Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+lamp-gemma-4b.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
+lamp-llama-3b.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
+training_results.png filter=lfs diff=lfs merge=lfs -text

Modelfile.lamp-gemma-4b

@@ -0,0 +1,4 @@
+FROM ./lamp-gemma-4b.Q8_0.gguf
+PARAMETER temperature 0.3
+PARAMETER num_predict 4096
+PARAMETER stop <end_of_turn>

Modelfile.lamp-llama-3b

@@ -0,0 +1,4 @@
+FROM ./lamp-llama-3b.Q8_0.gguf
+PARAMETER temperature 0.3
+PARAMETER num_predict 4096
+PARAMETER stop <|eot_id|>

README.md

@@ -0,0 +1,67 @@
+---
+license: apache-2.0
+tags:
+  - lamp
+  - iot
+  - pixel-art
+  - light-control
+  - fine-tuned
+  - gguf
+  - ollama
+---
+
+# LAMP Fine-Tuned Models
+
+Fine-tuned language models for the **LAMP** (Moonside Lamp Emulator) project. These models generate JSON light programs from natural language descriptions.
+
+## Models
+
+| Model | Base | Parameters | GGUF Size | Final Eval Loss |
+|-------|------|-----------|-----------|-----------------|
+| `lamp-llama-3b.Q8_0.gguf` | Llama 3.2 3B Instruct | 3.2B | 3.2 GB | 0.0294 |
+| `lamp-gemma-4b.Q8_0.gguf` | Gemma 3 4B IT | 4.3B | 3.9 GB | 0.0247 |
+
+## Training Details
+
+- **Method**: Full fine-tune (100% parameters trainable, not LoRA)
+- **Precision**: bf16
+- **Hardware**: NVIDIA H200 (140GB VRAM)
+- **Framework**: Unsloth + HuggingFace TRL
+- **Dataset**: 2,268 train / 253 validation examples
+- **Epochs**: 3
+- **Batch size**: 16 (4 per device x 4 gradient accumulation)
+- **Learning rate**: 2e-5 with cosine decay
+- **Optimizer**: AdamW
+
+## Training Results
+
+![Training Results](training_results.png)
+
+Both models converged well:
+- **Llama 3.2 3B**: Loss 1.38 -> 0.018, eval loss 0.0294
+- **Gemma 3 4B**: Loss 1.37 -> 0.018, eval loss 0.0247
+
+## Usage with Ollama
+
+1. Download the GGUF file
+2. Create a Modelfile:
+
+```
+# Modelfile.lamp-llama-3b
+FROM ./lamp-llama-3b.Q8_0.gguf
+PARAMETER temperature 0.3
+PARAMETER num_predict 4096
+PARAMETER stop <|eot_id|>
+```
+
+3. Create and run:
+```bash
+ollama create lamp-llama-3b -f Modelfile.lamp-llama-3b
+ollama run lamp-llama-3b "warm and cozy"
+```
+
+## Example
+
+**Input**: "Create a light program for: warm and cozy"
+
+**Output**: A JSON program controlling LED pixels with colors, animations, and timing for a warm ambient effect.

lamp-gemma-4b.Q8_0.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee2ed4dc203391abd941653f5b04598f90656b6d9be0cb1a6ac8f0290534e808
+size 4130401952

lamp-llama-3b.Q8_0.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02eac7bad9f7a3681923b75ea8006b24daf696df4b4170a3d0cee673c4bb8e9f
+size 3421898944