Training in progress, step 7500, checkpoint

checkpoint-7500/config.json

@@ -0,0 +1,33 @@
+{
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 1,
+  "dtype": "float32",
+  "embd_pdrop": 0.1,
+  "eos_token_id": 2,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 768,
+  "n_head": 12,
+  "n_inner": null,
+  "n_layer": 12,
+  "n_positions": 1024,
+  "pad_token_id": 2,
+  "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,
+  "transformers_version": "4.56.1",
+  "use_cache": true,
+  "vocab_size": 30000
+}

checkpoint-7500/generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": [
+    2
+  ],
+  "pad_token_id": 2,
+  "transformers_version": "4.56.1"
+}

checkpoint-7500/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bc8b2ec8fc08900b6377d3f829cfe132bff7e280eb3d597021dcf38ec1a9ae9
+size 435544704

checkpoint-7500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:084479db9a02b4c18460a783da95da07905f00f10b5015adb4f39b4096596a9d
+size 871183627

checkpoint-7500/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edab732a20560cb4435069ee4df15a46a5861803767816825ddffefad711c590
+size 14645

checkpoint-7500/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:642827d515e7c3d88d20c04224822701abbbc9250deb1d366045f974d08b8258
+size 1383

checkpoint-7500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca9b8a5501507d40ffae7081184eb5717a9cb61e1a824d7d0f7f2087efd5ab69
+size 1465

checkpoint-7500/special_tokens_map.json

@@ -0,0 +1,8 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "</s>",
+  "sep_token": "<sep>",
+  "unk_token": "<unk>"
+}

checkpoint-7500/tokenizer_config.json

@@ -0,0 +1,62 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<sep>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sep_token": "<sep>",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>"
+}

