Saving weights and logs of step 1000
Browse files
events.out.tfevents.1642748650.t1v-n-ccbf3e94-w-0.1812628.3.v2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3c3cd9ef4f724939cb87ada1bc665da4c603f0c91c6bedb7e9639ac0358b64b3
|
3 |
+
size 147136
|
flax_model.msgpack
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 498796983
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4ae7c72c78e81790343787e4b04020484e00f38a63bdcf6b498ff96d83ca1596
|
3 |
size 498796983
|
run_128.sh
CHANGED
@@ -10,7 +10,7 @@ python run_mlm_flax.py \
|
|
10 |
--per_device_train_batch_size="232" \
|
11 |
--per_device_eval_batch_size="232" \
|
12 |
--pad_to_max_length \
|
13 |
-
--learning_rate="
|
14 |
--warmup_steps="10000" \
|
15 |
--overwrite_output_dir \
|
16 |
--num_train_epochs="10" \
|
|
|
10 |
--per_device_train_batch_size="232" \
|
11 |
--per_device_eval_batch_size="232" \
|
12 |
--pad_to_max_length \
|
13 |
+
--learning_rate="4e-4" \
|
14 |
--warmup_steps="10000" \
|
15 |
--overwrite_output_dir \
|
16 |
--num_train_epochs="10" \
|