Update README.md

README.md

@@ -1,49 +1,82 @@
 ---
+language:
+- th
+- en
+library_name: transformers
 base_model:
+- Qwen/Qwen2.5-7B-Instruct
 - Qwen/Qwen2.5-7B
-library_name: transformers
-tags:
-- mergekit
-- merge
+pipeline_tag: text-generation
+---
+
+<img src="./Tsunami.webp" alt="Tsunami Model" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
+
+# Tsunami-0.5x-7B-Instruct
+**TSUNAMI**: Transformative Semantic Understanding and Natural Augmentation Model for Intelligence.
+
+**TSUNAMI** full name was created by ChatGPT.
+
+---
+
+### infomation
+**Tsunami-0.5x-7B-Instruct** is Thai Large Language Model that fine-tuned from **Qwen2.5-7B** around **100,000** rows in Thai-specific domain.
 
 ---
-# Butler-0.7xa.6-7B-beta-Instruct
 
-This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
+### Prompt Template
 
-## Merge Details
-### Merge Method
+This model uses `ChatML` prompt template:
+
+```
+<|im_start|>system
+{System}<|im_end|>
+<|im_start|>user
+{User}<|im_end|>
+<|im_start|>assistant
+{Assistant}
+````
+
+### How to use
 
-This model was merged using the [TIES](https://arxiv.org/abs/2306.01708) merge method using [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) as a base.
 
-### Models Merged
+```python
 
-The following models were included in the merge:
-* ../model_collection/7B-SFT-merge/Butler-0.7x.6-7B-cft-Instruct
-* ../model_collection/7B-SFT-merge/Butler-0.7x.6-7B-mrg-Instruct
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
-### Configuration
+model_name = "Tsunami-th/Tsunami-0.5x-7B-Instruct"
 
-The following YAML configuration was used to produce this model:
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype="auto",
+    device_map="auto"
+)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
 
-```yaml
+messages = [
+    {"role": "system", "content": "You are a helpful assistant."},
+    {"role": "user", "content": "สวัสดีครับ"}
+]
+text = tokenizer.apply_chat_template(
+    messages,
+    tokenize=False,
+    add_generation_prompt=True
+)
 
-name: Butler-0.7xa.6-7B-beta-Instruct
-models:
-  - model: ../model_collection/7B-SFT-merge/Butler-0.7x.6-7B-cft-Instruct
-    parameters:
-      density: 1
-      weight: 1
-  - model: ../model_collection/7B-SFT-merge/Butler-0.7x.6-7B-mrg-Instruct
-    parameters:
-      density: 1
-      weight: 1
-merge_method: ties
-base_model: Qwen/Qwen2.5-7B
-dtype: bfloat16
-parameters:
-  normalize: true
-  weight: 1
-  density: 1
+inputs = tokenizer(text, return_tensors="pt")
+inputs = inputs.to(model.device)
+with torch.no_grad():
+   output = model.generate(**inputs, max_new_tokens=512)
 
+response = tokenizer.decode(output[0, len(inputs['input_ids'][0]):], skip_special_tokens=True)
 ```
+
+---
+
+### Author
+ - Pollakrit Lorprasertkul | [email protected]
+
+---
+
+ - **Tsunami-0.5x-7B-Instruct** is the version 0.5x that did not train on the whole dataset.
+ - **Tsunami-1.0-7B-Instruct** is coming soon.