Training in progress, step 500
Browse files- config.json +9 -8
- model.safetensors +2 -2
- runs/Dec11_15-46-18_1639a7b6a09d/events.out.tfevents.1702309612.1639a7b6a09d.586.0 +3 -0
- runs/Dec11_15-46-18_1639a7b6a09d/events.out.tfevents.1702309630.1639a7b6a09d.586.1 +3 -0
- runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702309970.1639a7b6a09d.586.2 +3 -0
- runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310030.1639a7b6a09d.586.3 +3 -0
- runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310051.1639a7b6a09d.586.4 +3 -0
- runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310069.1639a7b6a09d.586.5 +3 -0
- runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310100.1639a7b6a09d.586.6 +3 -0
- runs/Dec11_15-55-13_1639a7b6a09d/events.out.tfevents.1702310119.1639a7b6a09d.586.7 +3 -0
- runs/Dec11_15-55-13_1639a7b6a09d/events.out.tfevents.1702310144.1639a7b6a09d.586.8 +3 -0
- runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310161.1639a7b6a09d.586.9 +3 -0
- runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310195.1639a7b6a09d.586.10 +3 -0
- runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310251.1639a7b6a09d.586.11 +3 -0
- runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310264.1639a7b6a09d.586.12 +3 -0
- runs/Dec11_16-00-23_1639a7b6a09d/events.out.tfevents.1702310429.1639a7b6a09d.586.13 +3 -0
- runs/Dec11_16-01-59_1639a7b6a09d/events.out.tfevents.1702310521.1639a7b6a09d.586.14 +3 -0
- runs/Dec11_16-03-01_1639a7b6a09d/events.out.tfevents.1702310592.1639a7b6a09d.586.15 +3 -0
- runs/Dec11_16-04-25_1639a7b6a09d/events.out.tfevents.1702310670.1639a7b6a09d.586.16 +3 -0
- runs/Dec11_16-07-40_1639a7b6a09d/events.out.tfevents.1702310864.1639a7b6a09d.586.17 +3 -0
- runs/Dec11_16-07-40_1639a7b6a09d/events.out.tfevents.1702310883.1639a7b6a09d.586.18 +3 -0
- runs/Dec11_16-08-39_1639a7b6a09d/events.out.tfevents.1702310925.1639a7b6a09d.586.19 +3 -0
- runs/Dec11_16-10-32_1639a7b6a09d/events.out.tfevents.1702311036.1639a7b6a09d.586.20 +3 -0
- runs/Dec11_16-15-09_1639a7b6a09d/events.out.tfevents.1702311317.1639a7b6a09d.586.21 +3 -0
- runs/Dec11_16-21-07_1639a7b6a09d/events.out.tfevents.1702311673.1639a7b6a09d.586.22 +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -4
- training_args.bin +1 -1
config.json
CHANGED
@@ -1,26 +1,26 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"T5ForConditionalGeneration"
|
5 |
],
|
6 |
"classifier_dropout": 0.0,
|
7 |
"d_ff": 2048,
|
8 |
"d_kv": 64,
|
9 |
-
"d_model":
|
10 |
"decoder_start_token_id": 0,
|
11 |
-
"dense_act_fn": "
|
12 |
"dropout_rate": 0.1,
|
13 |
"eos_token_id": 1,
|
14 |
-
"feed_forward_proj": "
|
15 |
"initializer_factor": 1.0,
|
16 |
"is_encoder_decoder": true,
|
17 |
-
"is_gated_act":
|
18 |
"layer_norm_epsilon": 1e-06,
|
19 |
"model_type": "t5",
|
20 |
"n_positions": 512,
|
21 |
-
"num_decoder_layers":
|
22 |
-
"num_heads":
|
23 |
-
"num_layers":
|
24 |
"output_past": true,
|
25 |
"pad_token_id": 0,
|
26 |
"relative_attention_max_distance": 128,
|
@@ -54,6 +54,7 @@
|
|
54 |
"prefix": "translate English to Romanian: "
|
55 |
}
|
56 |
},
|
|
|
57 |
"torch_dtype": "float32",
|
58 |
"transformers_version": "4.35.2",
|
59 |
"use_cache": true,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "google/flan-t5-base",
|
3 |
"architectures": [
|
4 |
"T5ForConditionalGeneration"
|
5 |
],
|
6 |
"classifier_dropout": 0.0,
|
7 |
"d_ff": 2048,
|
8 |
"d_kv": 64,
|
9 |
+
"d_model": 768,
|
10 |
"decoder_start_token_id": 0,
|
11 |
