Text Generation
PEFT
Safetensors
lora
vision-language
qwen3-vl
coding
math
darija
fine-tuned
beginner-friendly
conversational
Instructions to use CompiwerAI/Mtrini-SVL-1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use CompiwerAI/Mtrini-SVL-1.0 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-VL-8B-Instruct") model = PeftModel.from_pretrained(base_model, "CompiwerAI/Mtrini-SVL-1.0") - Notebooks
- Google Colab
- Kaggle
Add mtrini_svl_metadata.json
Browse files- mtrini_svl_metadata.json +15 -0
mtrini_svl_metadata.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": "Qwen/Qwen3-VL-8B-Instruct",
|
| 3 |
+
"dataset_a": "HuggingFaceH4/Bespoke-Stratos-17k",
|
| 4 |
+
"dataset_b": "open-r1/OpenR1-Math-220k",
|
| 5 |
+
"max_steps": 1000,
|
| 6 |
+
"batch_size": 4,
|
| 7 |
+
"gradient_accumulation": 4,
|
| 8 |
+
"effective_batch": 16,
|
| 9 |
+
"context_length": 2048,
|
| 10 |
+
"lora_r": 16,
|
| 11 |
+
"lora_alpha": 32,
|
| 12 |
+
"learning_rate": 0.0002,
|
| 13 |
+
"checkpoint_steps": 500,
|
| 14 |
+
"seed": 42
|
| 15 |
+
}
|