AlekseyKorshuk commited on
Commit
54bdc43
1 Parent(s): ba76bd3

huggingartists

Browse files
README.md CHANGED
@@ -14,7 +14,7 @@ widget:
14
  <div class="inline-flex flex-col" style="line-height: 1.5;">
15
  <div class="flex">
16
  <div
17
- style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url(&#39;https://images.genius.com/e4f988f1ee26618c5dd41b59b8ff2b43.1000x1000x1.jpg&#39;)">
18
  </div>
19
  </div>
20
  <div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
@@ -45,15 +45,15 @@ from datasets import load_dataset
45
  dataset = load_dataset("huggingartists/coldplay")
46
  ```
47
 
48
- [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/399heq03/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 Coldplay's lyrics.
53
 
54
- Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/3o6fr9bq) for full transparency and reproducibility.
55
 
56
- At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/3o6fr9bq/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
 
14
  <div class="inline-flex flex-col" style="line-height: 1.5;">
15
  <div class="flex">
16
  <div
17
+ style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url(&#39;https://images.genius.com/6cfcc2b1425286fe0d0b8c857c895b63.600x338x200.gif&#39;)">
18
  </div>
19
  </div>
20
  <div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
 
45
  dataset = load_dataset("huggingartists/coldplay")
46
  ```
47
 
48
+ [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/34tqcy7u/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 Coldplay's lyrics.
53
 
54
+ Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/23h7o09h) for full transparency and reproducibility.
55
 
56
+ At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/23h7o09h/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "huggingartists/coldplay",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
@@ -18,7 +18,9 @@
18
  "n_inner": null,
19
  "n_layer": 12,
20
  "n_positions": 1024,
 
21
  "resid_pdrop": 0.1,
 
22
  "scale_attn_weights": true,
23
  "summary_activation": null,
24
  "summary_first_dropout": 0.1,
@@ -35,7 +37,7 @@
35
  }
36
  },
37
  "torch_dtype": "float32",
38
- "transformers_version": "4.11.3",
39
  "use_cache": true,
40
  "vocab_size": 50257
41
  }
 
1
  {
2
+ "_name_or_path": "coldplay",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
 
18
  "n_inner": null,
19
  "n_layer": 12,
20
  "n_positions": 1024,
21
+ "reorder_and_upcast_attn": false,
22
  "resid_pdrop": 0.1,
23
+ "scale_attn_by_inverse_layer_idx": false,
24
  "scale_attn_weights": true,
25
  "summary_activation": null,
26
  "summary_first_dropout": 0.1,
 
37
  }
38
  },
39
  "torch_dtype": "float32",
40
+ "transformers_version": "4.20.1",
41
  "use_cache": true,
42
  "vocab_size": 50257
43
  }
evaluation.txt CHANGED
@@ -1 +1 @@
1
- {"eval_loss": 1.2095028162002563, "eval_runtime": 3.4191, "eval_samples_per_second": 21.643, "eval_steps_per_second": 2.925, "epoch": 19.0}
 
1
+ {"eval_loss": 1.0933133363723755, "eval_runtime": 1.4407, "eval_samples_per_second": 38.87, "eval_steps_per_second": 4.859, "epoch": 12.0}
flax_model.msgpack CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a8806c168019cbc1ae2ffed46c832caa9529a17675216fdd1c782cc1176e7331
3
  size 497764120
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a968012daa02f2aeede1b08301d9d8cdf0b0a0b66138b34e01d21479815aa9d6
3
  size 497764120
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:37ac05510aa6c132c3344e1ec9d0f1e37c7031d848f40767d745da3aa0e66a9b
3
  size 995604017
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a49be6f0dace143bae78e031981d54f03c5f1f75566034d46b4851b17b2efca
3
  size 995604017
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0a7158419658b030a3d8570b711350a55f73d17342aa9920b8e41aff9c03d4a1
3
- size 510403817
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc5669114394eec4c7fd9808510a62c68422cd295b0720550d9742974be536f
3
+ size 510396521
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1096f6f012654cb7b05c9d15f0c8c81dff6f30d1dbde8cf0b9489e06f1331a53
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3837ecb65e024d5d16e99eee7628f55aed7ce556e170bd1fa6cd6486cd8c4b50
3
  size 14503
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b9430296501fbe754e360495552f3a391e4c1ba2be1a491d6867af0670e74711
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1bc86d1066eb2b98a95a6b7ab12eba7384232826b3650f7109aba3e265fc996a
3
  size 623
special_tokens_map.json CHANGED
@@ -1 +1,5 @@
1
- {"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json CHANGED
@@ -1 +1,10 @@
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/coldplay", "tokenizer_class": "GPT2Tokenizer"}
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<|endoftext|>",
4
+ "eos_token": "<|endoftext|>",
5
+ "model_max_length": 1024,
6
+ "name_or_path": "huggingartists/coldplay",
7
+ "special_tokens_map_file": null,
8
+ "tokenizer_class": "GPT2Tokenizer",
9
+ "unk_token": "<|endoftext|>"
10
+ }
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "best_metric": 1.2095028162002563,
3
- "best_model_checkpoint": "output/coldplay/checkpoint-528",
4
  "epoch": 11.0,
5
- "global_step": 528,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -732,11 +732,49 @@
732
  "eval_samples_per_second": 22.445,
733
  "eval_steps_per_second": 3.033,
734
  "step": 528
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
735
  }
736
  ],
737
- "max_steps": 912,
738
- "num_train_epochs": 19,
739
- "total_flos": 549235851264000.0,
740
  "trial_name": null,
741
  "trial_params": null
742
  }
 
1
  {
2
+ "best_metric": 1.0933133363723755,
3
+ "best_model_checkpoint": "output/coldplay/checkpoint-550",
4
  "epoch": 11.0,
5
+ "global_step": 550,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
732
  "eval_samples_per_second": 22.445,
733
  "eval_steps_per_second": 3.033,
734
  "step": 528
735
+ },
736
+ {
737
+ "epoch": 10.6,
738
+ "learning_rate": 4.7401434185878923e-05,
739
+ "loss": 1.0609,
740
+ "step": 530
741
+ },
742
+ {
743
+ "epoch": 10.7,
744
+ "learning_rate": 2.8277931692736518e-05,
745
+ "loss": 1.2072,
746
+ "step": 535
747
+ },
748
+ {
749
+ "epoch": 10.8,
750
+ "learning_rate": 1.3101434185878659e-05,
751
+ "loss": 1.2221,
752
+ "step": 540
753
+ },
754
+ {
755
+ "epoch": 10.9,
756
+ "learning_rate": 3.3575229821524526e-06,
757
+ "loss": 1.0503,
758
+ "step": 545
759
+ },
760
+ {
761
+ "epoch": 11.0,
762
+ "learning_rate": 0.0,
763
+ "loss": 1.0194,
764
+ "step": 550
765
+ },
766
+ {
767
+ "epoch": 11.0,
768
+ "eval_loss": 1.0933133363723755,
769
+ "eval_runtime": 1.2839,
770
+ "eval_samples_per_second": 43.616,
771
+ "eval_steps_per_second": 5.452,
772
+ "step": 550
773
  }
774
  ],
775
+ "max_steps": 600,
776
+ "num_train_epochs": 12,
777
+ "total_flos": 572229550080000.0,
778
  "trial_name": null,
779
  "trial_params": null
780
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8634fa1cc14205deda566a1ce3471c09a00adb32b1ca2deb15e6602eb34457d2
3
- size 2863
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:092ba382e6c8b2a6645125b1644c9103ebd83510860ac51077dcbeccd57b05f0
3
+ size 3311