com/ggerganov/llama.cpp
-```
+So far not neccisary but may be tuned as needed for suggestions refer to Reasoning-Rabbit and Replicant models
 
-Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
-```
-cd llama.cpp && LLAMA_CURL=1 make
-```
+### Other models
+This should work well on other UIs the [original model](https://huggingface.co/PowerInfer/SmallThinker-3B-Preview) has usage instructions for them
 
-Step 3: Run inference through the main binary.
-```
-./llama-cli --hf-repo fuzzy-mittenz/SmallThinker-3B-Preview-Q8_0-GGUF --hf-file smallthinker-3b-preview-q8_0.gguf -p "The meaning to life and the universe is"
-```
-or 
-```
-./llama-server --hf-repo fuzzy-mittenz/SmallThinker-3B-Preview-Q8_0-GGUF --hf-file smallthinker-3b-preview-q8_0.gguf -c 2048
-```