Hugging Face's logo

theastronuts
/
sg_toxic_sentence_model

Text Generation
Transformers
Safetensors
gpt2
Generated from Trainer
text-generation-inference
Model card Files
xet
 Community
imda-lseokmin commited on
Commit
ed77c38
·
verified ·
1 Parent(s): 57a83d0

Upload 130 files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. all_results.json +6 -6
  2. checkpoint-1000/model.safetensors +1 -1
  3. checkpoint-1000/optimizer.pt +1 -1
  4. checkpoint-1000/rng_state.pth +1 -1
  5. checkpoint-1000/scheduler.pt +1 -1
  6. checkpoint-1000/trainer_state.json +11 -11
  7. checkpoint-1000/training_args.bin +1 -1
  8. checkpoint-1500/model.safetensors +1 -1
  9. checkpoint-1500/optimizer.pt +1 -1
  10. checkpoint-1500/rng_state.pth +1 -1
  11. checkpoint-1500/scheduler.pt +1 -1
  12. checkpoint-1500/trainer_state.json +15 -15
  13. checkpoint-1500/training_args.bin +1 -1
  14. checkpoint-2000/model.safetensors +1 -1
  15. checkpoint-2000/optimizer.pt +1 -1
  16. checkpoint-2000/rng_state.pth +1 -1
  17. checkpoint-2000/scheduler.pt +1 -1
  18. checkpoint-2000/trainer_state.json +19 -19
  19. checkpoint-2000/training_args.bin +1 -1
  20. checkpoint-2500/model.safetensors +1 -1
  21. checkpoint-2500/optimizer.pt +1 -1
  22. checkpoint-2500/rng_state.pth +1 -1
  23. checkpoint-2500/scheduler.pt +1 -1
  24. checkpoint-2500/trainer_state.json +23 -23
  25. checkpoint-2500/training_args.bin +1 -1
  26. checkpoint-3000/model.safetensors +1 -1
  27. checkpoint-3000/optimizer.pt +1 -1
  28. checkpoint-3000/rng_state.pth +1 -1
  29. checkpoint-3000/scheduler.pt +1 -1
  30. checkpoint-3000/trainer_state.json +27 -27
  31. checkpoint-3000/training_args.bin +1 -1
  32. checkpoint-3500/model.safetensors +1 -1
  33. checkpoint-3500/optimizer.pt +1 -1
  34. checkpoint-3500/rng_state.pth +1 -1
  35. checkpoint-3500/scheduler.pt +1 -1
  36. checkpoint-3500/trainer_state.json +31 -31
  37. checkpoint-3500/training_args.bin +1 -1
  38. checkpoint-4000/config.json +39 -0
  39. checkpoint-4000/generation_config.json +6 -0
  40. checkpoint-4000/merges.txt +0 -0
  41. checkpoint-4000/model.safetensors +3 -0
  42. checkpoint-4000/optimizer.pt +3 -0
  43. checkpoint-4000/rng_state.pth +3 -0
  44. checkpoint-4000/scheduler.pt +3 -0
  45. checkpoint-4000/special_tokens_map.json +5 -0
  46. checkpoint-4000/tokenizer.json +0 -0
  47. checkpoint-4000/tokenizer_config.json +19 -0
  48. checkpoint-4000/trainer_state.json +89 -0
  49. checkpoint-4000/training_args.bin +3 -0
  50. checkpoint-4000/vocab.json +0 -0
all_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 20.0,
3
- "total_flos": 3982090567680000.0,
4
- "train_loss": 3.3043636781382935,
5
- "train_runtime": 1550.8683,
6
- "train_samples": 381,
7
- "train_samples_per_second": 4.913,
8
- "train_steps_per_second": 2.463
9
  }
 
1
  {
2
  "epoch": 20.0,
3
+ "total_flos": 5058613739520000.0,
4
+ "train_loss": 3.4317847007562308,
5
+ "train_runtime": 1978.8511,
6
+ "train_samples": 484,
7
+ "train_samples_per_second": 4.892,
8
+ "train_steps_per_second": 2.446
9
  }
checkpoint-1000/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9374fbb2ab4ef9ed278af4c3dd610e1f31b821aeb5f658753ce0a7c2d429ccb3
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5beaf12ddc9d9a303c627e46cdb73d4ced10f0ef666329b7cd9020da7f23102
3
  size 497774208
checkpoint-1000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:efadce1692ab0bdae031711abd241c437d3ee5e9ef9b4993134be7ae94400a4e
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f0b53eba76d1a34d9047e1b95b10f475b951357f2602397dd079f8aa418c947
