FINE-TUNED BY HARMAN SANDHU
RAFT · RETRIEVAL-AUGMENTED FINE-TUNE

SLM-500M · RAFT

Mentor's 500M base, fully fine-tuned on 10,000 retrieval-augmented examples — a golden passage among distractors — to answer from the right source and refuse when the answer is absent.

HuggingFace weights →
517.8M
Parameters
32,768
Vocab
1,024
Context
4.8/10
Judge score
70.8%
Groundedness
10,000
RAFT examples
Model lineage
interactiveconnecting…

Give a passage and a question

Connecting to the inference endpoint…
The grounded answer will appear here.
retrieval-augmented

What this is

The 500M base given the RAFT habit through a full fine-tune (all 517.8M parameters): answer from the passage among distractors, refuse when it is not there. Uses the same 10,000-example RAFT dataset as the Gemma RAFT model.

2 epochs (1,875 steps). On the same 500 held-out questions and the arena's blind 0–10 rubric judge it scores 4.76/10 (groundedness 70.8%, fabrication 0.6%, token-F1 0.155). Token-F1 is low because RAFT answers quote-first and terse, which lexical overlap punishes; the judge score is the fairer read.

Architecture
ClassLlamaForCausalLM
Layers24
Hidden size1,280
Attention20 heads · dim 64 · full MHA
Feed-forwardSwiGLU · inner 3,456
PositionalRoPE · θ 10,000
NormRMSNorm · ε 1e-5
Context1,024 tokens
Vocabulary32,768
Embeddingstied input/output
Pretrained byDr. Sreedath Panat
Training
Init fromslm-500m base
MethodRAFT fine-tune (full)
Trainable params517.8M — all
Training data10,000 RAFT examples · 2 epochs
Optimizer steps1,875
What this model cost to build

$2.67 total Modal usage

our cost begins at fine-tuning — the base is imported.

StageDetailCost
500M fine-tuningSFT + RAFT trained as one Modal app ($2.67 for both) — not separable per model$2.67
Total$2.67

Figures are Modal GPU usage (time × rate) across this model's lineage; shared datasets are charged at this model's share. Whether base pretraining is included is stated above — it is for the models pretrained here, and excluded for imported bases. Evaluation-derived metrics come from an independent blind-judge harness on a frozen, decontaminated held-out set. Serving is billed separately and scales to zero.