Hugging Face's logo

cedpsam
/
EleutherAI_gpt-neo-125M-stablediffionprompts

Text Generation
Transformers
PyTorch
TensorBoard
Safetensors
gpt_neo
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
2
cedpsam commited on
Commit
a9ac737
•
1 Parent(s): 1d2c498

Training in progress, step 500

Browse files
Files changed (22) hide show
  1. last-checkpoint/config.json +54 -0
  2. last-checkpoint/optimizer.pt +3 -0
  3. last-checkpoint/pytorch_model.bin +3 -0
  4. last-checkpoint/rng_state.pth +3 -0
  5. last-checkpoint/scaler.pt +3 -0
  6. last-checkpoint/scheduler.pt +3 -0
  7. last-checkpoint/trainer_state.json +22 -0
  8. last-checkpoint/training_args.bin +3 -0
  9. pytorch_model.bin +1 -1
  10. runs/Oct20_17-51-04_5425517b47a3/1666288481.6889434/events.out.tfevents.1666288481.5425517b47a3.23.1 +3 -0
  11. runs/Oct20_17-51-04_5425517b47a3/1666288495.935108/events.out.tfevents.1666288495.5425517b47a3.23.2 +3 -0
  12. runs/Oct20_17-51-04_5425517b47a3/1666288500.2544196/events.out.tfevents.1666288500.5425517b47a3.23.3 +3 -0
  13. runs/Oct20_17-51-04_5425517b47a3/1666288502.2599976/events.out.tfevents.1666288502.5425517b47a3.23.4 +3 -0
  14. runs/Oct20_17-51-04_5425517b47a3/1666288503.689416/events.out.tfevents.1666288503.5425517b47a3.23.5 +3 -0
  15. runs/Oct20_17-51-04_5425517b47a3/1666288504.403424/events.out.tfevents.1666288504.5425517b47a3.23.6 +3 -0
  16. runs/Oct20_17-51-04_5425517b47a3/1666288505.0064352/events.out.tfevents.1666288505.5425517b47a3.23.7 +3 -0
  17. runs/Oct20_17-51-04_5425517b47a3/1666288505.4571786/events.out.tfevents.1666288505.5425517b47a3.23.8 +3 -0
  18. runs/Oct20_17-51-04_5425517b47a3/1666288506.2575643/events.out.tfevents.1666288506.5425517b47a3.23.9 +3 -0
  19. runs/Oct20_17-51-04_5425517b47a3/1666288507.0446784/events.out.tfevents.1666288507.5425517b47a3.23.10 +3 -0
  20. runs/Oct20_17-51-04_5425517b47a3/1666288508.4595587/events.out.tfevents.1666288508.5425517b47a3.23.11 +3 -0
  21. runs/Oct20_17-51-04_5425517b47a3/events.out.tfevents.1666288481.5425517b47a3.23.0 +3 -0
  22. training_args.bin +1 -1
last-checkpoint/config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "cedpsam/EleutherAI_gpt-neo-125M-stablediffionprompts",
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
+ "embed_dropout": 0,
33
+ "eos_token_id": 50256,
34
+ "gradient_checkpointing": false,
35
+ "hidden_size": 768,
36
+ "initializer_range": 0.02,
37
+ "intermediate_size": null,
38
+ "layer_norm_epsilon": 1e-05,
39
+ "max_position_embeddings": 2048,
40
+ "model_type": "gpt_neo",
41
+ "num_heads": 12,
42
+ "num_layers": 12,
43
+ "resid_dropout": 0,
44
+ "summary_activation": null,
45
+ "summary_first_dropout": 0.1,
46
+ "summary_proj_to_labels": true,
47
+ "summary_type": "cls_index",
48
+ "summary_use_proj": true,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.20.1",
51
+ "use_cache": true,
52
+ "vocab_size": 50257,
53
+ "window_size": 256
54
+ }
last-checkpoint/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:182476128fc820cf47fcd7097ceb6ad26aa2ea20ab4f2e1c78a441c35f086223
3
+ size 1001681601
last-checkpoint/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:611d2fc552ff584d3a9df83d55f896f1560cee17cf8833db5e465ec45870df91
3
+ size 551185105
last-checkpoint/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62851f63d26f31f0098aaf22bfc590da9d5dd492f2a149b0c502c13be7100a3d
3
+ size 14503
last-checkpoint/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3978f33a9eccdea52f77a26979e6a0077a50cd239694fba6bc25cd0891e8edf1
3
+ size 559
last-checkpoint/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4af5659a4ffafd6e5e4486d3da6802b3d89fc41fd3462038b6768c0dd81c78f7
3
+ size 623
last-checkpoint/trainer_state.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.006782603977318972,
5