3
  size 995642298
checkpoint-1000/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:14b5d9e136cfcc76d63c4356cc00ffab5521f8f3617e241d52b7c803985e22ed
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45e278484841d4e84c6f9efe8d71ec9ee77449ec60146753c000c2d11b5fabef
3
  size 14244
checkpoint-1000/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0d96409fac135efbe2309f3c0211dadf7288e17afa4f579d63e7c168779b9281
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c85fad065c755ecefb2f8a75492cbf53d2a0a965164b922ca6ec6f2c00c8572
3
  size 1064
checkpoint-1000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 5.2356020942408374,
5
  "eval_steps": 500,
6
  "global_step": 1000,
7
  "is_hyper_param_search": false,
@@ -9,22 +9,22 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  }
25
  ],
26
  "logging_steps": 500,
27
- "max_steps": 3820,
28
  "num_input_tokens_seen": 0,
29
  "num_train_epochs": 20,
30
  "save_steps": 500,
@@ -40,7 +40,7 @@
40
  "attributes": {}
41
  }
42
  },
43
- "total_flos": 1042555207680000.0,
44
  "train_batch_size": 2,
45
  "trial_name": null,
46
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 4.132231404958677,
5
  "eval_steps": 500,
6
  "global_step": 1000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  }
25
  ],
26
  "logging_steps": 500,
27
+ "max_steps": 4840,
28
  "num_input_tokens_seen": 0,
29
  "num_train_epochs": 20,
30
  "save_steps": 500,
 
40
  "attributes": {}
41
  }
42
  },
43
+ "total_flos": 1045168128000000.0,
44
  "train_batch_size": 2,
45
  "trial_name": null,
46
  "trial_params": null
checkpoint-1000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-1500/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9e46ba788734db115fb6e80cd918016b5d0a8176158c9a46ad41e5fd7374b2ee
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c65ce10ee1d47334c3ebfc529816eae470e6a4ae0c69bef56cd4127a0c151c06
3
  size 497774208
checkpoint-1500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc409abb6f627a63e5c6fc725c92793bb69ef51474783aa8aa68ecfb0d9ea20c
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7161de512a1c1e291cd113998581e7854f8cd64c0e03005738bb0f954c8e45f0
3
  size 995642298
checkpoint-1500/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1db5d3b38d3e26de8462f8cf8ea5bc12b75c70c7392fb2c277da7f198bcd4291
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3eb87d57ea3a2417e5776b8e55a7a6dc26700d72b1e665f349da16c0e100a0a
3
  size 14244
checkpoint-1500/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:33779dd91a5985bcac403d0c5b95438046f2b3d559c530858d88d590aef5839d
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8b349109fa08d1a5efa4c03ec422e526e6efe34b4094e8f1fdad11bf14202c4
3
  size 1064
checkpoint-1500/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 7.853403141361256,
5
  "eval_steps": 500,
6
  "global_step": 1500,
7
  "is_hyper_param_search": false,
@@ -9,29 +9,29 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  },
25
  {
26
- "epoch": 7.853403141361256,
27
- "grad_norm": 4.553157329559326,
28
- "learning_rate": 3.036649214659686e-05,
29
- "loss": 3.4509,
30
  "step": 1500
31
  }
32
  ],
33
  "logging_steps": 500,
34
- "max_steps": 3820,
35
  "num_input_tokens_seen": 0,
36
  "num_train_epochs": 20,
37
  "save_steps": 500,
@@ -47,7 +47,7 @@
47
  "attributes": {}
48
  }
49
  },
50
- "total_flos": 1564094103552000.0,
51
  "train_batch_size": 2,
52
  "trial_name": null,
53
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 6.198347107438017,
5
  "eval_steps": 500,
