Fix max sequence padding

Use transformers trainer
tangst2014 · Jul 4, 2023 · a33cb4f · a33cb4f
1 parent b99e3d7
commit a33cb4f
Show file tree

Hide file tree

Showing 4 changed files with 20 additions and 3,780 deletions.
diff --git a/ptuning/evaluate.sh b/ptuning/evaluate.sh
@@ -10,7 +10,7 @@ torchrun --standalone --nnodes=1 --nproc-per-node=$NUM_GPUS main.py \
     --overwrite_cache \
     --prompt_column content \
     --response_column summary \
-    --model_name_or_path chatglm2-6b \
+    --model_name_or_path THUDM/chatglm2-6b \
     --ptuning_checkpoint ./output/$CHECKPOINT/checkpoint-$STEP \
     --output_dir ./output/$CHECKPOINT \
     --overwrite_output_dir \

diff --git a/ptuning/main.py b/ptuning/main.py
@@ -178,7 +178,7 @@ def preprocess_function_eval(examples):
         return model_inputs
 
     def preprocess_function_train(examples):
-        max_seq_length = data_args.max_source_length + data_args.max_target_length
+        max_seq_length = data_args.max_source_length + data_args.max_target_length + 1
 
         model_inputs = {
             "input_ids": [],
@@ -335,7 +335,7 @@ def compute_metrics(eval_preds):
         tokenizer=tokenizer,
         data_collator=data_collator,
         compute_metrics=compute_metrics if training_args.predict_with_generate else None,
-        save_prefixencoder=model_args.pre_seq_len is not None
+        save_changed=model_args.pre_seq_len is not None
     )
 
     # Training