Training in progress, step 32000, checkpoint

checkpoint-32000/config.json

@@ -0,0 +1,39 @@
+{
+  "_name_or_path": "gpt2",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 5,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 4,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 128,
+  "n_embd": 128,
+  "n_head": 8,
+  "n_inner": null,
+  "n_layer": 8,
+  "n_positions": 128,
+  "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": "4.45.2",
+  "use_cache": true,
+  "vocab_size": 16384
+}

checkpoint-32000/generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 5,
+  "eos_token_id": 4,
+  "transformers_version": "4.45.2"
+}

checkpoint-32000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4194b324ea026afa1a62fde3f7d71ef1c86a740104f74d94745861639e3940e8
+size 14809584

checkpoint-32000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:745c678ccf788851ad3ad87505d1b3d95d5975468529efa4236a0edf8f4ccddb
+size 29680378

checkpoint-32000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4985f7362bd4bd22959ff1a6005f8640f99b5a34d77d9c5d34805973aff15c5e
+size 13990

checkpoint-32000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1444ea66cbbad7365e9a9bfc0a747283eaf112fe494318ef8d362428d5c2c86a
+size 1192

checkpoint-32000/special_tokens_map.json

@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "[BOS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "[EOS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "[EOS]"
+}

checkpoint-32000/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[EOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "[BOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "[BOS]",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "[EOS]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[EOS]",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

