Sentence Similarity
sentence-transformers
Safetensors
Northern Hindko
Southern Hindko
xlm-roberta
feature-extraction
hindko
embeddings
low-resource
urdu-script
perso-arabic
Eval Results (legacy)
text-embeddings-inference
Instructions to use junaid008/hindko-embeddings with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- sentence-transformers
How to use junaid008/hindko-embeddings with sentence-transformers:
from sentence_transformers import SentenceTransformer model = SentenceTransformer("junaid008/hindko-embeddings") sentences = [ "The weather is lovely today.", "It's so sunny outside!", "He drove to the stadium." ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [3, 3] - Notebooks
- Google Colab
- Kaggle
Download eval/training.json from junaid008/hindko-embeddings: direct link, hf CLI and curl.
- Browser
- Download file 2.17 kB
-
https://huggingface.co/junaid008/hindko-embeddings/resolve/main/eval/training.json
- Command line
-
hf download hf://junaid008/hindko-embeddings/eval/training.json
-
curl -L -o training.json https://huggingface.co/junaid008/hindko-embeddings/resolve/main/eval/training.json
2.17 kB
| { | |
| "runs": { | |
| "r2-bgem3": { | |
| "base": "BAAI/bge-m3", | |
| "lr": 1.5e-05, | |
| "epochs": 2, | |
| "batch": 256, | |
| "mini": 32, | |
| "max_seq_length": 512, | |
| "negatives": "on", | |
| "block": 32, | |
| "steps": 322, | |
| "train_seconds": 1608.7, | |
| "peak_gb": 25.564, | |
| "gpu": "NVIDIA A100-SXM4-80GB" | |
| }, | |
| "r2-e5b-plain": { | |
| "base": "intfloat/multilingual-e5-base", | |
| "lr": 3e-05, | |
| "epochs": 2, | |
| "batch": 256, | |
| "mini": 64, | |
| "max_seq_length": 512, | |
| "negatives": "off", | |
| "block": 0, | |
| "steps": 322, | |
| "train_seconds": 369.8, | |
| "peak_gb": 17.084, | |
| "gpu": "NVIDIA A100-SXM4-80GB" | |
| }, | |
| "r2-e5b": { | |
| "base": "intfloat/multilingual-e5-base", | |
| "lr": 3e-05, | |
| "epochs": 2, | |
| "batch": 256, | |
| "mini": 64, | |
| "max_seq_length": 512, | |
| "negatives": "on", | |
| "block": 32, | |
| "steps": 322, | |
| "train_seconds": 548.6, | |
| "peak_gb": 17.091, | |
| "gpu": "NVIDIA A100-SXM4-80GB" | |
| }, | |
| "r2-e5l": { | |
| "base": "intfloat/multilingual-e5-large", | |
| "lr": 2e-05, | |
| "epochs": 2, | |
| "batch": 256, | |
| "mini": 32, | |
| "max_seq_length": 512, | |
| "negatives": "on", | |
| "block": 32, | |
| "steps": 322, | |
| "train_seconds": 1620.1, | |
| "peak_gb": 25.447, | |
| "gpu": "NVIDIA A100-SXM4-80GB" | |
| }, | |
| "r2-e5s": { | |
| "base": "intfloat/multilingual-e5-small", | |
| "lr": 5e-05, | |
| "epochs": 2, | |
| "batch": 256, | |
| "mini": 128, | |
| "max_seq_length": 512, | |
| "negatives": "on", | |
| "block": 32, | |
| "steps": 322, | |
| "train_seconds": 309.6, | |
| "peak_gb": 14.999, | |
| "gpu": "NVIDIA A100-SXM4-80GB" | |
| } | |
| }, | |
| "train_rows": { | |
| "title": 4391, | |
| "verse_ict": 2949, | |
| "prose_q": 6939, | |
| "continuation": 12904, | |
| "neighbour": 14350, | |
| "total": 41533 | |
| }, | |
| "common": { | |
| "epochs": 2, | |
| "batch_size": 256, | |
| "warmup": "10% of steps, linear decay", | |
| "weight_decay": 0.01, | |
| "loss": "CachedMultipleNegativesRankingLoss(scale=20.0, mini_batch_size=mini)", | |
| "seed": 20261001, | |
| "prefixes_e5": { | |
| "title, verse_ict, prose_q, continuation": [ | |
| "query: ", | |
| "passage: " | |
| ], | |
| "neighbour": [ | |
| "query: ", | |
| "query: " | |
| ] | |
| }, | |
| "mixed_precision": "bf16 on GPUs with compute capability >= 8 (A100)", | |
| "saved_dtype": "float16" | |
| }, | |
| "released": { | |
| "name": "r2-e5l", | |
| "weights": "1.12 GB (1,068 MiB) fp16" | |
| } | |
| } |