6
  "global_step": 1500,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  },
25
  {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
  "step": 1500
31
  }
32
  ],
33
  "logging_steps": 500,
34
+ "max_steps": 4840,
35
  "num_input_tokens_seen": 0,
36
  "num_train_epochs": 20,
37
  "save_steps": 500,
 
47
  "attributes": {}
48
  }
49
  },
50
+ "total_flos": 1567752192000000.0,
51
  "train_batch_size": 2,
52
  "trial_name": null,
53
  "trial_params": null
checkpoint-1500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-2000/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c9de7024658ea0d43e1e1c53f32f4822af922273dc673662db08a8d6a141e952
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed127b06da69814fc6ee18bb21fac52cb507185099767a499f143a40a1d6511
3
  size 497774208
checkpoint-2000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d98936d517a4c41ba060cd5f77e4668686846b97556395a277778a952c639ac6
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef83972e43ff5876442393a3ae88d23eefd734f9a4aa11ae0a7472e72a7cb065
3
  size 995642298
checkpoint-2000/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fe24944c83fc32fd759afd3e676042b69a588c11b6add69a89f07481ed1d7b4e
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04846cbb4f4454903a172160e4a5f0fa9ef37302232d2ffb839e83ba05a8b144
3
  size 14244
checkpoint-2000/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eb020281ace5e9570b54946f63f46d18d91488bfbb3ae1454798e4c4ff1f70ab
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:216904d6402a16d79c455f88c537d1007db78a16d9b0e7c1430aad3c6e4868c0
3
  size 1064
checkpoint-2000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 10.471204188481675,
5
  "eval_steps": 500,
6
  "global_step": 2000,
7
  "is_hyper_param_search": false,
@@ -9,36 +9,36 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  },
25
  {
26
- "epoch": 7.853403141361256,
27
- "grad_norm": 4.553157329559326,
28
- "learning_rate": 3.036649214659686e-05,
29
- "loss": 3.4509,
30
  "step": 1500
31
  },
32
  {
33
- "epoch": 10.471204188481675,
34
- "grad_norm": 4.436212062835693,
35
- "learning_rate": 2.382198952879581e-05,
36
- "loss": 3.2407,
37
  "step": 2000
38
  }
39
  ],
40
  "logging_steps": 500,
41
- "max_steps": 3820,
42
  "num_input_tokens_seen": 0,
43
  "num_train_epochs": 20,
44
  "save_steps": 500,
@@ -54,7 +54,7 @@
54
  "attributes": {}
55
  }
56
  },
57
- "total_flos": 2085110415360000.0,
58
  "train_batch_size": 2,
59
  "trial_name": null,
60
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 8.264462809917354,
5
  "eval_steps": 500,
6
  "global_step": 2000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  },
25
  {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
  "step": 1500
31
  },
32
  {
33
+ "epoch": 8.264462809917354,
34
+ "grad_norm": 3.783989191055298,
35
+ "learning_rate": 2.9338842975206616e-05,
36
+ "loss": 3.5136,
37
  "step": 2000
38
  }
39
  ],
40
  "logging_steps": 500,
41
+ "max_steps": 4840,
42
  "num_input_tokens_seen": 0,
43
  "num_train_epochs": 20,
44
  "save_steps": 500,
 
54
  "attributes": {}
55
  }
56
  },
57
+ "total_flos": 2090336256000000.0,
58
  "train_batch_size": 2,
59
  "trial_name": null,
60
  "trial_params": null
checkpoint-2000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-2500/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5984eb709b5c93a7d5eb92abd745e613fb990be404ced8603000ec605c0c7f00
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28943bf5a085f93dc989addd5c066ed1d2637ad38b04edb8b10f61cf447d10ee
3
