GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Tanny03
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
Tanny03
Model Tree
Qwen/Qwen2.5-1.5B-Instruct
Input Modalities
Output Modalities
Supported Functionality
markdown
Classify the customer's banking request into one intent label.Request: {text}Intent:
Raw text, no chat template. Greedy decoding, at most 12 new tokens. Replace {text} with the input.
Golden sets are frozen random held-out splits; every system below was run on the same items. The hard-cases split is mined from this adapter's own failures, so it is report-only and sits near zero by construction for classification.
| system | split (n) | metric | score |
|---|---|---|---|
| this adapter, run 1 / run 2 | golden (770) | micro_accuracy | 0.9286 / 0.9299 |
| base model, 77 demonstrations | golden (770) | micro_accuracy | 0.5727 |
| GPT-4o-mini (frontier reference) | golden | micro_accuracy | 0.6870 |
| this adapter | hard cases (57), report-only | micro_accuracy | 0.0000 |
Latency with all four adapters served at once on one A10 (vLLM, concurrency 16): P50 82 ms · P95 136 ms.
QLoRA (4-bit NF4) on Qwen/Qwen2.5-1.5B-Instruct, LoRA rank 16, alpha 32, on all attention and MLP projections; prompt tokens masked from the loss. 8,495 training rows from mteb/banking77 (mit).
Full decision log, results and negative findings: https://github.com/tpawar03/AdapterOps.