Indian Flag
Government Of India
A-
A
A+
/BharatGen ORGANISATION
UpvoteUpvote
0

Shrutam-2

Shrutam-2 is a LLM based automatic speech recognition system for 12 major Indian languages. It bridges a Conformer speech encoder with a pretrained LLM decoder through a Mixture-of-Experts (MoE) projection layer, enabling high-quality, prompt-controllable transcription across diverse Indic languages.
  • Upvotes0upvotes
  • Downloads3downloads
  • Views329views
  • File size8.37 GB
  • Shares0Shares
  • Last updatedupdated 18 May 2026
UpvoteUpvote
0
LICENSE Attribution-Non-Commercial 4.0 International (CC BY-NC 4.0)
SECTOR Sector Agnostic

About Model

Unlike conventional CTC/Attention ASR systems that map audio directly to text tokens, Shrutam-2 reframes speech recognition as a conditional language generation task. A speech encoder produces frame-level audio representations, which are then projected into the LLM's embedding space and fed to a frozen LLM decoder alongside a text prompt.

Metadata Metadata

Attribution-Non-Commercial 4.0 International (CC BY-NC 4.0)

Sector Agnostic

bharatgenai

N.A.

BharatGen

N.A.

N.A.

N.A.

N.A.

N.A.

N.A.

N.A.

Restricted

18/05/26 11:24:38

N.A.

8.37 GB

Tags

  • Automatic Speech Recognition
  • Speech-to-Text

Related Models Related Models

Dhir-5B-4bit
Dhīr 5B MoE Q4_K_M GGUF (3.4 GB) for offline llama.cpp in English + 10 Indian languages. 5.09B total/1.79B active parameters; 91% token agreement (1.5M tokens), ≤1.8% perplexity vs BF16.
Large Language Models
ZENTEIQ
  • See Upvoters3
  • Downloads14
  • Views186
  • File Size3.13 GB

microsoft/Phi-3-vision-128k-instruct
Phi-3-vision-128k-instruct is a state-of-the-art multimodal model by Microsoft, designed to process both text and visual inputs with a context length of up to 128,000 tokens.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads12
  • Views281
  • File Size0

Phi-3-Medium-4K-Instruct - Scalable AI Model for Reasoning and Code Generation
A 14B parameter AI model optimized for 4K and 128K token context lengths, designed for advanced reasoning, math, and coding applications, supporting both research and commercial use.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads0
  • Views262
  • File Size0

Phi-3-Medium-128K-Instruct ONNX-CUDA - Optimized AI Model for NVIDIA GPU Inference
An ONNX-optimized version of Phi-3-Medium-128K-Instruct, designed for high-speed, efficient inference on NVIDIA GPUs, supporting FP16 and INT4 quantization for enhanced performance.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads2
  • Views164
  • File Size0

Phi-3-Small-8K-Instruct ONNX-CUDA - Optimized AI Model for NVIDIA GPU Inference
An ONNX-optimized version of Phi-3-Small-8K-Instruct, designed for efficient inference on NVIDIA GPUs, supporting FP16 and INT4 quantization for fast, scalable AI applications.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads1
  • Views173
  • File Size0

Phi-3-Mini-4K-Instruct GGUF - Optimized AI Model for GGUF Inference
A GGUF-optimized version of Phi-3-Mini-4K-Instruct, designed for efficient low-memory AI inference, supporting quantized and full-precision formats for balanced performance and quality.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters1
  • Downloads29
  • Views1,559
  • File Size0

Phi-3-Medium-128K-Instruct - Advanced Long-Context AI for Code and Logical Reasoning
A 14B parameter AI model optimized for 128K token context length, designed for complex reasoning, mathematical problem-solving, and efficient AI applications in commercial and research environments.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads3
  • Views132
  • File Size0

Phi-3-Medium-128K-Instruct ONNX-CPU - Optimized AI Model for CPU Inference
An ONNX-optimized version of Phi-3-Medium-128K-Instruct, designed for efficient long-context inference on CPUs, supporting structured reasoning, text generation, and code processing.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads2
  • Views208
  • File Size0

Phi-3-Mini-128K-Instruct ONNX - Optimized AI Model for Long-Context ONNX Inference
An ONNX-optimized version of Phi-3-Mini-128K-Instruct, quantized for fast, scalable inference on CPUs, GPUs, and mobile devices, supporting long-context AI applications.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads1
  • Views121
  • File Size0

Phi-3-Medium-4K-Instruct ONNX-DirectML - Optimized AI Model for Windows GPU Inference
An ONNX-optimized version of Phi-3-Medium-4K-Instruct, designed for efficient AI inference on Windows machines using DirectML, supporting INT4 quantization for high-performance execution on AMD, Intel, and NVIDIA GPUs.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads0
  • Views211
  • File Size0


