---
type: "raw_signal"
id: "glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43"
slug: "glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43"
title: "Glas reports fine-tuning GigaAM-v3 for Russian IT dictation"
description: "Glas reports fine-tuning GigaAM-v3 for Russian IT dictation."
retrieval_nugget: "Glas reports fine-tuning GigaAM-v3 for Russian IT dictation."
observed_at: "2026-10-20"
record_date: "2026-10-20"
date_kind: "observed_at"
why_it_matters: "Source: https://habr.com/ru/articles/1095524?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"
novelty: "notable"
verification_level: "source-linked"
signal_type: "update"
source_platform: "habr.com"
topics: []
entities: ["Glas","GigaAM-v3","Whisper"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "The story reports a new fine-tuning case for GigaAM-v3 in Glas, including WER and terminology gains; none of the archive documents cover this work."
prior_refs: []
source_url: "https://habr.com/ru/articles/1095524?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"
source_urls: ["https://habr.com/ru/articles/1095524?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "signal:glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/signals/glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43/","markdown":"https://newruntime.com/signals/glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43.md","json":"https://newruntime.com/signals/glas-reports-fine-tuning-gigaam-v3-for-russian-47aa43.json"}
---

# Glas reports fine-tuning GigaAM-v3 for Russian IT dictation

## Retrieval answer

Glas reports fine-tuning GigaAM-v3 for Russian IT dictation.

Glas reports fine-tuning GigaAM-v3 for Russian IT dictation. The team trained on 25 hours of live speech. On 525 held-out short recordings, word error rate fell from 8.81% to 5.76%, while the share of correctly written terminology rose from 43.8% to 86.7%; processing speed changed little. The authors also report more word omissions and weaker capitalization and punctuation on ordinary speech. They caution that the test labels came through the same automated transcription pipeline used for training, with selective manual review, leaving possible annotation-style overlap. The adaptation is version 0.01 and remains beta.

Source: https://habr.com/ru/articles/1095524?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)