Indian Flag
Government Of India
A-
A
A+
Home/Models/Param-1-Instruct

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.

About Model

Pre-Training Details: * Dataset: 7.5 Trillion tokens * Data Quality: Highly curated with standard filtering and multiple processing steps. * Scheduler: Cosine Annealing * Learning_rate: 3e-4 to 3e-6 * Training Setup: Running on 512 H100 GPUs * Framework: NVIDIA NeMo * Precision: bf16-mixed * Base Pre-Trained Checkpoint (Param 1): https://aikosh.indiaai.gov.in/home/models/details/bharatgen_param_1_indic_scale_bilingual_foundation_model.html SFT Training Details: * Dataset: 0.8 Million samples * Epochs: 3 * Scheduler: Cosine Annealing * Learning Rate: 5e-6 to 5e-8 * Training Hardware: 32 H200 GPUs * Framework: NVIDIA NeMo * Precision: bf16-mixed

Param-1-Instruct

Metadata Metadata

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

bharatgenai

Transformers

Transformers

Open

BharatGen

Other

07/05/26 10:02:42

Kundeshwar Vijay Pundalik

5.36 GB

BharatGen%20Logo%20%281%29.png ( 141.12 KB )

To preview this file, you need to be a registered user. Please complete the registration process to gain access and continue viewing the content.

Activity Overview Activity Overview

  • Downloads0
  • Downloads 0
  • File Size 5.36 GB
  • Views 8

Tags Tags

  • Instruction-Tuning
  • QnA
  • Model Fine-Tuning

License Control License Control

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

Version Control Version Control

FolderVersion 1(5.36 GB)
  • admin·5 day(s) ago
    • image/png
      BharatGen%20Logo%20%281%29.png
    • undefined
      chat_template.jinja
    • undefined
      config_parambharatgen.py
    • application/json
      config.json
    • application/json
      generation_config.json
    • undefined
      model-00001-of-00002.safetensors
    • undefined
      model-00002-of-00002.safetensors
    • application/json
      model.safetensors.index.json
    • undefined
      modeling_parambharatgen.py
    • more_horiz 3 more

More Models from BharatGen More Models from BharatGen

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.
pretrained
  • See Upvoters0
  • Downloads0
  • File Size10.42 GB
  • Views9
Updated 4 day(s) ago
BHARATGEN
View Details
share
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.
QnA
Instruction-Tuning
Model Fine-Tuning
  • See Upvoters0
  • Downloads0
  • File Size5.36 GB
  • Views9
Updated 4 day(s) ago
BHARATGEN
View Details
share
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.
Large Language Model
  • See Upvoters4
  • Downloads678
  • File Size13.79 GB
  • Views19,436
Updated 5 day(s) ago
BHARATGEN
View Details
share
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.
Mixture of Experts
Multilingual Text
pretrained
  • See Upvoters1
  • Downloads41
  • File Size57.29 GB
  • Views1,025
Updated 1 month(s) ago
BHARATGEN
View Details
share
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.
multilingual-TTS
Text-to-Speech
Multilingual Speech
Audio Synthesis
sooktam2
  • See Upvoters0
  • Downloads9
  • File Size1.25 GB
  • Views749
Updated 2 month(s) ago
BHARATGEN
View Details
share
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.
safetensors
mixtral
region:us
  • See Upvoters1
  • Downloads79
  • File Size0
  • Views1,174
Updated 4 month(s) ago
BHARATGEN
View Details
share
BharatGen-AgriParam
Large language model fine-tuned from Param-1-2.9B-Instruct on a high-quality, India-centric agriculture dataset.
Multiturn
QnA
  • See Upvoters0
  • Downloads1
  • File Size0
  • Views39
Updated 5 month(s) ago
BHARATGEN
View Details
share
BharatGen-FinanceParam
large language model fine-tuned from Param-1-2.9B-Instruct on a high-quality finance dataset.
Multiturn
QnA
  • See Upvoters0
  • Downloads7
  • File Size0
  • Views59
Updated 5 month(s) ago
BHARATGEN
View Details
share
BharatGen-LegalParam
Large language model fine-tuned from Param-1-2.9B-Instruct on an exhaustive India-centric legal dataset.
Multiturn
QnA
Summarization
  • See Upvoters0
  • Downloads0
  • File Size0
  • Views22
Updated 5 month(s) ago
BHARATGEN
View Details
share
BharatGen-AyurParam
Large language model fine-tuned from Param-1-2.9B-Instruct on a high-quality Ayurveda dataset
Ayurvedic
  • See Upvoters0
  • Downloads1
  • File Size0
  • Views23
Updated 5 month(s) ago
BHARATGEN
View Details
share

© 2026 - Copyright AIKosh. All rights reserved. This portal is developed by National e-Governance Division for AIKosh mission.