Update README.md

README.md

@@ -1,12 +1,13 @@
 ---
 license: apache-2.0
-library_name: transformers
+library_name: gguf
 base_model: Qwen/Qwen3-14B
 tags:
 - code
 - qwen3
 - gguf
 - fine-tuned
+- conversational
 model-index:
 - name: Codeas Model
   results: []
@@ -61,23 +62,30 @@ A fine-tuned **Qwen3-14B** model optimized for code generation and reasoning tas
 
 ### Ollama
 
-```bash
-ollama create codeas -f Modelfile
-ollama run codeas
-```
+Create a `Modelfile` with the following content:
+
+```dockerfile
+FROM ./codeas-model-Q6_K.gguf
 
-### Transformers (safetensors)
+PARAMETER temperature 0.6
+PARAMETER top_p 0.95
+PARAMETER top_k 20
 
-```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
+TEMPLATE """{{- if .System }}<|im_start|>system
+{{ .System }}<|im_end|>
+{{ end }}<|im_start|>user
+{{ .Prompt }}<|im_end|>
+<|im_start|>assistant
+"""
 
-model = AutoModelForCausalLM.from_pretrained("devaloper/thinkncode", torch_dtype="bfloat16", device_map="auto")
-tokenizer = AutoTokenizer.from_pretrained("devaloper/thinkncode")
+SYSTEM "You are Codeas, a helpful coding assistant."
+```
+
+Then run:
 
-messages = [{"role": "user", "content": "Write a binary search in Rust"}]
-inputs = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True).to(model.device)
-outputs = model.generate(inputs, max_new_tokens=1024)
-print(tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True))
+```bash
+ollama create codeas -f Modelfile
+ollama run codeas
 ```
 
 ## Hardware Requirements
@@ -85,7 +93,6 @@ print(tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True))
 | Format | VRAM / RAM |
 |--------|-----------|
 | Q6_K GGUF | ~14 GB |
-| BF16 (full) | ~30 GB |
 
 ## Training