  size 497774208
checkpoint-2500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e2ad2ecd854cbec1dda54bf9cda10a979a5341154044f567087733cf8cf386d1
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a337ae9f9dfd339adbf2c2ca80b968b301f19ee014299271051f6f833216730
3
  size 995642298
checkpoint-2500/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:855f9cc495b3b87180ebaf565bd4bd614066909697663573736256c818c0dc49
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de87d37e28b59d5460a468507f69949dd76f59c38b265d3d403b5b0bcdbeb182
3
  size 14244
checkpoint-2500/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2432a09a517e1cc05aad739783864b8631ac12f1dc79e5cd6b52f9cbea0c0e0e
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:209f428962b770daaca94913cd7a352c78e229ca85181d4f60bc89437e7a4f8a
3
  size 1064
checkpoint-2500/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 13.089005235602095,
5
  "eval_steps": 500,
6
  "global_step": 2500,
7
  "is_hyper_param_search": false,
@@ -9,43 +9,43 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  },
25
  {
26
- "epoch": 7.853403141361256,
27
- "grad_norm": 4.553157329559326,
28
- "learning_rate": 3.036649214659686e-05,
29
- "loss": 3.4509,
30
  "step": 1500
31
  },
32
  {
33
- "epoch": 10.471204188481675,
34
- "grad_norm": 4.436212062835693,
35
- "learning_rate": 2.382198952879581e-05,
36
- "loss": 3.2407,
37
  "step": 2000
38
  },
39
  {
40
- "epoch": 13.089005235602095,
41
- "grad_norm": 4.0396833419799805,
42
- "learning_rate": 1.7277486910994763e-05,
43
- "loss": 3.0836,
44
  "step": 2500
45
  }
46
  ],
47
  "logging_steps": 500,
48
- "max_steps": 3820,
49
  "num_input_tokens_seen": 0,
50
  "num_train_epochs": 20,
51
  "save_steps": 500,
@@ -61,7 +61,7 @@
61
  "attributes": {}
62
  }
63
  },
64
- "total_flos": 2606126727168000.0,
65
  "train_batch_size": 2,
66
  "trial_name": null,
67
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 10.330578512396695,
5
  "eval_steps": 500,
6
  "global_step": 2500,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  },
25
  {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
  "step": 1500
31
  },
32
  {
33
+ "epoch": 8.264462809917354,
34
+ "grad_norm": 3.783989191055298,
35
+ "learning_rate": 2.9338842975206616e-05,
36
+ "loss": 3.5136,
37
  "step": 2000
38
  },
39
  {
40
+ "epoch": 10.330578512396695,
41
+ "grad_norm": 3.957587242126465,
42
+ "learning_rate": 2.4173553719008264e-05,
43
+ "loss": 3.3546,
44
  "step": 2500
45
  }
46
  ],
47
  "logging_steps": 500,
48
+ "max_steps": 4840,
49
  "num_input_tokens_seen": 0,
50
  "num_train_epochs": 20,
51
  "save_steps": 500,
 
61
  "attributes": {}
62
  }
63
  },
64
+ "total_flos": 2612920320000000.0,
65
  "train_batch_size": 2,
66
  "trial_name": null,
67
  "trial_params": null
checkpoint-2500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-3000/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4746589da687f328f184c1432356ab6451cbc6a0df98b4e8216949b467e08df0
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:498d1e2760d2a9a113ca89dff4bc7dd936488225d23ed0377908d9fd3feb64e5
3
  size 497774208
checkpoint-3000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5facc9ff9b9b2726aec226bebe8180a7a80f6635289bf19db406d9d299a38b9
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f37b9d4314cabfc361de963f574c22d0ddf193c5d8146bc5474a2ca3f638e27
3
  size 995642298
checkpoint-3000/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:01267d7b4cf70c821d3f48d93d2fb0884bd26ecfb720249804977bc9f06135f1
