Saving weights and logs of step 20000
Browse files
events.out.tfevents.1674498755.t1v-n-9758a16f-w-0.1006414.0.v2 → events.out.tfevents.1674544407.t1v-n-9758a16f-w-0.1066395.0.v2
RENAMED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f38f6d9fa18938fd0ba282b44d7440a8507ecea5217448ec00125ef5f8058f55
|
3 |
+
size 2950146
|
flax_model.msgpack
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1625384886
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5e88922a1080d678c723871d9177d422ba926f7810feb7781c4395336ee29a4
|
3 |
size 1625384886
|
start_train.sh
CHANGED
@@ -7,7 +7,7 @@ python run_bart_dlm_flax.py \
|
|
7 |
--max_seq_length="1024" \
|
8 |
--per_device_train_batch_size="4" \
|
9 |
--per_device_eval_batch_size="4" \
|
10 |
-
--learning_rate="4e-
|
11 |
--weight_decay="0.01" \
|
12 |
--warmup_steps="10000" \
|
13 |
--overwrite_output_dir \
|
|
|
7 |
--max_seq_length="1024" \
|
8 |
--per_device_train_batch_size="4" \
|
9 |
--per_device_eval_batch_size="4" \
|
10 |
+
--learning_rate="4e-5" \
|
11 |
--weight_decay="0.01" \
|
12 |
--warmup_steps="10000" \
|
13 |
--overwrite_output_dir \
|