Merge branch '131-inconsistent-temperatures-kinetic-energies-between-ase-atoms-and-thermo-out-file' into 'master'

Resolve "Inconsistent temperatures/kinetic energies between ASE atoms and thermo.out file"

Closes #131

See merge request !223

Resolve "Add support for charges for GPUNEP"

Closes #128 and #115

See merge request !224

Resolve "Bump GPUMD to v4.8 in CI"

Closes #132

See merge request !226

remove old theme and pin sphinxcontrib to avoid PIP solver failure

See merge request !218

Add wrapper function for setting up finetuning

Closes #112

See merge request !225

Resolve "Add tutorial on TNEP"

Closes #116

See merge request !227

actually add the tnep tutorial to the docs

See merge request !228

Add per-atom information in read_structures

See merge request !220

Integrate reading/writing restart files with reading/writing nep.txt files

See merge request !231

Fix BEC NaN handling for structures without target data in read_structures

See merge request !234

Fixed handling of string values in write_nepfile

See merge request !236

Support higher-body descriptors and extended l_max format

See merge request !233

Fixed bug related to selection keyword argument in get_parity_data

See merge request !237

Support for model architecture manipulation

See merge request !235

Fix handling of errors with newer spglib version

See merge request !240

Added functionality for analyzing dielectric function

See merge request !232

Add IR and Raman spectrum functionality

See merge request !238

GPUMD v5.5 in CI and fix scheduled pipeline

See merge request !239

Bumped version to 3.4

See merge request !241

Updated links to ase documentation

See merge request !242

Bug/fix read model behavior for missing version line

See merge request !243

Raise clear ValueError for truncated/malformed nep.txt parameter counts

See merge request !245

Fixed bug in CPUNEP implementation of qNEP

See merge request !244

Add CALORINE_GPUMD_COMMAND environment variable for GPUNEP

See merge request !246

Preserve qNEP charge_mode (1 vs 2) through Model read/write roundtrip

See merge request !247

Fixed behavior

See merge request !248

Added function for fast prediction via nep batching

See merge request !249

Resolve "New functionality not showing in reference part of docs"

Closes #141

See merge request !252

Added functionality for reordering species in models

See merge request !250

Fix rank-2 tensor component-ordering behavior and unify Voigt-conversion handling

See merge request !251

Fix negative volume for triclinic cells in read_thermodynamic_data

Closes #140

See merge request !254

Fixed calculation of norm for per-atom rank-2 tensors in get_parity_data

See merge request !255

change from whitelist to blacklist in fine_tuning_nep89

Closes #135

See merge request !253

Fix flaky GPUNEP MD tests

See merge request !256

Route dipole-gradient `nep` invocations through CALORINE_NEP_COMMAND

See merge request !257

Add mode=fixed to setup_training for pre-defined train/test splits

See merge request !258

Fixed the handling of type-wise cutoff factors

Closes #138

See merge request !260

Close remaining coverage gaps

See merge request !261

Restructured tutorial suite

See merge request !259

Review documentation

See merge request !262

Release 3.5

See merge request !263