Training in progress, step 48000, checkpoint

checkpoint-48000/config.json

@@ -0,0 +1,28 @@
+{
+  "architectures": [
+    "RobertaForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 256,
+  "initializer_range": 0.02,
+  "intermediate_size": 2048,
+  "layer_norm_eps": 1e-05,
+  "mask_token_id": 4,
+  "max_position_embeddings": 512,
+  "model_type": "roberta",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 8,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 8192
+}

checkpoint-48000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54f15eeaffb61ce7d3f6d1bb484a4b1884ae410f72104b5c1e7782bf07552a8c
+size 59702184

checkpoint-48000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b09234b3ea71bb1276636c25e1d5ee98fad4ec2509f90a4888d4162474b65362
+size 119488058

checkpoint-48000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9efe6b6fe2bc613219cb860caf84a3f34d1d14e290e1614bbeacc69d991ea585
+size 14308

checkpoint-48000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27393a53852a74079c71db37b8fa4a5345694eeb5f709b6db29dcd6c216dc42d
+size 1000

checkpoint-48000/special_tokens_map.json

@@ -0,0 +1,40 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>"
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-48000/tokenizer_config.json

@@ -0,0 +1,63 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "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": "<pad>",
+      "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": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>"
+  ],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "model_max_length": 128,
+  "pad_token": "<pad>",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>"
+}