+ "global_step": 500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 9.886136363636363e-08,
13
+ "loss": 0.0253,
14
+ "step": 500
15
+ }
16
+ ],
17
+ "max_steps": 44000,
18
+ "num_train_epochs": 1,
19
+ "total_flos": 524503851466752.0,
20
+ "trial_name": null,
21
+ "trial_params": null
22
+ }
last-checkpoint/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a991bd9fa9f78a109c37f34127e236c70674b999767dd09c9c452dae8fb567ad
3
+ size 3439
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d92f649cc1be09c25d1de2eac713bc5fe2a01c10562d1c55cb2c1e2b0a4fe777
3
  size 551185105
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:611d2fc552ff584d3a9df83d55f896f1560cee17cf8833db5e465ec45870df91
3
  size 551185105
runs/Oct20_17-51-04_5425517b47a3/1666288481.6889434/events.out.tfevents.1666288481.5425517b47a3.23.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a548e18785771023413eccea4d8306067677c232683db99ffee8fa730300de6
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288495.935108/events.out.tfevents.1666288495.5425517b47a3.23.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51872cb717074dc6277d65c7e3b7d1743de5dc438971a29aaadb16d7e5cde075
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288500.2544196/events.out.tfevents.1666288500.5425517b47a3.23.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be0836ccbba8173facd92df821fb9f66ccdec1f975f7f9e6528871c6f7b90d58
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288502.2599976/events.out.tfevents.1666288502.5425517b47a3.23.4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1eff920fdfa9d43c4b02407a676ee8f913eea473ac401096d15a14def8c7ed4b
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288503.689416/events.out.tfevents.1666288503.5425517b47a3.23.5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f753ab8ab21c8dc1e6c6b6fcc1aa74f5eda8dfd41ee1486c6e888d6cb4e627a6
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288504.403424/events.out.tfevents.1666288504.5425517b47a3.23.6 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:262861295ae9d64cc284abd5db09acaf71d9d5533c81faa9bd4240adf26b5244
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288505.0064352/events.out.tfevents.1666288505.5425517b47a3.23.7 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a592537eee78cb504da0f07f15265d1f4ad2f127154c61792d2aed2a2740809b
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288505.4571786/events.out.tfevents.1666288505.5425517b47a3.23.8 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:589bfbb981fe4fbfc1e053570a7b594d6b5ef8c2a1dbd42f6b7836f16916359c
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288506.2575643/events.out.tfevents.1666288506.5425517b47a3.23.9 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91288529972f12f44bf91db1be3c19332059222e27c1293027f644b47fa74020
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288507.0446784/events.out.tfevents.1666288507.5425517b47a3.23.10 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:501cd7f1781338eba7b8f639a215bb2ba842fe7a3829a47a1b04770acb0d7d61
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/1666288508.4595587/events.out.tfevents.1666288508.5425517b47a3.23.11 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0092dc8d5db84008196f5cc527071806675dbc49cb73563c3efbadb806f4160
3
+ size 5438
runs/Oct20_17-51-04_5425517b47a3/events.out.tfevents.1666288481.5425517b47a3.23.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ffde6f248cdb4d923b90bfc0c723b57e8b8abb70779a8a30d19cc509046d8c70
3
+ size 45803
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e3c08e55cf4fa548ab6ad1d0976187b3df878db7839f630f4040317081083107
3
  size 3439
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a991bd9fa9f78a109c37f34127e236c70674b999767dd09c9c452dae8fb567ad
3
  size 3439