Datasets:

archit11
/

hyperswitch-token-aware-cpt

Modalities:

Text

Formats:

Size:

Libraries:

Dataset card Data Studio Files Files and versions

xet

Community

Dataset Viewer

Auto-converted to Parquet Duplicate

Split (1)

train · 1.08k rows

id stringlengths 23 58	type stringclasses 1 value	granularity stringclasses 4 values	content stringlengths 8.09k 87.1k	metadata dict
large_file_-1896727766380240850	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/mock_db.rs Tokens: 2,201 </path> <file> use std::sync::Arc; use common_utils::{errors::CustomResult, types::keymanager::KeyManagerState}; use diesel_models as store; use error_stack::ResultExt; use futures::lock::{Mutex, MutexGuard}; use...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/mock_db.rs", "files": null, "module": null, "num_files": null, "token_count": 2201 }
large_file_3755298731573652989	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/merchant_account.rs Tokens: 6,114 </path> <file> #[cfg(feature = "olap")] use std::collections::HashMap; use common_utils::ext_traits::AsyncExt; use diesel_models::merchant_account as storage; use error_stack::{report, ResultExt}; use hy...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/merchant_account.rs", "files": null, "module": null, "num_files": null, "token_count": 6114 }
large_file_-4292514424549763470	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/payment_method.rs Tokens: 7,297 </path> <file> pub use diesel_models::payment_method::PaymentMethod; use crate::redis::kv_store::KvStorePartition; impl KvStorePartition for PaymentMethod {} use common_enums::enums::MerchantStorageSchem...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/payment_method.rs", "files": null, "module": null, "num_files": null, "token_count": 7297 }
large_file_825304971069536194	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/merchant_connector_account.rs Tokens: 10,476 </path> <file> use async_bb8_diesel::AsyncConnection; use common_utils::{encryption::Encryption, ext_traits::AsyncExt}; use diesel_models::merchant_connector_account as storage; use error_stack...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/merchant_connector_account.rs", "files": null, "module": null, "num_files": null, "token_count": 10476 }
large_file_302692284225056352	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/configs.rs Tokens: 2,108 </path> <file> use diesel_models::configs as storage; use error_stack::report; use hyperswitch_domain_models::configs::ConfigInterface; use router_env::{instrument, tracing}; use crate::{ connection, erro...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/configs.rs", "files": null, "module": null, "num_files": null, "token_count": 2108 }
large_file_3038524595516078685	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/lib.rs Tokens: 3,574 </path> <file> use std::{fmt::Debug, sync::Arc}; use common_utils::types::TenantConfig; use diesel_models as store; use error_stack::ResultExt; use hyperswitch_domain_models::{ behaviour::{Conversion, ReverseConv...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/lib.rs", "files": null, "module": null, "num_files": null, "token_count": 3574 }
large_file_-854299013555929450	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/business_profile.rs Tokens: 3,433 </path> <file> use common_utils::ext_traits::AsyncExt; use diesel_models::business_profile::{self, ProfileUpdateInternal}; use error_stack::{report, ResultExt}; use hyperswitch_domain_models::{ behavi...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/business_profile.rs", "files": null, "module": null, "num_files": null, "token_count": 3433 }
large_file_-8980370308406515624	clm	file	<path> Repository: hyperswitch Crate: storage_impl File: crates/storage_impl/src/errors.rs Tokens: 2,457 </path> <file> pub use common_enums::{ApiClientError, ApplicationError, ApplicationResult}; pub use redis_interface::errors::RedisError; use crate::store::errors::DatabaseError; pub type StorageResult<T> = error_s...	{ "crate": "storage_impl", "file": "crates/storage_impl/src/errors.rs", "files": null, "module": null, "num_files": null, "token_count": 2457 }
large_file_6443090856209046879	clm	file	"<path>\nRepository: hyperswitch\nCrate: storage_impl\nFile: crates/storage_impl/src/customers.rs\nT(...TRUNCATED)	{"crate":"storage_impl","file":"crates/storage_impl/src/customers.rs","files":null,"module":null,"nu(...TRUNCATED)
large_file_-996606669328792716	clm	file	"<path>\nRepository: hyperswitch\nCrate: storage_impl\nFile: crates/storage_impl/src/kv_router_store(...TRUNCATED)	{"crate":"storage_impl","file":"crates/storage_impl/src/kv_router_store.rs","files":null,"module":nu(...TRUNCATED)

End of preview. Expand in Data Studio

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Hyperswitch Token-Aware CPT Dataset

This dataset contains 1,076 samples of Rust code from the Hyperswitch payment router project, optimized for Continued Pre-Training (CPT) with the Kwaipilot/KAT-Dev tokenizer.

Dataset Statistics

Total Samples: 1,076
Total Tokens: 5,687,255
Mean Tokens per Sample: 5,285
Token Range: 2,001 - 15,609

Token Distribution

< 4k tokens: 38.1% of samples
4k-10k tokens: 52.0% of samples
10k+ tokens: 9.9% of samples

Granularity Types

file: 721 samples (single large files)
module: 180 samples (multiple files from same module)
combined_files: 160 samples (small files combined by crate)
crate: 15 samples (entire small crates)

Top Crates

router - 371 samples
hyperswitch_connectors - 336 samples
analytics - 54 samples
diesel_models - 39 samples
api_models - 28 samples

Sample Structure

Each sample contains:

id: Unique identifier
type: Sample type (always "clm" for causal language modeling)
granularity: Level of code organization (file/module/combined_files/crate)

content: Full code with path metadata in format:

<path>
Repository: hyperswitch
Crate: [crate_name]
File: [file_path]
Tokens: [token_count]
</path>

<file>
[actual code content]
</file>

metadata: Contains crate, file info, and token count

Usage

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("archit11/hyperswitch-token-aware-cpt")

# Access samples
sample = dataset['train'][0]
print(f"Tokens: {sample['metadata']['token_count']:,}")
print(f"Crate: {sample['metadata']['crate']}")
print(f"Granularity: {sample['granularity']}")

# Filter by token count
medium_samples = dataset['train'].filter(
    lambda x: 4000 <= x['metadata']['token_count'] < 10000
)

# Filter by crate
router_samples = dataset['train'].filter(
    lambda x: x['metadata']['crate'] == 'router'
)

Training Recommendations

Context Length: 16k tokens (max sample is 15,609 tokens)
Tokenizer: Kwaipilot/KAT-Dev
Suggested Batch Size: 1-2 samples per batch (due to large context)
Format: Samples are pre-formatted with <path> and <file> tags

Source

Repository: https://github.com/juspay/hyperswitch
Language: Rust
License: Apache 2.0

Generation Method

Samples were generated using token-aware strategies:

Large files (2k-16k tokens) included as-is
Small files combined within same crate until reaching 2k+ tokens
Module clusters grouped by directory structure
Complete crates for small crates that fit within context

All token counts measured using the Kwaipilot/KAT-Dev tokenizer.

Downloads last month: 3