Related Models by Organization Related Models by Organization

sooktam2
Sooktam-2 is a multilingual Indic Text-to-Speech model by BharatGen supporting 12 languages including Hindi, Marathi, Tamil, Telugu, Bengali, Urdu, Punjabi and Indian English. It enables high-quality speech synthesis with reference-guided voice conditioning, preserving speaker voice, accent and prosody for natural and expressive generation.
BHARATGEN
  • See Upvoters0
  • Downloads28
  • Views373
  • File Size0

sooktam2
Sooktam-2 is a multilingual Indic Text-to-Speech model by BharatGen supporting 12 languages including Hindi, Marathi, Tamil, Telugu, Bengali, Urdu, Punjabi and Indian English. It enables high-quality speech synthesis with reference-guided voice conditioning, preserving speaker voice, accent and prosody for natural and expressive generation.
BHARATGEN
  • See Upvoters0
  • Downloads0
  • Views1
  • File Size0

Param-1-5B
Param-1-5B is a bilingual (English–Hindi) large language model developed under the Param-1 family. With 5 billion parameters, this model extends the capabilities of Param-1-2.9B by incorporating enhanced mathematical reasoning and code understanding/generation. The model is pretrained from scratch and designed to serve as a strong foundation for downstream tasks such as mathematical problem solving, and code-related understanding / generation.
BHARATGEN
  • See Upvoters0
  • Downloads1
  • Views217
  • File Size10.42 GB

Param-1-Instruct
BharatGen introduces the early checkpoint of SFT (Supervised Fine-Tuned) for Param 1, a bilingual language model trained from scratch in English and Hindi. With 2.9 billion parameters, this checkpoint builds upon the pretraining phase and serves as a foundation for more downstream tasks, safety testing, and customization.
BHARATGEN
  • See Upvoters0
  • Downloads19
  • Views203
  • File Size5.36 GB

BharatGen - Param 1 Indic-Scale Bilingual Foundation Model
Param1 is a 2.9 billion parameter language model pretrained on English and Hindi, designed for text completion.
BHARATGEN
  • See Upvoters4
  • Downloads753
  • Views21,703
  • File Size13.79 GB

Param2-17B-Thinking
BharatGen presents Param-2-17B-MoE-A2.4B, a large-scale Mixture-of-Experts (MoE) language model designed to deliver high model capacity while retaining the inference efficiency of a much smaller dense model. It uses a Hybrid MoE architecture with 17B total parameters, while activating only 2.4B parameters per token.
BHARATGEN
  • See Upvoters1
  • Downloads95
  • Views4,757
  • File Size57.29 GB

BharatGen Multilingual TTS - Sooktam2
Sooktam-2 is a multilingual Indic Text-to-Speech model by BharatGen supporting 12 languages including Hindi, Marathi, Tamil, Telugu, Bengali, Urdu, Punjabi and Indian English. It enables high-quality speech synthesis with reference-guided voice conditioning, preserving speaker voice, accent and prosody for natural and expressive generation.
BHARATGEN
  • See Upvoters0
  • Downloads26
  • Views2,657
  • File Size1.25 GB

BharatGen - Param-1-7B-MoE Advancing Multilingual GenAI for India
Param-1-7B-MoE is a multilingual large language model developed under the Param-1 family as part of BharatGen – A Suite of Generative AI Technologies for India. With 7 billion parameters and a Mixture of Experts (MoE) architecture, the model is designed to better understand and generate text across English, Hindi, and 14 additional Indian languages. The model is pretrained from scratch with a strong focus on linguistic diversity, cultural context, and large-scale multilingual representation.
BHARATGEN
  • See Upvoters1
  • Downloads90
  • Views2,017
  • File Size0

BharatGen-AgriParam
Large language model fine-tuned from Param-1-2.9B-Instruct on a high-quality, India-centric agriculture dataset.
BHARATGEN
  • See Upvoters0
  • Downloads55
  • Views696
  • File Size0

BharatGen-FinanceParam
large language model fine-tuned from Param-1-2.9B-Instruct on a high-quality finance dataset.
BHARATGEN
  • See Upvoters0
  • Downloads56
  • Views761
  • File Size0


Related Models by Sector Related Models by Sector

Dhir-5B-4bit
Dhīr 5B MoE Q4_K_M GGUF (3.4 GB) for offline llama.cpp in English + 10 Indian languages. 5.09B total/1.79B active parameters; 91% token agreement (1.5M tokens), ≤1.8% perplexity vs BF16.
Large Language Models
ZENTEIQ
  • See Upvoters3
  • Downloads14
  • Views186
  • File Size3.13 GB

microsoft/Phi-3-vision-128k-instruct
Phi-3-vision-128k-instruct is a state-of-the-art multimodal model by Microsoft, designed to process both text and visual inputs with a context length of up to 128,000 tokens.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads12
  • Views281
  • File Size0

Phi-3-Medium-4K-Instruct - Scalable AI Model for Reasoning and Code Generation
A 14B parameter AI model optimized for 4K and 128K token context lengths, designed for advanced reasoning, math, and coding applications, supporting both research and commercial use.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads0
  • Views262
  • File Size0

