Added smaller models, surely it can run a lot faster now with smaller models

```
"gpt2-small": "117M params: Very light text generation for prototyping.",
"tinybert-4": "14M params: Extremely compact, best for classification or short text outputs.",
"smollm2-135m": "135M params: Small generative tasks with lightweight footprint.",
"qwen2.5-0.5b-instruct": "Approx 500M params: Instruction-following moderate model.",
"tinyllama-1.1b": "Approx 1.1B params: General purpose small LLM, story & code generation.",
"llama3.2-1b": "Approx 1B params: Balanced small LLM for chat and generation."
```

app.py

@@ -18,7 +18,13 @@ MODEL_INFO = {
     "google/gemma-2-9b-it": "Analytical critic that provides detailed, structured feedback.",
     "meta-llama/Meta-Llama-3-8B-Instruct": "Balanced critic; creative and nuanced, slightly more lenient.",
     "phind/Phind-CodeLlama-34B-v2": "Expert coder model; verbose but deeply logical and precise.",
-    "stabilityai/stablelm-2-12b": "Fluent natural-language generator; great for fiction and tone consistency."
+    "stabilityai/stablelm-2-12b": "Fluent natural-language generator; great for fiction and tone consistency.",
+    "gpt2-small": "117M params: Very light text generation for prototyping.",
+    "tinybert-4": "14M params: Extremely compact, best for classification or short text outputs.",
+    "smollm2-135m": "135M params: Small generative tasks with lightweight footprint.",
+    "qwen2.5-0.5b-instruct": "Approx 500M params: Instruction-following moderate model.",
+    "tinyllama-1.1b": "Approx 1.1B params: General purpose small LLM, story & code generation.",
+    "llama3.2-1b": "Approx 1B params: Balanced small LLM for chat and generation."
 }
 
 # ---------- Helpers ----------