updated usage card

this is gemma 2b model from google which is fine-tuned on mental health conversations dataset

README.md

@@ -4,6 +4,9 @@ tags:
 - text-generation-inference
 - text-generation
 - peft
+- chatbot
+- depression
+- therapy
 library_name: transformers
 widget:
   - messages:
@@ -18,28 +21,29 @@ This model was trained using AutoTrain. For more information, please visit [Auto
 
 # Usage
 
-```python
+from transformers import AutoTokenizer, pipeline
+import torch
 
-from transformers import AutoModelForCausalLM, AutoTokenizer
+model = "Rhaps360/gemma-dep-ins-ft"
 
-model_path = "PATH_TO_THIS_REPO"
+tokenizer = AutoTokenizer.from_pretrained(model)
+pipeline = pipeline(
+    "text-generation",
+    model=model,
+    model_kwargs={"torch_dtype": torch.bfloat16},
+    device="cuda" if(torch.cuda.is_available()) else "cpu",
+    )
 
-tokenizer = AutoTokenizer.from_pretrained(model_path)
-model = AutoModelForCausalLM.from_pretrained(
-    model_path,
-    device_map="auto",
-    torch_dtype='auto'
-).eval()
-
-# Prompt content: "hi"
 messages = [
-    {"role": "user", "content": "hi"}
+    {"role": "user", "content": "### Context: the input message goes here. ### Response: "}
 ]
-
-input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
-output_ids = model.generate(input_ids.to('cuda'))
-response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
-
-# Model response: "Hello! How can I assist you today?"
-print(response)
-```
+prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+outputs = pipeline(
+    prompt,
+    max_new_tokens=300,
+    do_sample=True,
+    temperature=0.2,
+    top_k=50,
+    top_p=0.95
+)
+print(outputs[0]["generated_text"][len(prompt):])