Training in progress, epoch 3, checkpoint

checkpoint-6537/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<|endoftext|>": 50256
+}

checkpoint-6537/config.json

@@ -0,0 +1,42 @@
+{
+  "_name_or_path": "gpt2-xl",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 50256,
+  "do_sample": true,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 50256,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "max_length": 50,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 1600,
+  "n_head": 25,
+  "n_inner": null,
+  "n_layer": 48,
+  "n_positions": 1024,
+  "output_past": true,
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50
+    }
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.34.0",
+  "use_cache": true,
+  "vocab_size": 50257
+}

checkpoint-6537/generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 50256,
+  "do_sample": true,
+  "eos_token_id": 50256,
+  "max_length": 50,
+  "transformers_version": "4.34.0"
+}

checkpoint-6537/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59181e7f50d24e4b7b53a7b658e96a99690fe73df8f18ae0a6282d85242ca19e
+size 12461385454

checkpoint-6537/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf0f6e2424009cd3c2c96bd30bb79ba88d40def697e215e63dc9c597c9071807
+size 6230637102

checkpoint-6537/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3163cea678e52a4c13fb1775f2cba3cbd6720b9fd8efacdab1a6d7b041e55661
+size 14244

checkpoint-6537/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d3990df3adf97b65b16ec73c3e09c0b361ce3192012b2ec1de31719f5350efd
+size 1064

checkpoint-6537/special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>",
+  "unk_token": "<|endoftext|>"
+}

checkpoint-6537/tokenizer_config.json

@@ -0,0 +1,20 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "50256": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 1024,
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}

checkpoint-6537/trainer_state.json

@@ -0,0 +1,160 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 2179,
+  "global_step": 6537,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2,
+      "learning_rate": 3e-05,
+      "loss": 2.2464,
+      "step": 436
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3e-05,
+      "loss": 2.0544,
+      "step": 872
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3e-05,
+      "loss": 1.9968,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3e-05,
+      "loss": 1.9537,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.6693020521036767,
+      "eval_loss": 1.5623117685317993,
+      "eval_runtime": 220.2436,
+      "eval_samples_per_second": 33.622,
+      "eval_steps_per_second": 2.102,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 10.24983119513842,
+      "eval_f1": 15.257097040945366,
+      "eval_qa_bleu": 5.8382901471491575,
+      "eval_qa_exact_match": 0.09736664415935178,
+      "eval_recite_bleu": 17.0930417063702,
+      "eval_recite_exact_match": 0.0,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3e-05,
+      "loss": 1.9029,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3e-05,
+      "loss": 1.5215,
+      "step": 2616
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3e-05,
+      "loss": 1.5077,
+      "step": 3052
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3e-05,
+      "loss": 1.5016,
+      "step": 3488
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3e-05,
+      "loss": 1.4734,
+      "step": 3924
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6923618412861704,
+      "eval_loss": 1.2099343538284302,
+      "eval_runtime": 222.1542,
+      "eval_samples_per_second": 33.333,
+      "eval_steps_per_second": 2.084,
+      "step": 4358
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 10.61444969615125,
+      "eval_f1": 15.935672519396675,
+      "eval_qa_bleu": 6.524750946372503,
+      "eval_qa_exact_match": 0.10209318028359217,
+      "eval_recite_bleu": 18.618913529891735,
+      "eval_recite_exact_match": 0.0,
+      "step": 4358
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3e-05,
+      "loss": 1.4425,
+      "step": 4360
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3e-05,
+      "loss": 1.0827,
+      "step": 4796
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3e-05,
+      "loss": 1.0876,
+      "step": 5232
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3e-05,
+      "loss": 1.0887,
+      "step": 5668
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3e-05,
+      "loss": 1.0665,
+      "step": 6104
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7147465927772421,
+      "eval_loss": 0.9177509546279907,
+      "eval_runtime": 226.6441,
+      "eval_samples_per_second": 32.672,
+      "eval_steps_per_second": 2.043,
+      "step": 6537
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 12.397029034436192,
+      "eval_f1": 17.721572805797255,
+      "eval_qa_bleu": 8.139751517788653,
+      "eval_qa_exact_match": 0.11978392977717758,
+      "eval_recite_bleu": 21.769859861022653,
+      "eval_recite_exact_match": 0.0,
+      "step": 6537
+    }
+  ],
+  "logging_steps": 436,
+  "max_steps": 43580,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "total_flos": 3.020425488509952e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-6537/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3154d552425d8cb1b29494ee35253674917cf60979727c261838069c8f542789
+size 4728