jchwenger's picture
Upload 351 files (#2)
d9272c6 verified
# Copyright (c) 2022, salesforce.com, inc.
# All rights reserved.
# SPDX-License-Identifier: BSD-3-Clause
# For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
model:
arch: gpt_dialogue
model_type: base
datasets:
avsd_dialogue: # name of the dataset builder
vis_processor:
train:
name: "gpt_video_ft"
visual_ft: ["i3d_flow", "i3d_rgb"]
audio_ft: ["vggish"]
eval:
name: "gpt_video_ft"
visual_ft: ["i3d_flow", "i3d_rgb"]
audio_ft: ["vggish"]
text_processor:
train:
name: "gpt_dialogue"
max_turns: 3
use_caption: True
eval:
name: "gpt_dialogue"
max_turns: 3
use_caption: True
run:
task: dialogue
# optimizer
lr_sched: "linear_warmup_cosine_lr"
init_lr: 1e-5
min_lr: 0
weight_decay: 0.05
max_epoch: 20
batch_size_train: 16
batch_size_eval: 16
num_workers: 0
max_len: 20
min_len: 5
num_beams: 5
seed: 42
output_dir: "output/gpt2/dialogue_avsd"
amp: False
resume_ckpt_path: null
evaluate: False
train_splits: ["train"]
valid_splits: ["val"]
device: "cuda"
world_size: 1
dist_url: "env://"
distributed: True