End of training
Browse files- .gitignore +1 -0
- config.json +46 -0
- generation_config.json +6 -0
- pytorch_model.bin +3 -0
- runs/Jun08_06-11-36_c7ad634d6d65/1686204707.8237867/events.out.tfevents.1686204707.c7ad634d6d65.2893.1 +3 -0
- runs/Jun08_06-11-36_c7ad634d6d65/events.out.tfevents.1686204707.c7ad634d6d65.2893.0 +3 -0
- runs/Jun08_06-12-16_c7ad634d6d65/1686204739.478012/events.out.tfevents.1686204739.c7ad634d6d65.2893.3 +3 -0
- runs/Jun08_06-12-16_c7ad634d6d65/events.out.tfevents.1686204739.c7ad634d6d65.2893.2 +3 -0
- runs/Jun08_06-12-36_c7ad634d6d65/1686204758.8222342/events.out.tfevents.1686204758.c7ad634d6d65.2893.5 +3 -0
- runs/Jun08_06-12-36_c7ad634d6d65/events.out.tfevents.1686204758.c7ad634d6d65.2893.4 +3 -0
- runs/Jun08_06-13-04_c7ad634d6d65/1686204787.371913/events.out.tfevents.1686204787.c7ad634d6d65.2893.7 +3 -0
- runs/Jun08_06-13-04_c7ad634d6d65/events.out.tfevents.1686204787.c7ad634d6d65.2893.6 +3 -0
- training_args.bin +3 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
checkpoint-*/
|
config.json
ADDED
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "distilgpt2",
|
3 |
+
"_num_labels": 1,
|
4 |
+
"activation_function": "gelu_new",
|
5 |
+
"architectures": [
|
6 |
+
"GPT2LMHeadModel"
|
7 |
+
],
|
8 |
+
"attn_pdrop": 0.1,
|
9 |
+
"bos_token_id": 50256,
|
10 |
+
"embd_pdrop": 0.1,
|
11 |
+
"eos_token_id": 50256,
|
12 |
+
"id2label": {
|
13 |
+
"0": "LABEL_0"
|
14 |
+
},
|
15 |
+
"initializer_range": 0.02,
|
16 |
+
"label2id": {
|
17 |
+
"LABEL_0": 0
|
18 |
+
},
|
19 |
+
"layer_norm_epsilon": 1e-05,
|
20 |
+
"model_type": "gpt2",
|
21 |
+
"n_ctx": 1024,
|
22 |
+
"n_embd": 768,
|
23 |
+
"n_head": 12,
|
24 |
+
"n_inner": null,
|
25 |
+
"n_layer": 6,
|
26 |
+
"n_positions": 1024,
|
27 |
+
"reorder_and_upcast_attn": false,
|
28 |
+
"resid_pdrop": 0.1,
|
29 |
+
"scale_attn_by_inverse_layer_idx": false,
|
30 |
+
"scale_attn_weights": true,
|
31 |
+
"summary_activation": null,
|
32 |
+
"summary_first_dropout": 0.1,
|
33 |
+
"summary_proj_to_labels": true,
|
34 |
+
"summary_type": "cls_index",
|
35 |
+
"summary_use_proj": true,
|
36 |
+
"task_specific_params": {
|
37 |
+
"text-generation": {
|
38 |
+
"do_sample": true,
|
39 |
+
"max_length": 50
|
40 |
+
}
|
41 |
+
},
|
42 |
+
"torch_dtype": "float32",
|
43 |
+
"transformers_version": "4.29.2",
|
44 |
+
"use_cache": true,
|
45 |
+
"vocab_size": 50257
|
46 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 50256,
|
4 |
+
"eos_token_id": 50256,
|
5 |
+
"transformers_version": "4.29.2"
|
6 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f331b8df33423e6f43475aade8f7d58ca51d11ac976598d70c800f0e33ebd3d3
|
3 |
+
size 333970169
|
runs/Jun08_06-11-36_c7ad634d6d65/1686204707.8237867/events.out.tfevents.1686204707.c7ad634d6d65.2893.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:67f1f5c1a10ec390be7874469de304685ea7b9bb22134a1593cb713b1d8cceea
|
3 |
+
size 5913
|
runs/Jun08_06-11-36_c7ad634d6d65/events.out.tfevents.1686204707.c7ad634d6d65.2893.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:347d9a942411086c33fb97dbc5200a400c37dc0606fed2658ddb505e66ab33dd
|
3 |
+
size 4184
|
runs/Jun08_06-12-16_c7ad634d6d65/1686204739.478012/events.out.tfevents.1686204739.c7ad634d6d65.2893.3
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:20d343356db4c11e8023a4b9d1423298b97a1207306d6365a8db7d77b450667e
|
3 |
+
size 5913
|
runs/Jun08_06-12-16_c7ad634d6d65/events.out.tfevents.1686204739.c7ad634d6d65.2893.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7d3a71c8b481f0e05c1de05d9d39e13a784364d2ce96be71790740dccf5d4193
|
3 |
+
size 4184
|
runs/Jun08_06-12-36_c7ad634d6d65/1686204758.8222342/events.out.tfevents.1686204758.c7ad634d6d65.2893.5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7187ef150ce4f4fb8e8027ff8aabc956cca34b89d569253403026026c0e5b777
|
3 |
+
size 5913
|
runs/Jun08_06-12-36_c7ad634d6d65/events.out.tfevents.1686204758.c7ad634d6d65.2893.4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f81b44ff74665f429ea3ce0d783fca3893011ecf651559868c53468c45723cce
|
3 |
+
size 4184
|
runs/Jun08_06-13-04_c7ad634d6d65/1686204787.371913/events.out.tfevents.1686204787.c7ad634d6d65.2893.7
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fd3979aec8dcb4a7c3bff8c15b34b407b13dde309b8eb21fdf42735487cc6603
|
3 |
+
size 5913
|
runs/Jun08_06-13-04_c7ad634d6d65/events.out.tfevents.1686204787.c7ad634d6d65.2893.6
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7145c28c72a479ae74d07333477800900eb808f48b22bdea51e94073c8ec46af
|
3 |
+
size 5519
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:76731c75ee3b09e562f9ee8877bbc9b2c47c87caf05cdfe1af8618d428c5805d
|
3 |
+
size 3899
|