IDEA-AI4S/mol-vicuna-7b-v1.5-all-lora-ep5

sft

This model is a fine-tuned version of lmsys/vicuna-7b-v1.5 on the uspto_1k_tpl, the smol_270k_i2s, the smol_270k_s2i, the smol_270k_i2f, the smol_270k_f2i, the rcr_57k_rp, the rcr_10k_cp, the rcr_70k_sp, the molinst_125k_rs, the molinst_125k_fs, the bh_sm_10k_yr and the hte_4k_ras 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: 2e-05
• train_batch_size: 16
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• num_epochs: 5.0

Training results

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.16.1
• Tokenizers 0.15.0