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20cf46f56081916b9069740c1050bb58254597425e7386324a5f03a7b689a8f1
3
  size 14244
checkpoint-3000/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c42ca047edeed0b2a31eb765d8f3410c4dcd952eaaf6b9ed5083f20c322be934
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d94ad599c69f513b268445911a3ae6702b17dcd2e984bb3ff66863f183b549c1
3
  size 1064
checkpoint-3000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 15.706806282722512,
5
  "eval_steps": 500,
6
  "global_step": 3000,
7
  "is_hyper_param_search": false,
@@ -9,50 +9,50 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  },
25
  {
26
- "epoch": 7.853403141361256,
27
- "grad_norm": 4.553157329559326,
28
- "learning_rate": 3.036649214659686e-05,
29
- "loss": 3.4509,
30
  "step": 1500
31
  },
32
  {
33
- "epoch": 10.471204188481675,
34
- "grad_norm": 4.436212062835693,
35
- "learning_rate": 2.382198952879581e-05,
36
- "loss": 3.2407,
37
  "step": 2000
38
  },
39
  {
40
- "epoch": 13.089005235602095,
41
- "grad_norm": 4.0396833419799805,
42
- "learning_rate": 1.7277486910994763e-05,
43
- "loss": 3.0836,
44
  "step": 2500
45
  },
46
  {
47
- "epoch": 15.706806282722512,
48
- "grad_norm": 4.427651882171631,
49
- "learning_rate": 1.0732984293193717e-05,
50
- "loss": 2.9576,
51
  "step": 3000
52
  }
53
  ],
54
  "logging_steps": 500,
55
- "max_steps": 3820,
56
  "num_input_tokens_seen": 0,
57
  "num_train_epochs": 20,
58
  "save_steps": 500,
@@ -68,7 +68,7 @@
68
  "attributes": {}
69
  }
70
  },
71
- "total_flos": 3127665623040000.0,
72
  "train_batch_size": 2,
73
  "trial_name": null,
74
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 12.396694214876034,
5
  "eval_steps": 500,
6
  "global_step": 3000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  },
25
  {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
  "step": 1500
31
  },
32
  {
33
+ "epoch": 8.264462809917354,
34
+ "grad_norm": 3.783989191055298,
35
+ "learning_rate": 2.9338842975206616e-05,
36
+ "loss": 3.5136,
37
  "step": 2000
38
  },
39
  {
40
+ "epoch": 10.330578512396695,
41
+ "grad_norm": 3.957587242126465,
42
+ "learning_rate": 2.4173553719008264e-05,
43
+ "loss": 3.3546,
44
  "step": 2500
45
  },
46
  {
47
+ "epoch": 12.396694214876034,
48
+ "grad_norm": 4.028080463409424,
49
+ "learning_rate": 1.900826446280992e-05,
50
+ "loss": 3.2309,
51
  "step": 3000
52
  }
53
  ],
54
  "logging_steps": 500,
55
+ "max_steps": 4840,
56
  "num_input_tokens_seen": 0,
57
  "num_train_epochs": 20,
58
  "save_steps": 500,
 
68
  "attributes": {}
69
  }
70
  },
71
+ "total_flos": 3135504384000000.0,
72
  "train_batch_size": 2,
73
  "trial_name": null,
74
  "trial_params": null
checkpoint-3000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-3500/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:290e9206eae53ca9e160de097cbc949cb2afef4211a1e55040692ed5c3299138
3
  size 497774208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0b8d96769ec5982c127dde065f80ee526c47ec0da91dd52a85f67aaf594681d
3
  size 497774208
checkpoint-3500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:59471ba05320b8d78ff6680a5d6efa58b4d5230a93daf3696fe0795cdd86dd62
3
  size 995642298
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e50dfaee8346cfc52626492e300cacb4e4beffc8c9fb89d53e7dbbebc68383d0
3
  size 995642298
checkpoint-3500/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:260ebc76cad0016101f81a6ca73590c0649e503a1392064f1e4a2dccf84db39a
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a5e76dbfd141d2467f248a12991f5e466ba8e1037baafba005821f14c34692b
