fix model paths in the code & add info about the formatting

README.md

@@ -16,7 +16,7 @@ extra_gated_fields:
   Organization or Affiliation: text
   I ALLOW Stability AI to email me about new model releases: checkbox
 ---
-# `Stable Zephyr 3B`
+# `StableLM Zephyr 3B`
 
 ## Model Description
 
@@ -25,33 +25,52 @@ extra_gated_fields:
 
 ## Usage
 
-Get started generating text with `Stable Zephyr 3B` by using the following code snippet:
+`StableLM Zephyr 3B` uses the following instruction format:
+```
+<|user|>
+List 10 synonyms for the word "tiny"<|endoftext|>
+<|assistant|>
+1. Dwarf
+2. Little
+3. Petite
+4. Miniature
+5. Small
+6. Compact
+7. Cramped
+8. Wee
+9. Nibble
+10. Crumble<|endoftext|>
+```
+
+This format is also available through the tokenizer's `apply_chat_template` method:
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
-tokenizer = AutoTokenizer.from_pretrained("stabilityai/stablelm-zephyr-3b-dpo")
+
+tokenizer = AutoTokenizer.from_pretrained('stabilityai/stablelm-zephyr-3b')
 model = AutoModelForCausalLM.from_pretrained(
-  "stable-zephyr-3b",
+  'stabilityai/stablelm-zephyr-3b',
   trust_remote_code=True,
-  torch_dtype="auto",
+  device_map="auto"
 )
-model.cuda()
-prompt = "<|user|>\nIn the field of quantum physics, what is superposition, and how does it relate to the phenomenon of quantum entanglement?<|endoftext|>\n<|assistant|>\n"
-inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
+
+prompt = [{'role': 'user', 'content': 'List 10 synonyms for the word "tiny"'}]
+inputs = tokenizer.apply_chat_template(prompt, add_generation_prompt=True, return_tensors='pt')
+
 tokens = model.generate(
-  **inputs,
+  inputs.to(model.device),
   max_new_tokens=1024,
-  temperature=0.7,
-  top_p=0.95,
-  do_sample=True,
+  temperature=0.8,
+  do_sample=True
 )
-print(tokenizer.decode(tokens[0], skip_special_tokens=True))
+
+print(tokenizer.decode(tokens[0], skip_special_tokens=False))
 ```
 
 ## Model Details
 
 * **Developed by**: [Stability AI](https://stability.ai/)
-* **Model type**: `StableLM Zephyr 3B` models are auto-regressive language models based on the transformer decoder architecture.
+* **Model type**: `StableLM Zephyr 3B` model is an auto-regressive language model based on the transformer decoder architecture.
 * **Language(s)**: English
 * **Library**: [Alignment Handbook](https://github.com/huggingface/alignment-handbook.git)
 * **Finetuned from model**: [stabilityai/stablelm-3b-4e1t](https://huggingface.co/stabilityai/stablelm-3b-4e1t)