End of training
Browse files- .gitignore +1 -0
- config.json +55 -0
- generation_config.json +6 -0
- pytorch_model.bin +3 -0
- runs/Jun13_20-51-29_cc667b61cf5b/events.out.tfevents.1686689601.cc667b61cf5b.3782.0 +3 -0
- runs/Jun13_20-57-42_cc667b61cf5b/events.out.tfevents.1686689923.cc667b61cf5b.5766.0 +3 -0
- runs/Jun13_21-05-16_cc667b61cf5b/events.out.tfevents.1686690375.cc667b61cf5b.7978.0 +3 -0
- runs/Jun13_21-10-07_cc667b61cf5b/events.out.tfevents.1686690668.cc667b61cf5b.9240.0 +3 -0
- runs/Jun13_21-13-24_cc667b61cf5b/events.out.tfevents.1686690867.cc667b61cf5b.10555.0 +3 -0
- runs/Jun13_21-15-48_cc667b61cf5b/events.out.tfevents.1686691009.cc667b61cf5b.11452.0 +3 -0
- runs/Jun13_21-19-42_cc667b61cf5b/events.out.tfevents.1686691243.cc667b61cf5b.12699.0 +3 -0
- runs/Jun13_21-35-58_cc667b61cf5b/events.out.tfevents.1686692219.cc667b61cf5b.17487.0 +3 -0
- runs/Jun13_21-42-28_cc667b61cf5b/events.out.tfevents.1686692608.cc667b61cf5b.19387.0 +3 -0
- runs/Jun13_22-08-45_cc667b61cf5b/events.out.tfevents.1686694138.cc667b61cf5b.19387.1 +3 -0
- runs/Jun13_22-12-44_cc667b61cf5b/events.out.tfevents.1686694373.cc667b61cf5b.27358.0 +3 -0
- runs/Jun13_22-12-44_cc667b61cf5b/events.out.tfevents.1686694510.cc667b61cf5b.27358.1 +3 -0
- runs/Jun13_22-23-06_cc667b61cf5b/events.out.tfevents.1686695000.cc667b61cf5b.27358.2 +3 -0
- runs/Jun13_22-23-06_cc667b61cf5b/events.out.tfevents.1686695159.cc667b61cf5b.27358.3 +3 -0
- training_args.bin +3 -0
.gitignore
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
checkpoint-*/
|
config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/gpt-neo-125M",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPTNeoForCausalLM"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0,
|
| 8 |
+
"attention_layers": [
|
| 9 |
+
"global",
|
| 10 |
+
"local",
|
| 11 |
+
"global",
|
| 12 |
+
"local",
|
| 13 |
+
"global",
|
| 14 |
+
"local",
|
| 15 |
+
"global",
|
| 16 |
+
"local",
|
| 17 |
+
"global",
|
| 18 |
+
"local",
|
| 19 |
+
"global",
|
| 20 |
+
"local"
|
| 21 |
+
],
|
| 22 |
+
"attention_types": [
|
| 23 |
+
[
|
| 24 |
+
[
|
| 25 |
+
"global",
|
| 26 |
+
"local"
|
| 27 |
+
],
|
| 28 |
+
6
|
| 29 |
+
]
|
| 30 |
+
],
|
| 31 |
+
"bos_token_id": 50256,
|
| 32 |
+
"classifier_dropout": 0.1,
|
| 33 |
+
"embed_dropout": 0,
|
| 34 |
+
"eos_token_id": 50256,
|
| 35 |
+
"gradient_checkpointing": false,
|
| 36 |
+
"hidden_size": 768,
|
| 37 |
+
"initializer_range": 0.02,
|
| 38 |
+
"intermediate_size": null,
|
| 39 |
+
"layer_norm_epsilon": 1e-05,
|
| 40 |
+
"max_position_embeddings": 2048,
|
| 41 |
+
"model_type": "gpt_neo",
|
| 42 |
+
"num_heads": 12,
|
| 43 |
+
"num_layers": 12,
|
| 44 |
+
"resid_dropout": 0,
|
| 45 |
+
"summary_activation": null,
|
| 46 |
+
"summary_first_dropout": 0.1,
|
| 47 |
+
"summary_proj_to_labels": true,
|
| 48 |
+
"summary_type": "cls_index",
|
| 49 |
+
"summary_use_proj": true,
|
| 50 |
+
"torch_dtype": "float32",
|
| 51 |
+
"transformers_version": "4.30.2",
|
| 52 |
+
"use_cache": true,
|
| 53 |
+
"vocab_size": 50257,
|
| 54 |
+
"window_size": 256
|
| 55 |
+
}
|
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": "4.30.2"
|
| 6 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0df8d18dace4dd2226b4fd90e0837df20e9019641bc47bdfa4cff2aaaa8b254b
|
| 3 |
+
