Saving weights and logs of step 10000
Browse files
config.json
CHANGED
@@ -21,7 +21,7 @@
|
|
21 |
"pad_token_id": 1,
|
22 |
"position_embedding_type": "absolute",
|
23 |
"torch_dtype": "float32",
|
24 |
-
"transformers_version": "4.
|
25 |
"type_vocab_size": 1,
|
26 |
"use_cache": true,
|
27 |
"vocab_size": 50265
|
|
|
21 |
"pad_token_id": 1,
|
22 |
"position_embedding_type": "absolute",
|
23 |
"torch_dtype": "float32",
|
24 |
+
"transformers_version": "4.15.0.dev0",
|
25 |
"type_vocab_size": 1,
|
26 |
"use_cache": true,
|
27 |
"vocab_size": 50265
|
events.out.tfevents.1640261667.t1v-n-358ff5d1-w-0.1667932.3.v2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c6a8477d21766f937975208be48441ba54cdff6eb0cc792e7d1c0bca7734ffe4
|
3 |
+
size 40
|
events.out.tfevents.1640263998.t1v-n-358ff5d1-w-0.1676696.3.v2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d99dcbe152245c74123b25a17b0ecdba0a2d3c1e67121479130e1db38a20f5a8
|
3 |
+
size 40
|
events.out.tfevents.1640264276.t1v-n-358ff5d1-w-0.1678479.3.v2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5d30cf60bdf4528dd735af70ec9c2664d8b6e8b012b4989e6a115027abf7597
|
3 |
+
size 1470136
|
flax_model.msgpack
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8eb56d4bd0e72b32a965cdb0896990e2a6be3b5145b8514e0137f456ef5bf069
|
3 |
+
size 498796983
|
run_step1.sh
CHANGED
@@ -7,9 +7,9 @@
|
|
7 |
--cache_dir="/mnt/disks/flaxdisk/cache/" \
|
8 |
--max_seq_length="128" \
|
9 |
--weight_decay="0.01" \
|
10 |
-
--per_device_train_batch_size="
|
11 |
-
--per_device_eval_batch_size="
|
12 |
-
--learning_rate="
|
13 |
--warmup_steps="10000" \
|
14 |
--overwrite_output_dir \
|
15 |
--num_train_epochs="1000" \
|
|
|
7 |
--cache_dir="/mnt/disks/flaxdisk/cache/" \
|
8 |
--max_seq_length="128" \
|
9 |
--weight_decay="0.01" \
|
10 |
+
--per_device_train_batch_size="164" \
|
11 |
+
--per_device_eval_batch_size="164" \
|
12 |
+
--learning_rate="3e-4" \
|
13 |
--warmup_steps="10000" \
|
14 |
--overwrite_output_dir \
|
15 |
--num_train_epochs="1000" \
|