SpectraLLM_Pretrain_second_stage
This model is a fine-tuned version of /tos-bjml-ai4chem/chemshare/home/shuaikeshen/project/bk/LLaMA-Factory/saves/SpectraLLM_Pretrain/ on the pubchem_pretrain, the nmrbank_pretrain, the qm9s_pretrain, the hnmr_pretrain, the ms_fragment_pos_pretrain, the ms_neg_40ev_pretrain, the ms_pos_10ev_pretrain and the ms_pos_20ev_pretrain datasets.
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 32
- gradient_accumulation_steps: 8
- total_train_batch_size: 2048
- total_eval_batch_size: 256
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1.0
Training results
Framework versions
- Transformers 4.51.3
- Pytorch 2.7.0+cu126
- Datasets 3.5.0
- Tokenizers 0.21.1