size 551186797
|
runs/Jun13_20-51-29_cc667b61cf5b/events.out.tfevents.1686689601.cc667b61cf5b.3782.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2380dfce57400631da503656137faeab679e019f286eac0c9fc8229bbe4e7839
|
| 3 |
+
size 4799
|
runs/Jun13_20-57-42_cc667b61cf5b/events.out.tfevents.1686689923.cc667b61cf5b.5766.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd6b53584484bae0e3bc9d9aa8b0c41c541d1d4d87f48f8c2b5c4a776705ebfd
|
| 3 |
+
size 4799
|
runs/Jun13_21-05-16_cc667b61cf5b/events.out.tfevents.1686690375.cc667b61cf5b.7978.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d4d923fe82e33d359da759964a247ba1b50885782c2e0bc0e8ff344d19d4e43c
|
| 3 |
+
size 4799
|
runs/Jun13_21-10-07_cc667b61cf5b/events.out.tfevents.1686690668.cc667b61cf5b.9240.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fcbb58849d1ba56593948016fd815d46cee37ecbe4303c6ec014a45061f66fdc
|
| 3 |
+
size 4184
|
runs/Jun13_21-13-24_cc667b61cf5b/events.out.tfevents.1686690867.cc667b61cf5b.10555.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:997ac43deec06815f77112749ad0c065cd75b92c6ef6d106bc3b18f0cdf781b6
|
| 3 |
+
size 4184
|
runs/Jun13_21-15-48_cc667b61cf5b/events.out.tfevents.1686691009.cc667b61cf5b.11452.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c209f6996021eb9627b139bc7520680e4531e052fb74765cc04a1a4046cc4911
|
| 3 |
+
size 4799
|
runs/Jun13_21-19-42_cc667b61cf5b/events.out.tfevents.1686691243.cc667b61cf5b.12699.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2ff6c8d907a3e13a38d406cef4775ce8a7b3dfa35f5fc902fed0e769986cf9f
|
| 3 |
+
size 4799
|
runs/Jun13_21-35-58_cc667b61cf5b/events.out.tfevents.1686692219.cc667b61cf5b.17487.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc1a0f006eb29c8a8d7b85301ffa2a883a7e6dbca3fa12154fec27cf22669c95
|
| 3 |
+
size 4799
|
runs/Jun13_21-42-28_cc667b61cf5b/events.out.tfevents.1686692608.cc667b61cf5b.19387.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4992a5b0a374c262db8ca3528eb7f9c5b04d8dadf3bdacd6821a648aeeb19996
|
| 3 |
+
size 9510
|
runs/Jun13_22-08-45_cc667b61cf5b/events.out.tfevents.1686694138.cc667b61cf5b.19387.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd59334ca13fe9b49442b138a54c72230784a6dc1a1360bc7cb161c95c034823
|
| 3 |
+
size 4799
|
runs/Jun13_22-12-44_cc667b61cf5b/events.out.tfevents.1686694373.cc667b61cf5b.27358.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c043744cbab85a302af0cd751cceb413d34d3d1d611101b09ac5eae25b5e0405
|
| 3 |
+
size 5530
|
runs/Jun13_22-12-44_cc667b61cf5b/events.out.tfevents.1686694510.cc667b61cf5b.27358.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:655ff7e8f2eb73442e5983e0dbb260cad9c93d1e4dbcef15903e188382116ca9
|
| 3 |
+
size 359
|
runs/Jun13_22-23-06_cc667b61cf5b/events.out.tfevents.1686695000.cc667b61cf5b.27358.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:194c98a466754d1480edb4aae6b842f8220841f06a06eaada0a784a2129cfb20
|
| 3 |
+
size 5626
|
runs/Jun13_22-23-06_cc667b61cf5b/events.out.tfevents.1686695159.cc667b61cf5b.27358.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70fddeb92935b0b0cc044cfa29a5d4d546f2419c06da1b5693a896005161e618
|
| 3 |
+
size 359
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a03e0e056b7e00ef33eab6d38ba1d3b7dfcb9aaf2d7e7950ef496fd8c2e0c5d0
|
| 3 |
+
size 3963
|