Training in progress, step 200, checkpoint

checkpoint-200/config.json

@@ -0,0 +1,28 @@
+{
+  "_remove_final_layer_norm": false,
+  "activation_function": "relu",
+  "architectures": [
+    "OPTForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 2,
+  "do_layer_norm_before": true,
+  "dropout": 0.1,
+  "enable_bias": true,
+  "eos_token_id": 2,
+  "ffn_dim": 768,
+  "hidden_size": 768,
+  "init_std": 0.02,
+  "layer_norm_elementwise_affine": true,
+  "layerdrop": 0.0,
+  "max_position_embeddings": 512,
+  "model_type": "opt",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 8,
+  "pad_token_id": 1,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.3",
+  "use_cache": true,
+  "vocab_size": 50002,
+  "word_embed_proj_dim": 768
+}

checkpoint-200/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 2,
+  "eos_token_id": 2,
+  "pad_token_id": 1,
+  "transformers_version": "4.33.3"
+}

checkpoint-200/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fc2afa9f63785ea6aa02b4e57948aa316af5cffcb50f64b3660c8af59e68f15
+size 537476677

checkpoint-200/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08dbc1c5d49493a0d06863cbf2a7b50f5cdbd23169325c93244ba5a629269efa
+size 268727709

checkpoint-200/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20ba500168fc0f5eacf7f136eee72062671d004ec080c3fcd90a0580d8cde697
+size 14575

checkpoint-200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dab61d1012b849a4598f4dcd9b7c696f541cffa2c75575f5c0567a5c576492d8
+size 627

checkpoint-200/special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "<eos>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

checkpoint-200/tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<eos>",
+  "max_length": null,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_to_multiple_of": null,
+  "pad_token": "<pad>",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>"
+}

checkpoint-200/trainer_state.json

@@ -0,0 +1,299 @@
+{
+  "best_metric": 6.493106365203857,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/lm-training/src/outputs/2023-11-15/09-23-00/checkpoints/checkpoint-200",
+  "epoch": 1.6,
+  "eval_steps": 10,
+  "global_step": 200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.866666666666667e-05,
+      "loss": 9.9717,
+      "step": 10
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 9.324381828308105,
+      "eval_runtime": 2.0042,
+      "eval_samples_per_second": 498.949,
+      "eval_steps_per_second": 62.369,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7333333333333336e-05,
+      "loss": 9.1002,
+      "step": 20
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 8.892333984375,
+      "eval_runtime": 1.9424,
+      "eval_samples_per_second": 514.829,
+      "eval_steps_per_second": 64.354,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 8.7625,
+      "step": 30
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 8.54922103881836,
+      "eval_runtime": 1.9052,
+      "eval_samples_per_second": 524.878,
+      "eval_steps_per_second": 65.61,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.466666666666667e-05,
+      "loss": 8.4987,
+      "step": 40
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 8.234408378601074,
+      "eval_runtime": 2.0895,
+      "eval_samples_per_second": 478.585,
+      "eval_steps_per_second": 59.823,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 8.2485,
+      "step": 50
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 7.941398620605469,
+      "eval_runtime": 2.1108,
+      "eval_samples_per_second": 473.76,
+      "eval_steps_per_second": 59.22,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.2e-05,
+      "loss": 8.0802,
+      "step": 60
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 7.663424968719482,
+      "eval_runtime": 1.991,
+      "eval_samples_per_second": 502.251,
+      "eval_steps_per_second": 62.781,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.066666666666667e-05,
+      "loss": 7.7924,
+      "step": 70
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 7.419986724853516,
+      "eval_runtime": 1.9753,
+      "eval_samples_per_second": 506.241,
+      "eval_steps_per_second": 63.28,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.933333333333333e-05,
+      "loss": 7.5139,
+      "step": 80
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 7.188984394073486,
+      "eval_runtime": 1.9721,
+      "eval_samples_per_second": 507.08,
+      "eval_steps_per_second": 63.385,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8e-05,
+      "loss": 7.3932,
+      "step": 90
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 7.042924880981445,
+      "eval_runtime": 2.0739,
+      "eval_samples_per_second": 482.185,
+      "eval_steps_per_second": 60.273,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 7.334,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 6.937138557434082,
+      "eval_runtime": 2.0003,
+      "eval_samples_per_second": 499.914,
+      "eval_steps_per_second": 62.489,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.5333333333333336e-05,
+      "loss": 7.1699,
+      "step": 110
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 6.849837779998779,
+      "eval_runtime": 2.0147,
+      "eval_samples_per_second": 496.356,
+      "eval_steps_per_second": 62.045,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 7.2476,
+      "step": 120
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 6.779170036315918,
+      "eval_runtime": 2.0584,
+      "eval_samples_per_second": 485.82,
+      "eval_steps_per_second": 60.727,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.266666666666667e-05,
+      "loss": 6.8774,
+      "step": 130
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 6.709743022918701,
+      "eval_runtime": 2.4463,
+      "eval_samples_per_second": 408.786,
+      "eval_steps_per_second": 51.098,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.1333333333333334e-05,
+      "loss": 6.7219,
+      "step": 140
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 6.656557559967041,
+      "eval_runtime": 2.0119,
+      "eval_samples_per_second": 497.045,
+      "eval_steps_per_second": 62.131,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3e-05,
+      "loss": 6.6353,
+      "step": 150
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 6.612905502319336,
+      "eval_runtime": 2.0567,
+      "eval_samples_per_second": 486.217,
+      "eval_steps_per_second": 60.777,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8666666666666668e-05,
+      "loss": 6.6524,
+      "step": 160
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 6.580767631530762,
+      "eval_runtime": 1.9849,
+      "eval_samples_per_second": 503.794,
+      "eval_steps_per_second": 62.974,
+      "step": 160
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.733333333333333e-05,
+      "loss": 6.484,
+      "step": 170
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 6.5504679679870605,
+      "eval_runtime": 2.1014,
+      "eval_samples_per_second": 475.877,
+      "eval_steps_per_second": 59.485,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 6.5846,
+      "step": 180
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 6.531940460205078,
+      "eval_runtime": 3.1229,
+      "eval_samples_per_second": 320.22,
+      "eval_steps_per_second": 40.027,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.466666666666667e-05,
+      "loss": 6.558,
+      "step": 190
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 6.510298728942871,
+      "eval_runtime": 1.9686,
+      "eval_samples_per_second": 507.978,
+      "eval_steps_per_second": 63.497,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 6.4792,
+      "step": 200
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 6.493106365203857,
+      "eval_runtime": 1.8922,
+      "eval_samples_per_second": 528.475,
+      "eval_steps_per_second": 66.059,
+      "step": 200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_train_epochs": 3,
+  "save_steps": 10,
+  "total_flos": 23698405785600.0,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:343123aa056dacb81bf4aba5d1ee827cb26f5e4dd6a565a320e1d7e0430ea370
+size 4155