Update README.md

README.md

@@ -12,7 +12,7 @@ pipeline_tag: text-generation
 
 ## 🧠 Training Configuration
 
-This model was trained using the configuration file [`configs/train_small.yaml`](https://huggingface.co/donribbs/scraps-llm-model/blob/main/configs/train_small.yaml).
+This model was trained using the model architecture:
 
 ### ⚙️ Model Architecture
 | Parameter | Value | Description |
@@ -68,7 +68,7 @@ This model was trained using the configuration file [`configs/train_small.yaml`]
 | **Platform** | Google Colab Pro+ |
 | **GPU** | NVIDIA A100 (80 GB) |
 | **Runtime** | PyTorch 2.x |
-| **Training time** | ~4–5 hours |
+| **Training time** | ~18-22 hours |
 | **Mixed precision** | Enabled (AMP) |
 
 ---
@@ -83,7 +83,7 @@ This model was trained using the configuration file [`configs/train_small.yaml`]
 ---
 
 ### 🧩 Summary
-Scraps-LLM is a **136 M-parameter decoder-only Transformer** trained to generate complete cooking recipes from a list of input ingredients.  
+Scraps-LLM is a **138 M-parameter decoder-only Transformer** trained to generate complete cooking recipes from a list of input ingredients.  
 The model learns via **causal language modeling (next-token prediction)** on RecipeNLG data, producing structured, human-readable recipes that include titles and numbered steps.  
 It was later exported to **ONNX** for lightweight CPU inference and integrated into a Hugging Face Space demo.
 
@@ -102,5 +102,4 @@ It was later exported to **ONNX** for lightweight CPU inference and integrated i
 |------|-------------|
 | `export/scraps.onnx` | ONNX-optimized inference graph |
 | `tokenizer/bpe.json` | BPE vocabulary used for encoding/decoding |
-| `configs/train_small.yaml` | Original training configuration |
-| `best_model.pt` | PyTorch checkpoint (~136 M parameters) |
+| `best_model.pt` | PyTorch checkpoint (~138 M parameters) |