From d80ac2fd2df4e7fb8a28acfa512bb11472b5cc99 Mon Sep 17 00:00:00 2001
From: liugz18 <57401541+liugz18@users.noreply.github.com>
Date: 星期四, 18 七月 2024 21:34:55 +0800
Subject: [PATCH] Rename 'res' in line 514 to avoid with naming conflict with line 365

---
 examples/industrial_data_pretraining/llm_asr/conf/whisper_vicuna_linear.yaml |   12 ++++++------
 1 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/examples/industrial_data_pretraining/llm_asr/conf/whisper_vicuna_linear.yaml b/examples/industrial_data_pretraining/llm_asr/conf/whisper_vicuna_linear.yaml
index eacbd35..f4f8c01 100644
--- a/examples/industrial_data_pretraining/llm_asr/conf/whisper_vicuna_linear.yaml
+++ b/examples/industrial_data_pretraining/llm_asr/conf/whisper_vicuna_linear.yaml
@@ -58,14 +58,15 @@
 train_conf:
   accum_grad: 1
   grad_clip: 5
-  max_epoch: 150
+  max_epoch: 15
   keep_nbest_models: 10
   log_interval: 10
 
 optim: adamw
 optim_conf:
    lr: 0.0001
-   weight_decay: 0.000001
+   weight_decay: 0
+
 scheduler: warmuplr
 scheduler_conf:
    warmup_steps: 1500
@@ -73,11 +74,10 @@
 dataset: AudioLLMVicunaDataset
 dataset_conf:
     index_ds: IndexDSJsonl
-    batch_sampler: RankFullLocalShuffleBatchSampler
+    batch_sampler: CustomDistributedBatchSampler
     batch_type: example # example or length
-    batch_size: 8 # if batch_type is example, batch_size is the numbers of samples; if length, batch_size is source_token_len+target_token_len;
-    max_token_length: 2048 # filter samples if source_token_len+target_token_len > max_token_length,
-    buffer_size: 500
+    batch_size: 4 # if batch_type is example, batch_size is the numbers of samples; if length, batch_size is source_token_len+target_token_len;
+    max_token_length: 3000 # filter samples if source_token_len+target_token_len > max_token_length,
     shuffle: True
     num_workers: 4
 #    preprocessor_text: TextPreprocessRemovePunctuation

--
Gitblit v1.9.1