add/dem gpt2

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.41066931981946914,
+    "eval_loss": 3.3461549282073975,
+    "eval_runtime": 145.1896,
+    "eval_samples": 20735,
+    "eval_samples_per_second": 142.813,
+    "eval_steps_per_second": 17.853,
+    "perplexity": 28.393349017259617
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:805d541d7c4e7f33a2b387f813df049edd021f556cca165e73862604af15bc51
+size 510396521

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>"}

tokenizer_config.json

@@ -0,0 +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"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "train_loss": 3.3660196709954384,
+    "train_runtime": 188834.9175,
+    "train_samples": 1019783,
+    "train_samples_per_second": 54.004,
+    "train_steps_per_second": 2.25
+}

trainer_state.json

@@ -0,0 +1,181 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "global_step": 424910,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9999882328022406e-05,
+      "loss": 4.9809,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.5e-05,
+      "loss": 3.5631,
+      "step": 42491
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.3989329161199919,
+      "eval_loss": 3.4350521564483643,
+      "eval_runtime": 145.6618,
+      "eval_samples_per_second": 142.35,
+      "eval_steps_per_second": 17.795,
+      "step": 42491
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4e-05,
+      "loss": 3.4548,
+      "step": 84982
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.40287618978903256,
+      "eval_loss": 3.399595022201538,
+      "eval_runtime": 145.7047,
+      "eval_samples_per_second": 142.308,
+      "eval_steps_per_second": 17.789,
+      "step": 84982
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.5e-05,
+      "loss": 3.4073,
+      "step": 127473
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.40519529343857336,
+      "eval_loss": 3.382293701171875,
+      "eval_runtime": 145.8006,
+      "eval_samples_per_second": 142.215,
+      "eval_steps_per_second": 17.778,
+      "step": 127473
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3e-05,
+      "loss": 3.3747,
+      "step": 169964
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.4067526283111404,
+      "eval_loss": 3.369549512863159,
+      "eval_runtime": 146.2616,
+      "eval_samples_per_second": 141.766,
+      "eval_steps_per_second": 17.722,
+      "step": 169964
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.5e-05,
+      "loss": 3.3498,
+      "step": 212455
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.4080202176319472,
+      "eval_loss": 3.3622708320617676,
+      "eval_runtime": 145.9616,
+      "eval_samples_per_second": 142.058,
+      "eval_steps_per_second": 17.758,
+      "step": 212455
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2e-05,
+      "loss": 3.3297,
+      "step": 254946
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.4089392009022745,
+      "eval_loss": 3.35578989982605,
+      "eval_runtime": 146.2481,
+      "eval_samples_per_second": 141.78,
+      "eval_steps_per_second": 17.723,
+      "step": 254946
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.5e-05,
+      "loss": 3.3133,
+      "step": 297437
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.40958932460425806,
+      "eval_loss": 3.3527486324310303,
+      "eval_runtime": 145.9518,
+      "eval_samples_per_second": 142.067,
+      "eval_steps_per_second": 17.759,
+      "step": 297437
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1e-05,
+      "loss": 3.2995,
+      "step": 339928
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.4101536259016574,
+      "eval_loss": 3.3491063117980957,
+      "eval_runtime": 146.1545,
+      "eval_samples_per_second": 141.87,
+      "eval_steps_per_second": 17.735,
+      "step": 339928
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5e-06,
+      "loss": 3.2884,
+      "step": 382419
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.41050640914882025,
+      "eval_loss": 3.346827507019043,
+      "eval_runtime": 146.0559,
+      "eval_samples_per_second": 141.966,
+      "eval_steps_per_second": 17.747,
+      "step": 382419
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 3.2796,
+      "step": 424910
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.41066931981946914,
+      "eval_loss": 3.3461549282073975,
+      "eval_runtime": 145.5642,
+      "eval_samples_per_second": 142.446,
+      "eval_steps_per_second": 17.807,
+      "step": 424910
+    },
+    {
+      "epoch": 10.0,
+      "step": 424910,
+      "total_flos": 6.6615293067264e+17,
+      "train_loss": 3.3660196709954384,
+      "train_runtime": 188834.9175,
+      "train_samples_per_second": 54.004,
+      "train_steps_per_second": 2.25
+    }
+  ],
+  "max_steps": 424910,
+  "num_train_epochs": 10,
+  "total_flos": 6.6615293067264e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d42b6fe65522ba31e1dceb78cea3e0d262179912422cc097aefc934241775e7a
+size 3119