3
  size 14244
checkpoint-3500/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e3a19745f3f0cda237e16e0664ba9c728fc5cf0e8960611153deaae843e7c697
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9518e75129d8c588e3a304d09c53e05346c56d68c335c33798746198d379af7
3
  size 1064
checkpoint-3500/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 18.324607329842934,
5
  "eval_steps": 500,
6
  "global_step": 3500,
7
  "is_hyper_param_search": false,
@@ -9,57 +9,57 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 2.6178010471204187,
13
- "grad_norm": 4.953078269958496,
14
- "learning_rate": 4.3455497382198955e-05,
15
- "loss": 4.0934,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 5.2356020942408374,
20
- "grad_norm": 4.2992095947265625,
21
- "learning_rate": 3.691099476439791e-05,
22
- "loss": 3.7159,
23
  "step": 1000
24
  },
25
  {
26
- "epoch": 7.853403141361256,
27
- "grad_norm": 4.553157329559326,
28
- "learning_rate": 3.036649214659686e-05,
29
- "loss": 3.4509,
30
  "step": 1500
31
  },
32
  {
33
- "epoch": 10.471204188481675,
34
- "grad_norm": 4.436212062835693,
35
- "learning_rate": 2.382198952879581e-05,
36
- "loss": 3.2407,
37
  "step": 2000
38
  },
39
  {
40
- "epoch": 13.089005235602095,
41
- "grad_norm": 4.0396833419799805,
42
- "learning_rate": 1.7277486910994763e-05,
43
- "loss": 3.0836,
44
  "step": 2500
45
  },
46
  {
47
- "epoch": 15.706806282722512,
48
- "grad_norm": 4.427651882171631,
49
- "learning_rate": 1.0732984293193717e-05,
50
- "loss": 2.9576,
51
  "step": 3000
52
  },
53
  {
54
- "epoch": 18.324607329842934,
55
- "grad_norm": 4.466192245483398,
56
- "learning_rate": 4.18848167539267e-06,
57
- "loss": 2.8844,
58
  "step": 3500
59
  }
60
  ],
61
  "logging_steps": 500,
62
- "max_steps": 3820,
63
  "num_input_tokens_seen": 0,
64
  "num_train_epochs": 20,
65
  "save_steps": 500,
@@ -75,7 +75,7 @@
75
  "attributes": {}
76
  }
77
  },
78
- "total_flos": 3648681934848000.0,
79
  "train_batch_size": 2,
80
  "trial_name": null,
81
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 14.462809917355372,
5
  "eval_steps": 500,
6
  "global_step": 3500,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
  "step": 1000
24
  },
25
  {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
  "step": 1500
31
  },
32
  {
33
+ "epoch": 8.264462809917354,
34
+ "grad_norm": 3.783989191055298,
35
+ "learning_rate": 2.9338842975206616e-05,
36
+ "loss": 3.5136,
37
  "step": 2000
38
  },
39
  {
40
+ "epoch": 10.330578512396695,
41
+ "grad_norm": 3.957587242126465,
42
+ "learning_rate": 2.4173553719008264e-05,
43
+ "loss": 3.3546,
44
  "step": 2500
45
  },
46
  {
47
+ "epoch": 12.396694214876034,
48
+ "grad_norm": 4.028080463409424,
49
+ "learning_rate": 1.900826446280992e-05,
50
+ "loss": 3.2309,
51
  "step": 3000
52
  },
53
  {
54
+ "epoch": 14.462809917355372,
55
+ "grad_norm": 4.280130863189697,
56
+ "learning_rate": 1.3842975206611573e-05,
57
+ "loss": 3.1278,
58
  "step": 3500
59
  }
60
  ],
61
  "logging_steps": 500,
62
+ "max_steps": 4840,
63
  "num_input_tokens_seen": 0,
64
  "num_train_epochs": 20,
65
  "save_steps": 500,
 
75
  "attributes": {}
76
  }
77
  },
78
+ "total_flos": 3658088448000000.0,
79
  "train_batch_size": 2,
80
  "trial_name": null,
81
  "trial_params": null
checkpoint-3500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8340bb38e933c578d75c23425604f9ded7016172341fa692743654b53749296
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
  size 5112
