huggingartists

README.md

@@ -45,15 +45,15 @@ from datasets import load_dataset
 dataset = load_dataset("huggingartists/bill-wurtz")
 ```
 
-[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/3m596wb1/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
+[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/1y01b0sy/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
 
 ## Training procedure
 
 The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Bill Wurtz's lyrics.
 
-Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/tn25vkjp) for full transparency and reproducibility.
+Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1l9hr5w2) for full transparency and reproducibility.
 
-At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/tn25vkjp/artifacts) is logged and versioned.
+At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1l9hr5w2/artifacts) is logged and versioned.
 
 ## How to use
 

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "gpt2",
+  "_name_or_path": "huggingartists/bill-wurtz",
   "activation_function": "gelu_new",
   "architectures": [
     "GPT2LMHeadModel"

evaluation.txt

@@ -1 +1 @@
-{"eval_loss": 3.056016445159912, "eval_runtime": 0.7021, "eval_samples_per_second": 38.458, "eval_steps_per_second": 5.698, "epoch": 5.0}
+{"eval_loss": 4.039820671081543, "eval_runtime": 0.5683, "eval_samples_per_second": 36.954, "eval_steps_per_second": 5.279, "epoch": 22.0}

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94a38c311db28275ebd22bd1a96b5f1ae0705768584e0b02b91c6d3f51f16d1a
+oid sha256:aa1422428e38213576ff235f2001350e546a4c4643f2b2f472d00667df459b33
 size 995603825

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1232337b23975f069a764f8e344e5636ceb4c40315428483d2aca49f6796a108
-size 14503
+oid sha256:e1d50a042836fb7244e3a57318b3664fd2cb3c84d1ecf895aeecf11cd286b9d1
+size 14567

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e9d0d89fe51412bc7eecc43912be03d94b03466976a036bba499f0a41644d0f8
+oid sha256:b1aee7c44e1d1ea5ba4c4593abd0231fd0c216d063ddd82618d0f2d2c0e4a5dc
 size 623

tokenizer_config.json

@@ -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"}
+{"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/bill-wurtz", "tokenizer_class": "GPT2Tokenizer"}

trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 3.056016445159912,
-  "best_model_checkpoint": "output/bill-wurtz/checkpoint-54",
-  "epoch": 2.0,
-  "global_step": 54,
+  "best_metric": 2.6812992095947266,
+  "best_model_checkpoint": "output/bill-wurtz/checkpoint-140",
+  "epoch": 5.0,
+  "global_step": 140,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -82,11 +82,151 @@
       "eval_samples_per_second": 42.746,
       "eval_steps_per_second": 6.333,
       "step": 54
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00013676865759867644,
+      "loss": 2.5878,
+      "step": 55
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.762599468231201,
+      "eval_runtime": 0.5151,
+      "eval_samples_per_second": 40.769,
+      "eval_steps_per_second": 5.824,
+      "step": 56
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00013040646433810595,
+      "loss": 2.8115,
+      "step": 60
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00010509740044895205,
+      "loss": 2.5949,
+      "step": 65
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 2.6643,
+      "step": 70
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 3.210259955104798e-05,
+      "loss": 2.5837,
+      "step": 75
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 6.793535661894062e-06,
+      "loss": 2.7461,
+      "step": 80
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.6894850730895996,
+      "eval_runtime": 0.5306,
+      "eval_samples_per_second": 39.579,
+      "eval_steps_per_second": 5.654,
+      "step": 84
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.313424013235498e-07,
+      "loss": 2.4693,
+      "step": 85
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.496636030269314e-05,
+      "loss": 2.4734,
+      "step": 90
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 4.594285634987545e-05,
+      "loss": 2.5555,
+      "step": 95
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 8.386493606940314e-05,
+      "loss": 2.498,
+      "step": 100
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.0001171075251893971,
+      "loss": 2.3626,
+      "step": 105
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.0001354800547756731,
+      "loss": 2.4603,
+      "step": 110
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 2.6920154094696045,
+      "eval_runtime": 0.5456,
+      "eval_samples_per_second": 38.49,
+      "eval_steps_per_second": 5.499,
+      "step": 112
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00013335039645915404,
+      "loss": 2.3237,
+      "step": 115
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.00011137140040750914,
+      "loss": 2.2052,
+      "step": 120
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.62807630606869e-05,
+      "loss": 2.2121,
+      "step": 125
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.883557549653544e-05,
+      "loss": 2.2551,
+      "step": 130
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.0514719932939762e-05,
+      "loss": 2.2849,
+      "step": 135
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 2.2654,
+      "step": 140
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 2.6812992095947266,
+      "eval_runtime": 0.5264,
+      "eval_samples_per_second": 39.891,
+      "eval_steps_per_second": 5.699,
+      "step": 140
     }
   ],
-  "max_steps": 135,
-  "num_train_epochs": 5,
-  "total_flos": 56439078912000.0,
+  "max_steps": 616,
+  "num_train_epochs": 22,
+  "total_flos": 145278369792000.0,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74110011f543625430a4ba42009eb3a204c303837c8ad074906c1a42d538d532
+oid sha256:41eee99e83d6fdcd4ae4efff4b1f5e17120c0670785282724daf91004a53de18
 size 2671