datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
sarvam-tts-dataset
Sarvam TTS Dataset: Indian English and Hindi
A curated speech dataset for TTS model training, containing 396 clips in Indian English (en-IN) and Hindi (hi-IN) totalling 57.3 minutes (3440 seconds).
Pipeline source code: https://github.com/Ayush147258/sarvam-tts-dataset
Dataset Summary
Stat
Value
Total clips
396
Total duration
57.3 minutes (3440 seconds)
Clip duration range
5.0s to 26.5s
Mean clip duration
8.7s
Median clip duration
7.5s… See the full description on the dataset page: https://huggingface.co/datasets/ayush712145/sarvam-tts-dataset.sarvam-top10-review
Sarvam Top-10 Task Review
10 tasks from the 118-task gworkspace_hard_v1 benchmark, selected for trace
richness (actions per goal ≥ 1.5). 5 tasks where Sarvam-105B passed; 5 engaged
failures. Each row is a Sarvam-105B rollout (n=1); the other three models'
results live in the peer_results column.
Selection
Passes (5): all Sarvam-105B passes in the dataset, per the trace file.
Failures (5): Sarvam ≥7 actions, failed, capped at 2 per family for diversity.… See the full description on the dataset page: https://huggingface.co/datasets/rajatagarwal457/sarvam-top10-review.sarvamai__OpenHathi-7B-Hi-v0.1-Base-details
Dataset Card for Evaluation run of sarvamai/OpenHathi-7B-Hi-v0.1-Base
Dataset automatically created during the evaluation run of model sarvamai/OpenHathi-7B-Hi-v0.1-Base
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/sarvamai__OpenHathi-7B-Hi-v0.1-Base-details.mtbenchtemplaw_embed_v1