GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Silviase
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Silviase
Model Tree
Qwen/Qwen3.5-2B
Input Modalities
Output Modalities
Supported Functionality
Full-image OCR uses CC-OCR character P/R/F1 and clipped order-free CER (image means), with NED as a diagnostic. Training set-F1 rewards are not full-image OCR evaluation. Paper metric version: jawildtext-metrics v2 (2026-09-28). Any earlier external exact/soft set-F1 report is superseded by this evaluation.
{ "dev": { "qa": { "n": 500, "mean_reward": 0.226 }, "evidence": { "n": 500, "mean_reward": 0.5944236083818675 }, "ocr": { "n": 500, "mean_reward": 0.8816601479574927 } }, "ocr": { "metrics_version": "jawildtext-metrics v2 (2026-09-28)", "images": 580, "capped": 54, "empty": 0, "precision": 0.8702455674979207, "recall": 0.8508958354778651, "f1": 0.8404795379080438, "ned": 0.3936620992559648, "cer": 0.20126234615278146, "character_micro_f1": 0.6207167838915784, "exact_region_recall": 0.41560191423128745, "cer_detail": { "clipped_images": 55, "exact_images": 487, "exact_rate": 0.8396551724137931, "mean_gap": 0.006494104728268872, "median_image_cer": 0.0934659090909091, "reference_characters": 263496, "pooled_cer": null, "substitutions": null, "deletions": null, "insertions": null } }, "qa": { "expected": 1025, "complete": 1025, "capped": 93, "metrics": { "content": { "value": 0.408780487804878, "ci95": [ 0.3782171895971522, 0.4387703656567469 ], "units": 1025, "images": 745, "excluded_units": 0, "valid_draws": 2000 }, "boxed": { "value": 0.3882926829268293, "ci95": [ 0.3597621407333994, 0.4175552108915923 ], "units": 1025, "images": 745, "excluded_units": 0, "valid_draws": 2000 }, "format_valid": { "value": 0.8790243902439024, "ci95": [ 0.8574202631510653, 0.8982628332387975 ], "units": 1025, "images": 745, "excluded_units": 0, "valid_draws": 2000 } }, "scope": "All available outputs for this model; not a paired intersection", "repetitions": 2000, "seed": 42, "unit": "image cluster", "limitation": "Single training seed; intervals exclude training-seed variability" } }