Vyom Mind 1.0 (Qwen2.5-7B LoRA)
Vyom Mind 1.0 is the customized conversational intelligence checkpoint for Project Vyom. It is built on top of Qwen/Qwen2.5-7B-Instruct using parameter-efficient QLoRA fine-tuning.
Model Details
- Developer: Project Vyom
- Base Checkpoint: Qwen/Qwen2.5-7B-Instruct
- Fine-Tuning Method: QLoRA (4-bit NF4 Quantization + LoRA rank 16, alpha 32)
- Target Modules: q_proj, k_proj, _proj, o_proj, gate_proj, up_proj, down_proj
- Training Infrastructure: Kaggle GPU (Tesla T4 x2)
- Training Notebook: Kaggle Notebook
Capabilities
- Conversational reasoning and code synthesis
- ChatML prompt template compliance
- Grounded web research integration
- Contextual follow-up suggestions generation