+
"dense_act_fn": "gelu_new",
|
12 |
"dropout_rate": 0.1,
|
13 |
"eos_token_id": 1,
|
14 |
+
"feed_forward_proj": "gated-gelu",
|
15 |
"initializer_factor": 1.0,
|
16 |
"is_encoder_decoder": true,
|
17 |
+
"is_gated_act": true,
|
18 |
"layer_norm_epsilon": 1e-06,
|
19 |
"model_type": "t5",
|
20 |
"n_positions": 512,
|
21 |
+
"num_decoder_layers": 12,
|
22 |
+
"num_heads": 12,
|
23 |
+
"num_layers": 12,
|
24 |
"output_past": true,
|
25 |
"pad_token_id": 0,
|
26 |
"relative_attention_max_distance": 128,
|
|
|
54 |
"prefix": "translate English to Romanian: "
|
55 |
}
|
56 |
},
|
57 |
+
"tie_word_embeddings": false,
|
58 |
"torch_dtype": "float32",
|
59 |
"transformers_version": "4.35.2",
|
60 |
"use_cache": true,
|
model.safetensors
CHANGED
@@ -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 sha256:81836dcc3753731d286092e763646bb690d01e1758ef413c0f42a802bb399a0d
|
3 |
+
size 990345064
|
runs/Dec11_15-46-18_1639a7b6a09d/events.out.tfevents.1702309612.1639a7b6a09d.586.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:31657bc229484a7b784be689922119a6809d8d39aedcee50cb63ed0cf59feb8c
|
3 |
+
size 5239
|
runs/Dec11_15-46-18_1639a7b6a09d/events.out.tfevents.1702309630.1639a7b6a09d.586.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c7bc4b896a43bfd3419e5ae709cf88738b9e7ffe632133b6b6b330355d0e3244
|
3 |
+
size 7043
|
runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702309970.1639a7b6a09d.586.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f2ff2f4713f779e7d36aaf5ea2deb6b94449e63f66852e0ee3dca451f4cef77e
|
3 |
+
size 7043
|
runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310030.1639a7b6a09d.586.3
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:44dcc8f61a6034f17fd8d991ccf4982c5de6d18e6b9de720efcd9783988e7a5d
|
3 |
+
size 5239
|
runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310051.1639a7b6a09d.586.4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8f2e9b98ba75d790601b77d80caeae16d0dcd71a80575a12fb8f205c3445003e
|
3 |
+
size 5239
|
runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310069.1639a7b6a09d.586.5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2ce084dc294f5fae1f9ca4151a9b100e000c11a89d497599dbeafe4346b805ee
|
3 |
+
size 5239
|
runs/Dec11_15-50-04_1639a7b6a09d/events.out.tfevents.1702310100.1639a7b6a09d.586.6
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:68053d001881afbbb64b04da3c3bb1b74508e14359edb4a30ac216d05fe2e0d4
|
3 |
+
size 5239
|
runs/Dec11_15-55-13_1639a7b6a09d/events.out.tfevents.1702310119.1639a7b6a09d.586.7
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:94e0828898955a3217bf4880f6ebcd12dfc387c21c130d5a10bd0274e677d24f
|
3 |
+
size 5239
|
runs/Dec11_15-55-13_1639a7b6a09d/events.out.tfevents.1702310144.1639a7b6a09d.586.8
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1a5766bdafcf0be35c8f279e99cb61821ed87656eb2f2fab1f486303a716eb3c
|
3 |
+
size 5239
|
runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310161.1639a7b6a09d.586.9
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:59d3493400dc2caaac35986d3b76aad0382ab38db86492b354ec788aa14b2957
|
3 |
+
size 5239
|
runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310195.1639a7b6a09d.586.10
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c750f9e3dc4de352233a974f942bb4f76b970c3e55a8695923a70d5eef92382c
|
3 |
+
size 7043
|
runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310251.1639a7b6a09d.586.11