checkpoint-7500/trainer_state.json

@@ -0,0 +1,311 @@
+{
+  "best_global_step": 5200,
+  "best_metric": 5.385416507720947,
+  "best_model_checkpoint": null,
+  "epoch": 18.75,
+  "eval_steps": 500,
+  "global_step": 7500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0025,
+      "grad_norm": 15.394726753234863,
+      "learning_rate": 0.0,
+      "loss": 10.3867,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4460233449935913,
+      "learning_rate": 7.98e-05,
+      "loss": 7.063,
+      "step": 400
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 6.196335315704346,
+      "eval_runtime": 15.8434,
+      "eval_samples_per_second": 1621.435,
+      "eval_steps_per_second": 6.375,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2636526823043823,
+      "learning_rate": 9.600267379679145e-05,
+      "loss": 5.9603,
+      "step": 800
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 5.856795310974121,
+      "eval_runtime": 15.8854,
+      "eval_samples_per_second": 1617.144,
+      "eval_steps_per_second": 6.358,
+      "step": 800
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.307215690612793,
+      "learning_rate": 9.065508021390375e-05,
+      "loss": 5.7367,
+      "step": 1200
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 5.718450546264648,
+      "eval_runtime": 15.8896,
+      "eval_samples_per_second": 1616.719,
+      "eval_steps_per_second": 6.356,
+      "step": 1200
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.3407551050186157,
+      "learning_rate": 8.530748663101604e-05,
+      "loss": 5.6043,
+      "step": 1600
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 5.625463962554932,
+      "eval_runtime": 15.8287,
+      "eval_samples_per_second": 1622.94,
+      "eval_steps_per_second": 6.381,
+      "step": 1600
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.3907631635665894,
+      "learning_rate": 7.995989304812834e-05,
+      "loss": 5.4977,
+      "step": 2000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 5.553786277770996,
+      "eval_runtime": 15.879,
+      "eval_samples_per_second": 1617.802,
+      "eval_steps_per_second": 6.361,
+      "step": 2000
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.4962470531463623,
+      "learning_rate": 7.461229946524065e-05,
+      "loss": 5.405,
+      "step": 2400
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 5.4988250732421875,
+      "eval_runtime": 16.1294,
+      "eval_samples_per_second": 1592.678,
+      "eval_steps_per_second": 6.262,
+      "step": 2400
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.5504164695739746,
+      "learning_rate": 6.926470588235294e-05,
+      "loss": 5.3231,
+      "step": 2800
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 5.458596706390381,
+      "eval_runtime": 15.8758,
+      "eval_samples_per_second": 1618.122,
+      "eval_steps_per_second": 6.362,
+      "step": 2800
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.695698618888855,
+      "learning_rate": 6.391711229946524e-05,
+      "loss": 5.2489,
+      "step": 3200
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 5.429712295532227,
+      "eval_runtime": 15.9068,
+      "eval_samples_per_second": 1614.967,
+      "eval_steps_per_second": 6.349,
+      "step": 3200
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.746219277381897,
+      "learning_rate": 5.8569518716577545e-05,
+      "loss": 5.1805,
+      "step": 3600
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 5.407948017120361,
+      "eval_runtime": 15.8959,
+      "eval_samples_per_second": 1616.082,
+      "eval_steps_per_second": 6.354,
+      "step": 3600
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.9468352794647217,
+      "learning_rate": 5.322192513368984e-05,
+      "loss": 5.1167,
+      "step": 4000
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 5.395551681518555,
+      "eval_runtime": 15.9345,
+      "eval_samples_per_second": 1612.162,
+      "eval_steps_per_second": 6.338,
+      "step": 4000
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 2.0324604511260986,
+      "learning_rate": 4.787433155080214e-05,
+      "loss": 5.0566,
+      "step": 4400
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 5.387964725494385,
+      "eval_runtime": 15.9346,
+      "eval_samples_per_second": 1612.149,
+      "eval_steps_per_second": 6.338,
+      "step": 4400
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 2.1527271270751953,
+      "learning_rate": 4.2526737967914444e-05,
+      "loss": 4.9995,
+      "step": 4800
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 5.385544300079346,
+      "eval_runtime": 15.9015,
+      "eval_samples_per_second": 1615.511,
+      "eval_steps_per_second": 6.352,
+      "step": 4800
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 2.369941234588623,
+      "learning_rate": 3.717914438502674e-05,
+      "loss": 4.9456,
+      "step": 5200
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 5.385416507720947,
+      "eval_runtime": 15.9781,
+      "eval_samples_per_second": 1607.767,
+      "eval_steps_per_second": 6.321,
+      "step": 5200
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 2.4394657611846924,
+      "learning_rate": 3.1831550802139034e-05,
+      "loss": 4.8941,
+      "step": 5600
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 5.389502048492432,
+      "eval_runtime": 15.8441,
+      "eval_samples_per_second": 1621.358,
+      "eval_steps_per_second": 6.375,
+      "step": 5600
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 2.6221773624420166,
+      "learning_rate": 2.6483957219251336e-05,
+      "loss": 4.8489,
+      "step": 6000
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 5.398255348205566,
+      "eval_runtime": 15.9582,
+      "eval_samples_per_second": 1609.767,
+      "eval_steps_per_second": 6.329,
+      "step": 6000
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.7822163105010986,
+      "learning_rate": 2.1136363636363638e-05,
+      "loss": 4.8092,
+      "step": 6400
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 5.399804592132568,
+      "eval_runtime": 15.931,
+      "eval_samples_per_second": 1612.519,
+      "eval_steps_per_second": 6.34,
+      "step": 6400
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.907365083694458,
+      "learning_rate": 1.5788770053475936e-05,
+      "loss": 4.7745,
+      "step": 6800
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 5.40533447265625,
+      "eval_runtime": 15.9329,
+      "eval_samples_per_second": 1612.321,
+      "eval_steps_per_second": 6.339,
+      "step": 6800
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 2.930150270462036,
+      "learning_rate": 1.0441176470588236e-05,
+      "loss": 4.745,
+      "step": 7200
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 5.412296295166016,
+      "eval_runtime": 15.9489,
+      "eval_samples_per_second": 1610.704,
+      "eval_steps_per_second": 6.333,
+      "step": 7200
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 8000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.25169726947328e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-7500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0272e76bd6120910c15c2c932a2fa3c639ef9fa59dac7faba46b78396af23c47
+size 5905