AlekseyKorshuk commited on
Commit
4008979
1 Parent(s): 17aae1f

huggingartists

Browse files
README.md CHANGED
@@ -45,15 +45,15 @@ from datasets import load_dataset
45
  dataset = load_dataset("huggingartists/viktor-tsoi")
46
  ```
47
 
48
- [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/3dbxo21p/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 Виктор Цой (Viktor Tsoi)'s lyrics.
53
 
54
- Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1qe51xj0) for full transparency and reproducibility.
55
 
56
- At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1qe51xj0/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
 
45
  dataset = load_dataset("huggingartists/viktor-tsoi")
46
  ```
47
 
48
+ [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/1uufz4th/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 Виктор Цой (Viktor Tsoi)'s lyrics.
53
 
54
+ Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/8mogk3d7) for full transparency and reproducibility.
55
 
56
+ At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/8mogk3d7/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/viktor-tsoi",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
evaluation.txt CHANGED
@@ -1 +1 @@
1
- {"eval_loss": 1.742375135421753, "eval_runtime": 2.6042, "eval_samples_per_second": 20.736, "eval_steps_per_second": 2.688, "epoch": 1.0}
 
1
+ {"eval_loss": 1.6593416929244995, "eval_runtime": 2.657, "eval_samples_per_second": 21.453, "eval_steps_per_second": 3.011, "epoch": 2.0}
flax_model.msgpack CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:91e10663a0f3244857439b56cd04237c31b6eae728e7946713ce843d8d28e75e
3
  size 497764120
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d49064efb266e519ece7257c9d60deebe252fb6406d169e20dadeff277f7dc37
3
  size 497764120
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9b6393030e2096bada970db9d27e3ff22446a05a9077ef19b5fc54c55e11b45d
3
  size 995603825
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db593a7972d8e20eed34acf31aa0d79678447e39973eff379958dc4fbba2231e
3
  size 995603825
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:546047b53d33d8741418a5178e3adc522fb1e0d0301daf0f8e99e986b5d9777c
3
  size 510403817
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c750cb8a3a0856153a37e616ee6dce7c26d113b033fd26b7f7d02ab5d7e8639
3
  size 510403817
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:43f420abc20ff158a09ca170c7b11375c8ca382b141e051a66f1a65570d74fea
3
- size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14c00e5accc59ae7985545e12496021de015b318dd1489d964d4a6d8d6ce5297
3
+ size 14567
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8d4369506dedd54e65c6967a3362585e3aa7b5afb87732c009d3c12491c2fbd1
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:377ca4976456890405a303851a7b4d7bfd45747ee263ed38f7554f3c6e6a1db4
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/viktor-tsoi", "tokenizer_class": "GPT2Tokenizer"}
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "best_metric": 1.742375135421753,
3
- "best_model_checkpoint": "output/viktor-tsoi/checkpoint-41",
4
- "epoch": 1.0,
5
- "global_step": 41,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -62,11 +62,67 @@
62
  "eval_samples_per_second": 20.944,
63
  "eval_steps_per_second": 2.715,
64
  "step": 41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
65
  }
66
  ],
67
- "max_steps": 41,
68
- "num_train_epochs": 1,
69
- "total_flos": 42459955200000.0,
70
  "trial_name": null,
71
  "trial_params": null
72
  }
 
1
  {
2
+ "best_metric": 1.6593416929244995,
3
+ "best_model_checkpoint": "output/viktor-tsoi/checkpoint-82",
4
+ "epoch": 2.0,
5
+ "global_step": 82,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
62
  "eval_samples_per_second": 20.944,
63
  "eval_steps_per_second": 2.715,
64
  "step": 41
65
+ },
66
+ {
67
+ "epoch": 1.1,
68
+ "learning_rate": 3.197007505031765e-06,
69
+ "loss": 1.858,
70
+ "step": 45
71
+ },
72
+ {
73
+ "epoch": 1.22,
74
+ "learning_rate": 1.5675842264214697e-05,
75
+ "loss": 1.8206,
76
+ "step": 50
77
+ },
78
+ {
79
+ "epoch": 1.34,
80
+ "learning_rate": 3.58284204500588e-05,
81
+ "loss": 1.7518,
82
+ "step": 55
83
+ },
84
+ {
85
+ "epoch": 1.46,
86
+ "learning_rate": 6.0732717017669706e-05,
87
+ "loss": 1.744,
88
+ "step": 60
89
+ },
90
+ {
91
+ "epoch": 1.59,
92
+ "learning_rate": 8.677773105069102e-05,
93
+ "loss": 1.7542,
94
+ "step": 65
95
+ },
96
+ {
97
+ "epoch": 1.71,
98
+ "learning_rate": 0.00011018706319231134,
99
+ "loss": 1.7612,
100
+ "step": 70
101
+ },
102
+ {
103
+ "epoch": 1.83,
104
+ "learning_rate": 0.00012756647503932202,
105
+ "loss": 1.7551,
106
+ "step": 75
107
+ },
108
+ {
109
+ "epoch": 1.95,
110
+ "learning_rate": 0.0001363960370713319,
111
+ "loss": 1.7343,
112
+ "step": 80
113
+ },
114
+ {
115
+ "epoch": 2.0,
116
+ "eval_loss": 1.6593416929244995,
117
+ "eval_runtime": 2.6581,
118
+ "eval_samples_per_second": 21.444,
119
+ "eval_steps_per_second": 3.01,
120
+ "step": 82
121
  }
122
  ],
123
+ "max_steps": 82,
124
+ "num_train_epochs": 2,
125
+ "total_flos": 84397326336000.0,
126
  "trial_name": null,
127
  "trial_params": null
128
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b49b2cf9a36957cbbc6e44492b8b37d4e9c2e9a29bf05d217e090cc84a1dda16
3
  size 2671
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4864274d1168011bbafdc53bc5f4a82d25908496ba29f580633e85995da2e5f
3
  size 2671