checkpoint-4000/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.41.1",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
checkpoint-4000/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": "4.41.1"
6
+ }
checkpoint-4000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-4000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ced2264599496d92e52768ed58acea9a6bad234e12a9a5f03183e913f17c7a47
3
+ size 497774208
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e235c25dce549e97d8a44cf666533f9774cbe0f9864afa5d1bc30603ebab2c83
3
+ size 995642298
checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:804c1a8a5cfa020f9fdb6aa26f7f5fbc7e2972fe277c86ad744be191fe6ff510
3
+ size 14244
checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b3aa2fe8174d0d642b21725c8efca027fcde9dbb86b41f88ac923024e12494f
3
+ size 1064
checkpoint-4000/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
checkpoint-4000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-4000/tokenizer_config.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "50256": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": true,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "tokenizer_class": "GPT2Tokenizer",
18
+ "unk_token": "<|endoftext|>"
19
+ }
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 16.52892561983471,
5
+ "eval_steps": 500,
6
+ "global_step": 4000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 2.0661157024793386,
13
+ "grad_norm": 4.206299781799316,
14
+ "learning_rate": 4.4834710743801654e-05,
15
+ "loss": 4.2802,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 4.132231404958677,
20
+ "grad_norm": 4.235241889953613,
21
+ "learning_rate": 3.9669421487603306e-05,
22
+ "loss": 3.9317,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 6.198347107438017,
27
+ "grad_norm": 4.2089643478393555,
28
+ "learning_rate": 3.4504132231404964e-05,
29
+ "loss": 3.7086,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 8.264462809917354,
34
+ "grad_norm": 3.783989191055298,
35
+ "learning_rate": 2.9338842975206616e-05,
36
+ "loss": 3.5136,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 10.330578512396695,
41
+ "grad_norm": 3.957587242126465,
42
+ "learning_rate": 2.4173553719008264e-05,
43
+ "loss": 3.3546,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 12.396694214876034,
48
+ "grad_norm": 4.028080463409424,
49
+ "learning_rate": 1.900826446280992e-05,
50
+ "loss": 3.2309,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 14.462809917355372,
55
+ "grad_norm": 4.280130863189697,
56
+ "learning_rate": 1.3842975206611573e-05,
57
+ "loss": 3.1278,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 16.52892561983471,
62
+ "grad_norm": 3.9069178104400635,
63
+ "learning_rate": 8.677685950413224e-06,
64
+ "loss": 3.0433,
65
+ "step": 4000
66
+ }
67
+ ],
68
+ "logging_steps": 500,
69
+ "max_steps": 4840,
70
+ "num_input_tokens_seen": 0,
71
+ "num_train_epochs": 20,
72
+ "save_steps": 500,
73
+ "stateful_callbacks": {
74
+ "TrainerControl": {
75
+ "args": {
76
+ "should_epoch_stop": false,
77
+ "should_evaluate": false,
78
+ "should_log": false,
79
+ "should_save": true,
80
+ "should_training_stop": false
81
+ },
82
+ "attributes": {}
83
+ }
84
+ },
85
+ "total_flos": 4180672512000000.0,
86
+ "train_batch_size": 2,
87
+ "trial_name": null,
88
+ "trial_params": null
89
+ }
checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9430e9d69ca56d23cd7dae5d3e97d08be48f937801607042800d8747f06ded5a
3
+ size 5112
checkpoint-4000/vocab.json ADDED
The diff for this file is too large to render. See raw diff