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e96e22816a5abec05533f8023de3c9dfe8d8ef576b46abc09d190502fe56193b
|
3 |
+
size 7043
|
runs/Dec11_15-55-57_1639a7b6a09d/events.out.tfevents.1702310264.1639a7b6a09d.586.12
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2b4c2e15edf16b709d1b368a3d3494bc4e74935a74585d3c038692fadca9140c
|
3 |
+
size 7043
|
runs/Dec11_16-00-23_1639a7b6a09d/events.out.tfevents.1702310429.1639a7b6a09d.586.13
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bd6b80c9c62fad3683483f47ccc67aa86641c7e55ec26bcbbbc9d6c87a692004
|
3 |
+
size 7043
|
runs/Dec11_16-01-59_1639a7b6a09d/events.out.tfevents.1702310521.1639a7b6a09d.586.14
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a8d1ead661c12ed8d043d9dbfe0a9007db9d2673c121244a6102e58635579df3
|
3 |
+
size 7043
|
runs/Dec11_16-03-01_1639a7b6a09d/events.out.tfevents.1702310592.1639a7b6a09d.586.15
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:240c78fd137624f18781485056698b0d67430a0c3d33429c0b18c5cac8e9d52e
|
3 |
+
size 7043
|
runs/Dec11_16-04-25_1639a7b6a09d/events.out.tfevents.1702310670.1639a7b6a09d.586.16
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e65894b54f85c38d4c9d8e0000e3a2ee1b3afdcaac15826a504e9e2a4ed9d660
|
3 |
+
size 7043
|
runs/Dec11_16-07-40_1639a7b6a09d/events.out.tfevents.1702310864.1639a7b6a09d.586.17
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4a22bc3ab415d05fcee80663f57fd294361c77c25c62025720cc7da0e53628a1
|
3 |
+
size 7043
|
runs/Dec11_16-07-40_1639a7b6a09d/events.out.tfevents.1702310883.1639a7b6a09d.586.18
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4b4b31a5fdc558009a5ee0783f6ae67e4a55ae3cc54cc3f8929ac3187be97d51
|
3 |
+
size 7043
|
runs/Dec11_16-08-39_1639a7b6a09d/events.out.tfevents.1702310925.1639a7b6a09d.586.19
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7969d5f7dbe39c8749daa13e8501d4a0184a6515e838f54d5deb8bc4b5f190c6
|
3 |
+
size 7043
|
runs/Dec11_16-10-32_1639a7b6a09d/events.out.tfevents.1702311036.1639a7b6a09d.586.20
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dbb59b27ed689fd12d12f24cc2f104236d8b67c3fb92d073918d855e2f1778bc
|
3 |
+
size 9956
|
runs/Dec11_16-15-09_1639a7b6a09d/events.out.tfevents.1702311317.1639a7b6a09d.586.21
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:62035391bc8306d5ce61bbfba0d38a47dce0829350c62dc03e667a6e730ac9f3
|
3 |
+
size 9963
|
runs/Dec11_16-21-07_1639a7b6a09d/events.out.tfevents.1702311673.1639a7b6a09d.586.22
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:89cf03838a9623b7378a4dbc0db702519d6776d64e645e7d4d1787984e5c5edf
|
3 |
+
size 5403
|
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
CHANGED
@@ -930,12 +930,9 @@
|
|
930 |
"clean_up_tokenization_spaces": true,
|
931 |
"eos_token": "</s>",
|
932 |
"extra_ids": 100,
|
933 |
-
"max_length": 256,
|
934 |
"model_max_length": 512,
|
935 |
"pad_token": "<pad>",
|
936 |
-
"
|
937 |
"tokenizer_class": "T5Tokenizer",
|
938 |
-
"truncation_side": "right",
|
939 |
-
"truncation_strategy": "longest_first",
|
940 |
"unk_token": "<unk>"
|
941 |
}
|
|
|
930 |
"clean_up_tokenization_spaces": true,
|
931 |
"eos_token": "</s>",
|
932 |
"extra_ids": 100,
|
|
|
933 |
"model_max_length": 512,
|
934 |
"pad_token": "<pad>",
|
935 |
+
"sp_model_kwargs": {},
|
936 |
"tokenizer_class": "T5Tokenizer",
|
|
|
|
|
937 |
"unk_token": "<unk>"
|
938 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4728
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f81b08d125067c555d937938bab03ddbafbc9d7761f647f9460d9dd22255f981
|
3 |
size 4728
|