Sift-1B is a hyper-specialized 1.5B-parameter Small Language Model (SLM) engineered specifically to sift structured JSON parameters out of unstructured user prompts and shift routing workloads away from expensive cloud LLMs.

Trained via a two-stage Supervised Fine-Tuning (SFT) + Direct Preference Optimization (DPO) pipeline on top of Qwen/Qwen2.5-1.5B-Instruct, Sift-1B strips away all conversational fluff ("Sure! Here is your JSON:") and outputs strict, machine-readable JSON on the very first attempt.

Developed & Designed by: Sanatan Singh
Base Architecture: Qwen/Qwen2.5-1.5B-Instruct
Model Type: Causal Language Model + QLoRA Adapter + GGUF Quantization (q4_k_m)
License: MIT
GitHub Repository: SanatanSinghVishen/Sift-1B
GGUF Quantized: SanatanSinghVishen/sift-1b-gguf
DPO Golden Adapter: SanatanSinghVishen/sift-1b-dpo
SFT Adapter: SanatanSinghVishen/sift-1b-sft

📊 Comprehensive Benchmark Results

Evaluated across 50 holdout test cases containing UNSEEN function schemas (not present in training) comparing Base Qwen vs. SFT vs. DPO (Golden Release Checkpoint-750):

Evaluation Metric	Base Model (Qwen2.5-1.5B)	Sift-1B (SFT)	🏆 Sift-1B (DPO Golden)	Delta vs Base
Tool Selection Accuracy	70.0%	98.0%	100.0% ✅	+30.0%
Parameter Extraction Accuracy	34.0%	80.0%	88.0% ✅	+54.0%
JSON Parse / Validity Rate	96.0%	98.0%	100.0% ✅	+4.0%
Zero Markdown / Fluff Rate	76.0%	100.0%	100.0% ✅	+24.0%
Zero Hallucination Rate	100.0%	100.0%	100.0% ✅	0% Hallucinations
Average Latency (TTFT)	2,277 ms	1,734 ms	1,714 ms ⚡	25% Faster

⚙️ Technical Specifications & GGUF Metadata

GGUF Architecture Parameters

Quantization Format: q4_k_m (4-bit medium K-quantization)
GGUF File Size: 940.4 MB (0.94 GB)
Context Length (n_ctx): 32,768 tokens
Embedding Dimensions (n_embd): 1,536
Feed-Forward Network Dimension (n_ff): 8,960
Attention Heads (n_head): 12
Key-Value Heads (n_head_kv): 2 (Grouped-Query Attention / GQA)
RMS Norm Epsilon: 1e-6
RoPE Theta Base (rope_theta): 1,000,000.0
Vocabulary Size: 151,936 tokens (ChatML format)

Training Hyperparameters

Stage 1: Supervised Fine-Tuning (SFT)

Dataset: 10,000 ChatML multi-turn function calling samples derived from Salesforce/xlam-60k
Quantization: QLoRA 4-bit NF4 (load_in_4bit=True)
LoRA Rank ($r$): 16 | LoRA Alpha ($\alpha$): 32 | Dropout: 0
Target Modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
Batch Size: 1 per device | Gradient Accumulation: 8 (Effective batch size = 8)
Learning Rate: 2.0e-4 (Cosine schedule, warmup_ratio=0.05)
Epochs: 2 (2,500 total steps) | Optimizer: adamw_8bit

Stage 2: Direct Preference Optimization (DPO)

Dataset: Synthetically mutated preference pairs ($y_w$ = clean JSON, $y_l$ = markdown fluff/hallucinations)
DPO Temperature ($\beta$): 0.1 | Loss Type: sigmoid
Learning Rate: 5.0e-6 (Cosine schedule, warmup_ratio=0.1)
Epochs: 1 (2,500 total steps, golden checkpoint at step 750)

🚀 Quick Start Code Examples

1. Ollama (Local CLI & Server)

bash

# Run directly from Hugging Face Hub:
ollama run hf.co/SanatanSinghVishen/sift-1b-gguf