Phi-3-Medium-128K-Instruct ONNX-CUDA - Optimized AI Model for NVIDIA GPU Inference
An ONNX-optimized version of Phi-3-Medium-128K-Instruct, designed for high-speed, efficient inference on NVIDIA GPUs, supporting FP16 and INT4 quantization for enhanced performance.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads2
  • Views164
  • File Size0

Phi-3-Small-8K-Instruct ONNX-CUDA - Optimized AI Model for NVIDIA GPU Inference
An ONNX-optimized version of Phi-3-Small-8K-Instruct, designed for efficient inference on NVIDIA GPUs, supporting FP16 and INT4 quantization for fast, scalable AI applications.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads1
  • Views173
  • File Size0

Phi-3-Mini-4K-Instruct GGUF - Optimized AI Model for GGUF Inference
A GGUF-optimized version of Phi-3-Mini-4K-Instruct, designed for efficient low-memory AI inference, supporting quantized and full-precision formats for balanced performance and quality.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters1
  • Downloads29
  • Views1,559
  • File Size0

Phi-3-Medium-128K-Instruct - Advanced Long-Context AI for Code and Logical Reasoning
A 14B parameter AI model optimized for 128K token context length, designed for complex reasoning, mathematical problem-solving, and efficient AI applications in commercial and research environments.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads3
  • Views132
  • File Size0

Phi-3-Medium-128K-Instruct ONNX-CPU - Optimized AI Model for CPU Inference
An ONNX-optimized version of Phi-3-Medium-128K-Instruct, designed for efficient long-context inference on CPUs, supporting structured reasoning, text generation, and code processing.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads2
  • Views208
  • File Size0

Phi-3-Mini-128K-Instruct ONNX - Optimized AI Model for Long-Context ONNX Inference
An ONNX-optimized version of Phi-3-Mini-128K-Instruct, quantized for fast, scalable inference on CPUs, GPUs, and mobile devices, supporting long-context AI applications.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads1
  • Views121
  • File Size0

Phi-3-Medium-4K-Instruct ONNX-DirectML - Optimized AI Model for Windows GPU Inference
An ONNX-optimized version of Phi-3-Medium-4K-Instruct, designed for efficient AI inference on Windows machines using DirectML, supporting INT4 quantization for high-performance execution on AMD, Intel, and NVIDIA GPUs.
MICROSOFT CORPORATION (INDIA) PVT. LTD.
  • See Upvoters0
  • Downloads0
  • Views211
  • File Size0


Related Models by Task Type Related Models by Task Type

AI4Bharat - IndicWav2Vec-Hindi: Hindi Speech Recognition Model
A model trained to transcribe Hindi speech into text, utilizing the wav2vec architecture.
AI4BHARAT
  • See Upvoters0
  • Downloads61
  • Views1,591
  • File Size0

AI4Bharat - IndicWav2Vec-Odia: Odia Speech Recognition Model
A model trained to transcribe Odia speech into text, based on the wav2vec architecture.
AI4BHARAT
  • See Upvoters0
  • Downloads77
  • Views2,035
  • File Size0

IITG-Mizo-ASR-v1
A mizo ASR model.
DIGITAL INDIA BHASHINI DIVISION
  • See Upvoters0
  • Downloads8
  • Views55
  • File Size5.75 GB

Indic-Transcribe-Flex
Multilingual speech recognition for 27 Indian languages — with native-script, mixed-script, and romanized output.
BODHAN.AI
  • See Upvoters0
  • Downloads2
  • Views63
  • File Size4.21 GB

Indic-Transcribe-Core
Multilingual speech recognition for 25 Indian languages, in each language's native script.
BODHAN.AI
  • See Upvoters0
  • Downloads1
  • Views72
  • File Size4.22 GB

Northeast STT Multilingual Speech to Text Model
A multilingual Speech-to-Text (STT) model for eight Northeast Indian languages, fine-tuned from Whisper Medium using over 150,000 speech-text pairs from public and institutional datasets. The model expands speech recognition support for low-resource indigenous languages, including Khasi, Garo, Mizo, Kokborok, Nagamese, Assamese, Chakma, and Wancho.
MWIRE LABS
  • See Upvoters0
  • Downloads4
  • Views161
  • File Size0

AI4Bharat - IndicConformer Automatic Speech Recognition (ASR) Model for Nepali
This model is a conformer-Large model, consisting of 120M parameters, as the encoder, with a hybrid CTC-RNNT decoder.
AI4BHARAT
  • See Upvoters0
  • Downloads9
  • Views252
  • File Size0

BharatGen - ASR: Hindi
This is a 30M-parameter hybrid CTC-RNNT Branchformer model designed for Automatic Speech Recognition (ASR) in the Hindi language.
BHARATGEN
  • See Upvoters0
  • Downloads6
  • Views788
  • File Size94.49 MB