checkpoint-48000/trainer_state.json

@@ -0,0 +1,309 @@
+{
+  "best_metric": 2.1982405185699463,
+  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_last/fr_mlm/fr_childes_13/checkpoint-48000",
+  "epoch": 60.0,
+  "eval_steps": 2000,
+  "global_step": 48000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2.5,
+      "eval_loss": 7.2409749031066895,
+      "eval_runtime": 1.6336,
+      "eval_samples_per_second": 1264.105,
+      "eval_steps_per_second": 79.58,
+      "step": 2000
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.1403313875198364,
+      "learning_rate": 1e-05,
+      "loss": 7.2222,
+      "step": 4000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 5.968332767486572,
+      "eval_runtime": 1.6315,
+      "eval_samples_per_second": 1265.736,
+      "eval_steps_per_second": 79.683,
+      "step": 4000
+    },
+    {
+      "epoch": 7.5,
+      "eval_loss": 5.824122428894043,
+      "eval_runtime": 1.6212,
+      "eval_samples_per_second": 1273.756,
+      "eval_steps_per_second": 80.188,
+      "step": 6000
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.6872384548187256,
+      "learning_rate": 2e-05,
+      "loss": 5.6899,
+      "step": 8000
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 5.726555824279785,
+      "eval_runtime": 1.6918,
+      "eval_samples_per_second": 1220.594,
+      "eval_steps_per_second": 76.841,
+      "step": 8000
+    },
+    {
+      "epoch": 12.5,
+      "eval_loss": 5.629961013793945,
+      "eval_runtime": 1.6387,
+      "eval_samples_per_second": 1260.18,
+      "eval_steps_per_second": 79.333,
+      "step": 10000
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 2.588118076324463,
+      "learning_rate": 3e-05,
+      "loss": 5.4663,
+      "step": 12000
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 5.5523834228515625,
+      "eval_runtime": 1.6106,
+      "eval_samples_per_second": 1282.116,
+      "eval_steps_per_second": 80.714,
+      "step": 12000
+    },
+    {
+      "epoch": 17.5,
+      "eval_loss": 5.512193202972412,
+      "eval_runtime": 1.5999,
+      "eval_samples_per_second": 1290.687,
+      "eval_steps_per_second": 81.254,
+      "step": 14000
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 4.425747871398926,
+      "learning_rate": 4e-05,
+      "loss": 5.3077,
+      "step": 16000
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 5.460000991821289,
+      "eval_runtime": 1.6043,
+      "eval_samples_per_second": 1287.195,
+      "eval_steps_per_second": 81.034,
+      "step": 16000
+    },
+    {
+      "epoch": 22.5,
+      "eval_loss": 5.401474475860596,
+      "eval_runtime": 1.6476,
+      "eval_samples_per_second": 1253.315,
+      "eval_steps_per_second": 78.901,
+      "step": 18000
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 4.271786212921143,
+      "learning_rate": 5e-05,
+      "loss": 5.1915,
+      "step": 20000
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 5.336738586425781,
+      "eval_runtime": 1.5922,
+      "eval_samples_per_second": 1296.931,
+      "eval_steps_per_second": 81.647,
+      "step": 20000
+    },
+    {
+      "epoch": 27.5,
+      "eval_loss": 5.1496429443359375,
+      "eval_runtime": 2.7104,
+      "eval_samples_per_second": 761.889,
+      "eval_steps_per_second": 47.964,
+      "step": 22000
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 5.581806182861328,
+      "learning_rate": 6e-05,
+      "loss": 4.8173,
+      "step": 24000
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 4.08081579208374,
+      "eval_runtime": 3.0759,
+      "eval_samples_per_second": 671.348,
+      "eval_steps_per_second": 42.264,
+      "step": 24000
+    },
+    {
+      "epoch": 32.5,
+      "eval_loss": 3.3694963455200195,
+      "eval_runtime": 2.9664,
+      "eval_samples_per_second": 696.136,
+      "eval_steps_per_second": 43.825,
+      "step": 26000
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 4.369658470153809,
+      "learning_rate": 7e-05,
+      "loss": 3.387,
+      "step": 28000
+    },
+    {
+      "epoch": 35.0,
+      "eval_loss": 3.072216510772705,
+      "eval_runtime": 3.0861,
+      "eval_samples_per_second": 669.14,
+      "eval_steps_per_second": 42.125,
+      "step": 28000
+    },
+    {
+      "epoch": 37.5,
+      "eval_loss": 2.8991713523864746,
+      "eval_runtime": 3.0259,
+      "eval_samples_per_second": 682.437,
+      "eval_steps_per_second": 42.962,
+      "step": 30000
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 4.331390380859375,
+      "learning_rate": 8e-05,
+      "loss": 2.8035,
+      "step": 32000
+    },
+    {
+      "epoch": 40.0,
+      "eval_loss": 2.727933168411255,
+      "eval_runtime": 3.0386,
+      "eval_samples_per_second": 679.589,
+      "eval_steps_per_second": 42.783,
+      "step": 32000
+    },
+    {
+      "epoch": 42.5,
+      "eval_loss": 2.61684513092041,
+      "eval_runtime": 3.0184,
+      "eval_samples_per_second": 684.128,
+      "eval_steps_per_second": 43.069,
+      "step": 34000
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 3.8377530574798584,
+      "learning_rate": 9e-05,
+      "loss": 2.531,
+      "step": 36000
+    },
+    {
+      "epoch": 45.0,
+      "eval_loss": 2.557036876678467,
+      "eval_runtime": 3.0738,
+      "eval_samples_per_second": 671.798,
+      "eval_steps_per_second": 42.292,
+      "step": 36000
+    },
+    {
+      "epoch": 47.5,
+      "eval_loss": 2.4636330604553223,
+      "eval_runtime": 2.9501,
+      "eval_samples_per_second": 699.986,
+      "eval_steps_per_second": 44.067,
+      "step": 38000
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 3.9490303993225098,
+      "learning_rate": 0.0001,
+      "loss": 2.3359,
+      "step": 40000
+    },
+    {
+      "epoch": 50.0,
+      "eval_loss": 2.419264793395996,
+      "eval_runtime": 3.1681,
+      "eval_samples_per_second": 651.813,
+      "eval_steps_per_second": 41.034,
+      "step": 40000
+    },
+    {
+      "epoch": 52.5,
+      "eval_loss": 2.352754592895508,
+      "eval_runtime": 2.6693,
+      "eval_samples_per_second": 773.624,
+      "eval_steps_per_second": 48.703,
+      "step": 42000
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 3.742015838623047,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 2.1893,
+      "step": 44000
+    },
+    {
+      "epoch": 55.0,
+      "eval_loss": 2.299854278564453,
+      "eval_runtime": 1.7519,
+      "eval_samples_per_second": 1178.69,
+      "eval_steps_per_second": 74.203,
+      "step": 44000
+    },
+    {
+      "epoch": 57.5,
+      "eval_loss": 2.2412161827087402,
+      "eval_runtime": 2.9705,
+      "eval_samples_per_second": 695.158,
+      "eval_steps_per_second": 43.763,
+      "step": 46000
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 3.492009162902832,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 2.0715,
+      "step": 48000
+    },
+    {
+      "epoch": 60.0,
+      "eval_loss": 2.1982405185699463,
+      "eval_runtime": 2.9731,
+      "eval_samples_per_second": 694.561,
+      "eval_steps_per_second": 43.725,
+      "step": 48000
+    }
+  ],
+  "logging_steps": 4000,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 125,
+  "save_steps": 4000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4972960636928e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-48000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f019ce7e3a851c79459abf282e5a1f1c8e1b3e048abd2b918873801766590a02
+size 5368