---
license: mit
datasets:
- Bifrost-AI/Solana-Vanguard-Challenge
language:
- en
metrics:
- accuracy
- code_eval
base_model:
- Bifrost-AI/NextCoder-Mirage-sol-7B
pipeline_tag: text-generation
tags:
- code
- finance
- chat
- text-generation
- large-language-model
library_name: transformers
---
# NextCoder Mirage SOL 7B
### This fine-tuned variant of the NextCoder 7B model was supervised fine-tuned on blockchain-specific datasets(Bifrost-AI/Solana-Vanguard-Challenge), optimized for downstream tasks in blockchain coding and smart contract development on the Solana ecosystem.
The **Solana Vanguard Challenge** dataset, comprising 1,000 diverse and in-depth questions, offers full-spectrum coverage of the Solana ecosystem. It spans fundamental blockchain concepts, advanced on-chain programming in Rust and the Anchor framework, client-side integration in TypeScript, detailed security strategies, and performance as well as regulatory considerations. 

NextCoder Mirage SOL 7B is in active development with additional fine-tuning sessions, & benchmark statistics coming soon!

## Provided Quants

| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/Bifrost-AI/NextCoder-Mirage-sol-7B-GGUF/resolve/main/nextcoder-Mirage-7.6B-Q4_K_S.gguf) | Q4_K_S | 2.3 | fast, recommended |
| [GGUF](https://huggingface.co/Bifrost-AI/NextCoder-Mirage-sol-7B-GGUF/resolve/main/nextcoder-Mirage-7.6B-Q5_K_S.gguf) | Q5_K_S | 2.8 | fast, recommended |
| [GGUF](https://huggingface.co/Bifrost-AI/NextCoder-Mirage-sol-7B-GGUF/resolve/main/nextcoder-Mirage-7.6B-Q6_K.gguf) | Q6_K | 3.1 | very good quality |
| [GGUF](https://huggingface.co/Bifrost-AI/NextCoder-Mirage-sol-7B-GGUF/resolve/main/nextcoder-Mirage-7.6B-Q8_0.gguf) | Q8_0 | 4.0 | fast, best quality |
| [GGUF](https://huggingface.co/Bifrost-AI/NextCoder-Mirage-sol-7B-GGUF/resolve/main/nextcoder-Mirage-7.6B-F16.gguf) | F16 | 7.7 | 16 bpw, highest quality |


## Training Session:
- Time: 9 hours & 56 minutes
- GPU: NVIDIA GeForce RTX 3090
- Batches: 500
- Context-Size: 2043
- Batch-size: 1
- Learning-rate: 2e-5
- Training-loss: 1.09
- Eval-loss: 0.89

## Dataset Composition
- **Total Questions:** 1,000
- **Languages Covered:**
  - **Rust:** On-chain smart contract development, security best practices, advanced state management, CPIs, PDAs, and more.
  - **TypeScript:** Client-side integration using @solana/web3.js, wallet adapters, Metaplex for NFT protocols, dynamic transaction composition, and front-end dApp development.
- **Planned Extensions:**
  - **C# (Solnet):** To be integrated later for .NET ecosystem coverage.