Update README.md

README.md

@@ -9,6 +9,11 @@ base_model: TheBloke/typhoon-7B-GPTQ
 model-index:
 - name: typhoon-7b-chat-alpaca
   results: []
+datasets:
+- Thaweewat/alpaca-cleaned-52k-th
+language:
+- th
+pipeline_tag: text-generation
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -18,19 +23,58 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [TheBloke/typhoon-7B-GPTQ](https://huggingface.co/TheBloke/typhoon-7B-GPTQ) on the None dataset.
 
-## Model description
+## Usage
 
-More information needed
+```python
+from peft import AutoPeftModelForCausalLM
+from transformers import GenerationConfig, AutoTokenizer
+import torch
+import time
 
-## Intended uses & limitations
 
-More information needed
+def generate_response(input_text: str) -> str:
+    """
+    Generate a response for the given input text using the Typhoon-7B model.
 
-## Training and evaluation data
+    Parameters:
+    input_text (str): The input text prompt.
 
-More information needed
+    Returns:
+    str: The generated response.
+    """
+    # Initialize the tokenizer and model only once
+    tokenizer = AutoTokenizer.from_pretrained("Thaweewat/typhoon-7b-chat-alpaca")
 
-## Training procedure
+    model = AutoPeftModelForCausalLM.from_pretrained(
+        "Thaweewat/typhoon-7b-chat-alpaca",
+        low_cpu_mem_usage=True,
+        return_dict=True,
+        torch_dtype=torch.float16,
+        device_map="cuda")
+
+    generation_config = GenerationConfig(
+        do_sample=True,
+        top_k=1,
+        temperature=0.5,
+        max_new_tokens=300, 
+        pad_token_id=tokenizer.eos_token_id)
+
+    # Tokenize input
+    inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
+
+    # Generate outputs
+    st_time = time.time()
+    outputs = model.generate(**inputs, generation_config=generation_config)
+
+    # Decode and print response
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    print(f"Response time: {time.time() - st_time} seconds")
+    return response
+
+# Sample usage:
+input_text = "###Human: ใครคือนายกไทยคนปัจจุบัน ###Assistant: "
+print(generate_response(input_text))
+```
 
 ### Training hyperparameters
 
@@ -41,13 +85,8 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
-- training_steps: 1000
 - mixed_precision_training: Native AMP
 
-### Training results
-
-
-
 ### Framework versions
 
 - PEFT 0.7.1