Update README.md

README.md

@@ -20,15 +20,15 @@ extra_gated_fields:
   I ALLOW Stability AI to email me about new model releases: checkbox
 license: other
 ---
-# `StableLM 2 Chat`
+# `StableLM 2 12B Chat`
 
 ## Model Description
 
-`Stable LM 2 Chat` is a 12 billion parameter instruction tuned language model inspired by [HugginFaceH4's Zephyr 7B](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) training pipeline. The model is trained on a mix of publicly available datasets and synthetic datasets, utilizing [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290).
+`Stable LM 2 12B Chat` is a 12 billion parameter instruction tuned language modeltrained on a mix of publicly available datasets and synthetic datasets, utilizing [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290).
 
 ## Usage
 
-`StableLM 2 Chat` uses the following instruction ChatML format
+`StableLM 2 12B Chat` uses the following instruction ChatML format
 This format is also available through the tokenizer's `apply_chat_template` method:
 
 ```python
@@ -50,15 +50,16 @@ inputs = tokenizer.apply_chat_template(
 
 tokens = model.generate(
     inputs.to(model.device),
-    max_new_tokens=1024,
-    temperature=0.5,
+    max_new_tokens=100,
+    temperature=0.7,
     do_sample=True
 )
+output = tokenizer.decode(tokens[:, inputs.input_ids.shape[-1]:][0], skip_special_tokens=False)
 
-print(tokenizer.decode(tokens[0], skip_special_tokens=False))
+print(output)
 ```
 
-StableLM 2 Chat also supports function call usage this is an example how you can use it:
+StableLM 2 12B Chat also supports function call usage this is an example how you can use it:
 ```python
 system_prompt_func = """\
 You are a helpful assistant with access to the following functions. You must use them if required -\n
@@ -81,26 +82,7 @@ You are a helpful assistant with access to the following functions. You must use
         ]
       }
     }
-    },
-    {
-        "type": "function",
-        "function": {
-        "name": "EditImage",
-        "description": "This is capable of changing, editing, adjusting, or modifying an image by describing changes to the image through a text prompt.",
-        "parameters": {
-            "type": "object",
-            "properties": {
-            "prompt": {
-                "type": "string",
-                "description": "The instruction used to edit the image."
-            }
-            },
-            "required": [
-                "prompt"
-            ]
-        }
-        }
-    }
+  }
 ]
 """
 messages = [{'role': 'system', 'content': system_prompt_func}, "user": "Help me to generate a picture of Eiffel Tower in the night!"]
@@ -116,15 +98,16 @@ tokens = model.generate(
     temperature=0.5,
     do_sample=True
 )
+output = tokenizer.decode(tokens[:, inputs.input_ids.shape[-1]:][0], skip_special_tokens=False)
 
-print(tokenizer.decode(tokens[0], skip_special_tokens=False))
+print(output)
 ```
 
 
 ## Model Details
 
 * **Developed by**: [Stability AI](https://stability.ai/)
-* **Model type**: `StableLM 2 Chat` model is an auto-regressive language model based on the transformer decoder architecture.
+* **Model type**: `StableLM 2 12B Chat` model is an auto-regressive language model based on the transformer decoder architecture.
 * **Language(s)**: English
 * **Paper**: [Stable LM 2 Chat Technical Report](https://drive.google.com/file/d/1JYJHszhS8EFChTbNAf8xmqhKjogWRrQF/view?usp=sharing)
 * **Library**: [Alignment Handbook](https://github.com/huggingface/alignment-handbook.git)
@@ -157,8 +140,8 @@ The dataset is comprised of a mixture of open datasets large-scale datasets avai
 
 ### Training Infrastructure
 
-* **Hardware**: `StableLM 2 Chat` was trained on the Stability AI cluster across 8 nodes with 8 A100 80GBs GPUs for each nodes.
-* **Code Base**: We use our internal script for SFT steps and used [HuggingFace Alignment Handbook script](https://github.com/huggingface/alignment-handbook) for DPO training.
+* **Hardware**: `StableLM 2 12B Chat` was trained on the Stability AI cluster across 8 nodes with 8 A100 80GBs GPUs for each nodes.
+* **Code Base**: We use our internal script for SFT training and [HuggingFace Alignment Handbook](https://github.com/huggingface/alignment-handbook) for DPO training.
 
 ## Use and Limitations