checkpoint-32000/trainer_state.json

@@ -0,0 +1,282 @@
+{
+  "best_metric": 3.487799882888794,
+  "best_model_checkpoint": "/Users/frapadovani/Desktop/babyLM_controlled/models_trained/convergence_french/random_sentence_french/checkpoint-32000",
+  "epoch": 0.9475024427797353,
+  "eval_steps": 2000,
+  "global_step": 32000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.059218902673733455,
+      "grad_norm": 1.0812722444534302,
+      "learning_rate": 0.0001,
+      "loss": 4.7523,
+      "step": 2000
+    },
+    {
+      "epoch": 0.059218902673733455,
+      "eval_loss": 4.0236735343933105,
+      "eval_runtime": 5.173,
+      "eval_samples_per_second": 425.477,
+      "eval_steps_per_second": 26.677,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11843780534746691,
+      "grad_norm": 1.2557106018066406,
+      "learning_rate": 0.0001,
+      "loss": 3.8958,
+      "step": 4000
+    },
+    {
+      "epoch": 0.11843780534746691,
+      "eval_loss": 3.8087611198425293,
+      "eval_runtime": 5.2666,
+      "eval_samples_per_second": 417.913,
+      "eval_steps_per_second": 26.203,
+      "step": 4000
+    },
+    {
+      "epoch": 0.17765670802120037,
+      "grad_norm": 1.3326870203018188,
+      "learning_rate": 0.0001,
+      "loss": 3.7344,
+      "step": 6000
+    },
+    {
+      "epoch": 0.17765670802120037,
+      "eval_loss": 3.704986333847046,
+      "eval_runtime": 5.2277,
+      "eval_samples_per_second": 421.023,
+      "eval_steps_per_second": 26.398,
+      "step": 6000
+    },
+    {
+      "epoch": 0.23687561069493382,
+      "grad_norm": 1.38993239402771,
+      "learning_rate": 0.0001,
+      "loss": 3.6379,
+      "step": 8000
+    },
+    {
+      "epoch": 0.23687561069493382,
+      "eval_loss": 3.6455252170562744,
+      "eval_runtime": 5.1485,
+      "eval_samples_per_second": 427.5,
+      "eval_steps_per_second": 26.804,
+      "step": 8000
+    },
+    {
+      "epoch": 0.29609451336866727,
+      "grad_norm": 1.3884963989257812,
+      "learning_rate": 0.0001,
+      "loss": 3.5687,
+      "step": 10000
+    },
+    {
+      "epoch": 0.29609451336866727,
+      "eval_loss": 3.6062421798706055,
+      "eval_runtime": 5.1466,
+      "eval_samples_per_second": 427.661,
+      "eval_steps_per_second": 26.814,
+      "step": 10000
+    },
+    {
+      "epoch": 0.35531341604240074,
+      "grad_norm": 1.435062289237976,
+      "learning_rate": 0.0001,
+      "loss": 3.5165,
+      "step": 12000
+    },
+    {
+      "epoch": 0.35531341604240074,
+      "eval_loss": 3.5764389038085938,
+      "eval_runtime": 5.132,
+      "eval_samples_per_second": 428.881,
+      "eval_steps_per_second": 26.89,
+      "step": 12000
+    },
+    {
+      "epoch": 0.41453231871613416,
+      "grad_norm": 1.3539327383041382,
+      "learning_rate": 0.0001,
+      "loss": 3.4738,
+      "step": 14000
+    },
+    {
+      "epoch": 0.41453231871613416,
+      "eval_loss": 3.555619239807129,
+      "eval_runtime": 5.1586,
+      "eval_samples_per_second": 426.667,
+      "eval_steps_per_second": 26.752,
+      "step": 14000
+    },
+    {
+      "epoch": 0.47375122138986764,
+      "grad_norm": 1.3827875852584839,
+      "learning_rate": 0.0001,
+      "loss": 3.4389,
+      "step": 16000
+    },
+    {
+      "epoch": 0.47375122138986764,
+      "eval_loss": 3.5388290882110596,
+      "eval_runtime": 5.8029,
+      "eval_samples_per_second": 379.292,
+      "eval_steps_per_second": 23.781,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5329701240636011,
+      "grad_norm": 1.316589593887329,
+      "learning_rate": 0.0001,
+      "loss": 3.4067,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5329701240636011,
+      "eval_loss": 3.5234413146972656,
+      "eval_runtime": 5.1383,
+      "eval_samples_per_second": 428.353,
+      "eval_steps_per_second": 26.857,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5921890267373345,
+      "grad_norm": 1.3910475969314575,
+      "learning_rate": 0.0001,
+      "loss": 3.3778,
+      "step": 20000
+    },
+    {
+      "epoch": 0.5921890267373345,
+      "eval_loss": 3.5121970176696777,
+      "eval_runtime": 7.6572,
+      "eval_samples_per_second": 287.442,
+      "eval_steps_per_second": 18.022,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6514079294110681,
+      "grad_norm": 1.510622262954712,
+      "learning_rate": 0.0001,
+      "loss": 3.3511,
+      "step": 22000
+    },
+    {
+      "epoch": 0.6514079294110681,
+      "eval_loss": 3.5116806030273438,
+      "eval_runtime": 5.6879,
+      "eval_samples_per_second": 386.964,
+      "eval_steps_per_second": 24.262,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7106268320848015,
+      "grad_norm": 1.4063044786453247,
+      "learning_rate": 0.0001,
+      "loss": 3.3309,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7106268320848015,
+      "eval_loss": 3.5099916458129883,
+      "eval_runtime": 5.2486,
+      "eval_samples_per_second": 419.351,
+      "eval_steps_per_second": 26.293,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7698457347585349,
+      "grad_norm": 1.459807276725769,
+      "learning_rate": 0.0001,
+      "loss": 3.309,
+      "step": 26000
+    },
+    {
+      "epoch": 0.7698457347585349,
+      "eval_loss": 3.498032808303833,
+      "eval_runtime": 5.3512,
+      "eval_samples_per_second": 411.307,
+      "eval_steps_per_second": 25.788,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8290646374322683,
+      "grad_norm": 1.4196616411209106,
+      "learning_rate": 0.0001,
+      "loss": 3.2905,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8290646374322683,
+      "eval_loss": 3.4919426441192627,
+      "eval_runtime": 5.2649,
+      "eval_samples_per_second": 418.049,
+      "eval_steps_per_second": 26.211,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8882835401060019,
+      "grad_norm": 1.431018352508545,
+      "learning_rate": 0.0001,
+      "loss": 3.2685,
+      "step": 30000
+    },
+    {
+      "epoch": 0.8882835401060019,
+      "eval_loss": 3.492326021194458,
+      "eval_runtime": 5.1879,
+      "eval_samples_per_second": 424.254,
+      "eval_steps_per_second": 26.6,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9475024427797353,
+      "grad_norm": 1.4345752000808716,
+      "learning_rate": 0.0001,
+      "loss": 3.2518,
+      "step": 32000
+    },
+    {
+      "epoch": 0.9475024427797353,
+      "eval_loss": 3.487799882888794,
+      "eval_runtime": 5.3438,
+      "eval_samples_per_second": 411.877,
+      "eval_steps_per_second": 25.824,
+      "step": 32000
+    }
+  ],
+  "logging_steps": 2000,
+  "max_steps": 33773,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 623810445312000.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-32000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fce3a8508ad2aa1c0ae1cc54c6fe1747e2312953499cad9c035bc201c305894e
+size 5496