Update README.md

README.md

@@ -5,18 +5,43 @@ license: apache-2.0
 tags:
 - text-generation-inference
 - transformers
-- unsloth
 - llama
 - trl
 base_model: unsloth/llama-3-70b-bnb-4bit
 ---
 
-# Uploaded  model
-
-- **Developed by:** tema7707
+# Uploaded Model
+- **Developed by:** [Salt AI](https://getsalt.ai/)
 - **License:** apache-2.0
-- **Finetuned from model :** unsloth/llama-3-70b-bnb-4bit
+- **Finetuned from model:** unsloth/llama-3-70b-bnb-4bit
+
+This is a llama3-70b model that was finetuned on ComfyUI data. This model can generate ComfyUI JSONs from text.
+
+## Example Prompts and Outputs
+
+|           Prompt                                    | Generated Workflow |
+|:---------------------------------------------------:|:----------------|
+| `Create a simple ComfyUI pipeline that performs basic text-to-image generation.` | ![Example1](https://huggingface.co/tema7707/comfyui-llm-alpha/resolve/main/example-1.png) |
+| `Create a simple ComfyUI pipeline that performs basic text-to-video generation.` | ![Example2](https://huggingface.co/tema7707/comfyui-llm-alpha/resolve/main/example-2.png) |
+| `Create a simple ComfyUI pipeline that performs face swap and face enhancement.` | ![Example3](https://huggingface.co/tema7707/comfyui-llm-alpha/resolve/main/example-3.png) |
+| `Create a simple ComfyUI pipeline that performs text-to-image generation with Instant ID and Openpose Controlnet.` | ![Example4](https://huggingface.co/tema7707/comfyui-llm-alpha/resolve/main/example-4.png) |
+| `Create a ComfyUI pipeline that performs text-to-image generation using Lora. Format the pipeline using groups.` | ![Example5](https://huggingface.co/tema7707/comfyui-llm-alpha/resolve/main/example-5.png) |
+
+## Disclaimer
+
+This is an alpha version. It can generate pipelines that may not work as intended and may require manual fixes.
+
+The model was trained utilizing the Alpaca format for prompts. Consequently, it is recommended to structure the prompts in this format to potentially enhance the quality of the outputs:
+```
+Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
+
+### Instruction:
+Below is a description of the ComfyUI pipeline; please generate a ComfyUI JSON for it.
+
+### Input:
+{your idea}
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+### Response:
+```
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
+To improve quality, it is advisable to explicitly specify the main nodes to be used in the pipeline.