Skip to content

Commit ca8e503

Browse files
committed
Ugh, lost some training runs because files got saved to the wrong place
1 parent 02f0706 commit ca8e503

File tree

4 files changed

+12
-12
lines changed

4 files changed

+12
-12
lines changed

olmocr/train/configs/qwen25_vl_b100_x1_day2_json.yaml renamed to olmocr/train/configs/qwen25_vl_b100_x1_day3_json.yaml

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Project metadata
44
project_name: olmocr-qwen-vl-training
5-
run_name: qwen2.5-vl-7b-finetune-day2-json
5+
run_name: qwen2.5-vl-7b-finetune-day3-json
66

77
# Model configuration
88
model:
@@ -58,7 +58,7 @@ dataset:
5858

5959
# Training configuration
6060
training:
61-
output_dir: /home/ubuntu/olmocr-trainer/
61+
output_dir: /weka/oe-data-default/jakep/olmocr-trainer/
6262
num_train_epochs: 1
6363

6464
# Batch size and accumulation
@@ -72,7 +72,7 @@ training:
7272

7373
# Learning rate
7474
learning_rate: 2e-5
75-
lr_scheduler_type: cosine
75+
lr_scheduler_type: linear
7676
warmup_ratio: 0.1
7777

7878
# Optimization

olmocr/train/configs/qwen25_vl_b100_x1_day2_json_1280.yaml renamed to olmocr/train/configs/qwen25_vl_b100_x1_day3_json_1280.yaml

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Project metadata
44
project_name: olmocr-qwen-vl-training
5-
run_name: qwen2.5-vl-7b-finetune-day2-1280
5+
run_name: qwen2.5-vl-7b-finetune-day3-1280
66

77
# Model configuration
88
model:
@@ -58,7 +58,7 @@ dataset:
5858

5959
# Training configuration
6060
training:
61-
output_dir: /home/ubuntu/olmocr-trainer/
61+
output_dir: /weka/oe-data-default/jakep/olmocr-trainer/
6262
num_train_epochs: 1
6363

6464
# Batch size and accumulation
@@ -72,7 +72,7 @@ training:
7272

7373
# Learning rate
7474
learning_rate: 2e-5
75-
lr_scheduler_type: cosine
75+
lr_scheduler_type: linear
7676
warmup_ratio: 0.1
7777

7878
# Optimization

olmocr/train/configs/qwen25_vl_b100_x1_day2_json_1280_noanchor.yaml renamed to olmocr/train/configs/qwen25_vl_b100_x1_day3_json_1280_noanchor.yaml

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Project metadata
44
project_name: olmocr-qwen-vl-training
5-
run_name: qwen2.5-vl-7b-finetune-day2-1280-noanchor
5+
run_name: qwen2.5-vl-7b-finetune-day3-1280-noanchor
66

77
# Model configuration
88
model:
@@ -58,7 +58,7 @@ dataset:
5858

5959
# Training configuration
6060
training:
61-
output_dir: /home/ubuntu/olmocr-trainer/
61+
output_dir: /weka/oe-data-default/jakep/olmocr-trainer/
6262
num_train_epochs: 1
6363

6464
# Batch size and accumulation
@@ -72,7 +72,7 @@ training:
7272

7373
# Learning rate
7474
learning_rate: 2e-5
75-
lr_scheduler_type: cosine
75+
lr_scheduler_type: linear
7676
warmup_ratio: 0.1
7777

7878
# Optimization

olmocr/train/configs/qwen25_vl_b100_x1_day2_json_1600.yaml renamed to olmocr/train/configs/qwen25_vl_b100_x1_day3_json_1600.yaml

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
# Project metadata
44
project_name: olmocr-qwen-vl-training
5-
run_name: qwen2.5-vl-7b-finetune-day2-1600
5+
run_name: qwen2.5-vl-7b-finetune-day3-1600
66

77
# Model configuration
88
model:
@@ -58,7 +58,7 @@ dataset:
5858

5959
# Training configuration
6060
training:
61-
output_dir: /home/ubuntu/olmocr-trainer/
61+
output_dir: /weka/oe-data-default/jakep/olmocr-trainer/
6262
num_train_epochs: 1
6363

6464
# Batch size and accumulation
@@ -72,7 +72,7 @@ training:
7272

7373
# Learning rate
7474
learning_rate: 2e-5
75-
lr_scheduler_type: cosine
75+
lr_scheduler_type: linear
7676
warmup_ratio: 0.1
7777

7878
# Optimization

0 commit comments

Comments
 (0)