Update README.md

README.md

@@ -4,9 +4,18 @@ tags: []
 ---
 
 # Model Card for Model ID
+glm2_large_decoder_bgc_16k_percdsmask_10ep_Feb14
 
-<!-- Provide a quick summary of what the model is/does. -->
+command: finetune.py --model_config=configs/base_large_decode_bgc.yaml --dataset_path=/media/16TBNVME/datasets/hf/bgc_16384 --exp_name=glm2_large_decoder_bgc_16k_percdsmask_10ep
+Distributed environment: MULTI_GPU  Backend: nccl
+Num processes: 8
+Process index: 0
+Local process index: 0
+Device: cuda:0
 
+Mixed precision type: bf16
+
+Training/evaluation parameters AccelerateTrainingArguments(do_train=True, do_eval=True, mask_prob=1.0, mask_prob_uniform_sample=True, mask_mixed_strategy=False, reweight_loss=False, finetune_model_name='tattabio/gLM2_650M_decoder_v0', per_device_train_batch_size=1, per_device_eval_batch_size=1, learning_rate=4e-05, weight_decay=0.1, adamw_beta_1=0.9, adamw_beta_2=0.95, num_train_epochs=4, max_train_steps=None, max_eval_steps=25, validation_split_fraction=0.02, gradient_accumulation_steps=8, lr_scheduler_type='cosine', num_warmup_steps=5000, output_dir='/media/16TBNVME/dev_models', seed=46, with_tracking=True, checkpointing_steps=24000, save_model_steps='epoch', report_to='tensorboard', eval_steps=1000, logging_steps=10, dataloader_num_workers=8, max_grad_norm=1.0)
 
 
 ## Model Details