Saving weights and logs of step 2500
Browse files- config.json +1 -1
- events.out.tfevents.1638631417.t1v-n-d4d6e0cd-w-0.542156.0.v2 +3 -0
- flax_model.msgpack +1 -1
- run_oscar.sh +2 -1
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": "4.11.0.dev0",
|
| 25 |
"type_vocab_size": 1,
|
| 26 |
"use_cache": true,
|
| 27 |
"vocab_size": 50265
|
events.out.tfevents.1638631417.t1v-n-d4d6e0cd-w-0.542156.0.v2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:521b740bed8c2132d8c653b414b2c975cc0ed1f7885133830775400e26ec64c5
|
| 3 |
+
size 367912
|
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:ce01861ebbc1c824d1302b035d0b8da42e16e8e61782e64fb9203f44db39a6b5
|
| 3 |
size 498796983
|
run_oscar.sh
CHANGED
|
@@ -3,7 +3,8 @@
|
|
| 3 |
--model_type="roberta" \
|
| 4 |
--config_name="./" \
|
| 5 |
--tokenizer_name="./" \
|
| 6 |
-
--dataset_name="
|
|
|
|
| 7 |
--max_seq_length="128" \
|
| 8 |
--weight_decay="0.01" \
|
| 9 |
--per_device_train_batch_size="128" \
|
|
|
|
| 3 |
--model_type="roberta" \
|
| 4 |
--config_name="./" \
|
| 5 |
--tokenizer_name="./" \
|
| 6 |
+
--dataset_name="oscar" \
|
| 7 |
+
--dataset_config_name="unshuffled_deduplicated_no" \
|
| 8 |
--max_seq_length="128" \
|
| 9 |
--weight_decay="0.01" \
|
| 10 |
--per_device_train_batch_size="128" \
|