Documented Nemotron SFT
According to the model card, the variant was further trained with supervised fine-tuning on a curated chat subset of nvidia/Nemotron-Post-Training-Dataset-v1.
Open Source Model Profile · cjvt
GaMS-9B-Instruct-Nemotron
GaMS-9B-Instruct-Nemotron is a 9.24B-parameter Gemma2 text-generation variant from cjvt. According to the model card, it applies supervised fine-tuning to GaMS-9B-Instruct on Nemotron chat data.
- Publisher
- cjvt
- Task
- text-generation
- Model type
- gemma2
- License
- gemma
- Library
- Unknown
- Publication status
- Accepted · not indexed
Model overview
The model is published by cjvt as a Gemma2 text-generation fine-tune. The captured configuration identifies Gemma2ForCausalLM with model type gemma2, and Safetensors metadata reports 9,241,705,984 parameters. According to the model card, it is a GaMS-9B-Instruct variant further trained with SFT on a curated chat subset of nvidia/Nemotron-Post-Training-Dataset-v1.
Recorded capabilities
Slovenian plus English mixture
According to the model card, training data included about 80k Slovenian instruction-response pairs and about 20k English examples.
Thesis and programme context
According to the model card, the instruction-tuned GaMS 9B was developed in a master thesis project by Timotej Petric within the PoVeJMo programme.
Gemma license
Card data records gemma, with hub tags repeating a license:gemma entry.
Use cases in the source record
- Slovenian and English instruction-following chat using the documented SFT chat-data mixture.
- Research reuse within the PoVeJMo and SloLLaMai Slovenian-language modelling programme described in the card.
Limitations and unknowns
- No evaluation results, context-window value, or hardware requirements were extracted.
- Training detail beyond the SFT dataset mixture and programme acknowledgement was not extracted.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: cjvt/GaMS-9B-Instruct-Nemotron
Captured: Unknown. Processed: 2026-09-07T19:35:55.020253+00:00.
Model Card for GaMS-9B-Instruct-Nemotron GaMS-9B-Instruct-Nemotron is a variant of GaMS-9B-Instruct, further trained with supervised fine-tuning (SFT) on a curated subset of the chat part of nvidia/Nemotron-Post-Training-Dataset-v1 . The training data included ~80k Slovenian instruction–response pairs (based on translations with additional modifications to adjust identity and context) and ~20k English examples. This instruction tuned version of GaMS 9B was developed as part of a master thesis project by Timotej Petrič. Further details will follow. Acknowledgment The model was developed within the PoVeJMo research program (Adaptive N…
F001F002F003F004F005F006F007F008F009F010F011F012F013