AlekseyKorshuk commited on
Commit
bf33705
1 Parent(s): 2cfacc7

huggingartists

Browse files
README.md CHANGED
@@ -45,15 +45,15 @@ from datasets import load_dataset
45
  dataset = load_dataset("huggingartists/death-grips")
46
  ```
47
 
48
- [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/ec6to0ba/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Death Grips's lyrics.
53
 
54
- Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/14p94gs8) for full transparency and reproducibility.
55
 
56
- At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/14p94gs8/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
 
45
  dataset = load_dataset("huggingartists/death-grips")
46
  ```
47
 
48
+ [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/1z4yxb78/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Death Grips's lyrics.
53
 
54
+ Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/18v4x337) for full transparency and reproducibility.
55
 
56
+ At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/18v4x337/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "gpt2",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
 
1
  {
2
+ "_name_or_path": "huggingartists/death-grips",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
evaluation.txt CHANGED
@@ -1 +1 @@
1
- {"eval_loss": 3.4941582679748535, "eval_runtime": 0.5943, "eval_samples_per_second": 74.031, "eval_steps_per_second": 10.095, "epoch": 1.0}
 
1
+ {"eval_loss": 3.1992340087890625, "eval_runtime": 0.5011, "eval_samples_per_second": 75.835, "eval_steps_per_second": 9.978, "epoch": 61.0}
flax_model.msgpack CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1757221700272e752c8d7c1a8aaeb5afb46e475c9d7766873751da252b23c78
3
  size 497764120
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93ea9acbaf03e5cd973f67cc943630d2e10e1ec2156b1a368c852eef0b80c447
3
  size 497764120
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:51504115416c0c0ceb5999766325d8fc015e4cfd213b87fec34e0c2f4b840ab6
3
  size 995603825
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91885fb2d0a23add8ab65706c01695bf0d25c6bdb6fa85982f61c7a4403d7562
3
  size 995603825
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3d2c7cc19add62d7e343e93cd67b3fc2447d313257ae21733cac467b3fd084a1
3
  size 510403817
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45a4d5fb1bb4bc74f1d211806c05b5291969db23c26535e118f8d45dd0493b88
3
  size 510403817
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8aa36eb6f9781a66d2ff9dda6822a5ac1492c196faa65d6cc177e35a39f3abeb
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80bc2e05cd33aba954c4fa5280f46e5f77924d62e7982e05ecf36aff4194107e
3
  size 14503
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:789f53ac5f4078426fcc32c51995193c87b589ef511dca806dd5fcd3735b78a5
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2222362d80f6c4a5547283104536d5c4564191fb641a9b7f7b9e346e9a58c2ee
3
  size 623
tokenizer_config.json CHANGED
@@ -1 +1 @@
1
- {"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "gpt2", "tokenizer_class": "GPT2Tokenizer"}
 
1
+ {"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "huggingartists/death-grips", "tokenizer_class": "GPT2Tokenizer"}
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "best_metric": 3.4941582679748535,
3
- "best_model_checkpoint": "output/death-grips/checkpoint-32",
4
- "epoch": 1.0,
5
- "global_step": 32,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -50,11 +50,69 @@
50
  "eval_samples_per_second": 75.33,
51
  "eval_steps_per_second": 10.272,
52
  "step": 32
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  }
54
  ],
55
- "max_steps": 32,
56
- "num_train_epochs": 1,
57
- "total_flos": 32922796032000.0,
58
  "trial_name": null,
59
  "trial_params": null
60
  }
 
1
  {
2
+ "best_metric": 3.1992340087890625,
3
+ "best_model_checkpoint": "output/death-grips/checkpoint-66",
4
+ "epoch": 2.0,
5
+ "global_step": 66,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
50
  "eval_samples_per_second": 75.33,
51
  "eval_steps_per_second": 10.272,
52
  "step": 32
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_loss": 3.25174617767334,
57
+ "eval_runtime": 0.4993,
58
+ "eval_samples_per_second": 76.107,
59
+ "eval_steps_per_second": 10.014,
60
+ "step": 33
61
+ },
62
+ {
63
+ "epoch": 1.06,
64
+ "learning_rate": 1.2396913677783124e-06,
65
+ "loss": 3.1933,
66
+ "step": 35
67
+ },
68
+ {
69
+ "epoch": 1.21,
70
+ "learning_rate": 1.4676757700644785e-05,
71
+ "loss": 3.1172,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 1.36,
76
+ "learning_rate": 4.0102530108070535e-05,
77
+ "loss": 3.4529,
78
+ "step": 45
79
+ },
80
+ {
81
+ "epoch": 1.52,
82
+ "learning_rate": 7.186411942550872e-05,
83
+ "loss": 3.2121,
84
+ "step": 50
85
+ },
86
+ {
87
+ "epoch": 1.67,
88
+ "learning_rate": 0.00010290000000000001,
89
+ "loss": 3.1379,
90
+ "step": 55
91
+ },
92
+ {
93
+ "epoch": 1.82,
94
+ "learning_rate": 0.000126309992352219,
95
+ "loss": 3.1008,
96
+ "step": 60
97
+ },
98
+ {
99
+ "epoch": 1.97,
100
+ "learning_rate": 0.0001368893738885136,
101
+ "loss": 3.1177,
102
+ "step": 65
103
+ },
104
+ {
105
+ "epoch": 2.0,
106
+ "eval_loss": 3.1992340087890625,
107
+ "eval_runtime": 0.4988,
108
+ "eval_samples_per_second": 76.185,
109
+ "eval_steps_per_second": 10.024,
110
+ "step": 66
111
  }
112
  ],
113
+ "max_steps": 2013,
114
+ "num_train_epochs": 61,
115
+ "total_flos": 66890760192000.0,
116
  "trial_name": null,
117
  "trial_params": null
118
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:79b0711c457b001263266f6599453533724d726bf5a7988a0de95942c1d57764
3
  size 2863
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:820c88075277a4ce2d8bfc6da26d7fd133ba368b384db802a6a4afafc3e34877
3
  size 2863